<<< TOPAZ::$255$DUA108:[NOTES$LIBRARY]ARTICLES.NOTE;1 >>>
-< Welcome to the Articles conference >-
================================================================================
Note 36.23 The DECUServe Journal - Internet Version 23 of 23
EISNER::MCMAHON_B "DECUServe Journal" 3729 lines 15-DEC-1994 18:52
-< December 1994 >-
--------------------------------------------------------------------------------
The DECUServe Journal
=====================
December, 1994
From the Editor's Keyboard
==========================
As announced, this issue of the DECUServe Journal is devoted to
obscure, hard-to-find, and vital scraps of information on older (or
if you prefer, "mature") DEC systems. By that, we (very broadly)
mean 5- to 10-year-old equipment, such as the MicroVAX II (but
certainly not limited to VAX hardware) that shows up on the used
equipment market, still graces many production environments, and is
sometimes found as a home system.
DECUServe's archives are a tremendous (possibly unequaled) resource
for this older hardware, so much so that we found the volume of
information slightly overwhelming. We certainly found enough to
fill more than one special issue, so if you like this sort of thing,
let us know -- there's plenty more where this came from!
During our excavations, we were struck by the consistency of the way
DECUServe has operated through the years. Whether a series of notes
dated from 1988 or 1994 seemed to make little difference, except for
the technology being discussed. The "DECUServe style" has remained
constant. Then as now, the technical quality of information is
high, the noise level is low, the conversational tone is informal
but respectful, and it's not at all unusual to have questions
answered within a day. Many of the names are familiar, too,
although their DECUS hats may have changed several times.
Speaking of familiar names, we are very grateful to Terry Kennedy
for pointing us in the right direction when we started collecting
information for this special issue. You'll be seeing a lot of him
in the following articles. Terry, thank you for your helpful
advice... and for contributing a significant chunk of the original
information in this issue!
Concerning Journal administrivia, you may have noticed that your
faithful editors have also been rather consistently tardy editors in
recent months. Following long-established programming practice,
we've decided to document this consistency, thereby turning it into
a feature! (Insert obligatory software development joke here.)
"Editorial cut-off" for the Journal will be the first of the month,
and the finished product will appear by the 15th.
This timetable adjustment allows us to make sure recently-active
threads have really come to a conclusion (it would be a shame to
extract and publish a stream the day before someone posts the
definitive, dazzlingly brilliant solution), and gives us some leeway
to handle scheduling disasters in the real world as well (hint:
this is finals week at Grinnell College). We hope this will help us
The DECUServe Journal December, 1994 Page 2
From the Editor's Keyboard
to put the DECUServe Journal on a more reliable schedule.
As we are fond of pointing out, we welcome your comments. We're
especially interested in feedback on this dedicated-theme issue. Is
it good, bad, or ugly? Would you like to see more such issues?
What other topics (e.g., e-mail, security, internetworking) might be
worth a special issue? Contact information can be found at the end
of each issue. Let us hear your thoughts... or we'll continue
potting along with our own!
Table of Contents
=================
CONTENTS
From the Editor's Keyboard . . . . . . . . . . . . . 1
Table of Contents . . . . . . . . . . . . . . . . . 2
BA23 Fire Hazard . . . . . . . . . . . . . . . . . . 3
Disk Equivalences . . . . . . . . . . . . . . . . . 7
MicroVAX 2000 Comm Ports . . . . . . . . . . . . . 16
Old Q-bus Hardware . . . . . . . . . . . . . . . . 19
Printronix 600 Setup . . . . . . . . . . . . . . . 28
11/70 Tips . . . . . . . . . . . . . . . . . . . . 31
RK07 Tips . . . . . . . . . . . . . . . . . . . . 37
Reviving a PDP-11/24 . . . . . . . . . . . . . . . 39
Backing Up With RSX . . . . . . . . . . . . . . . 50
SCSI for RT-11 . . . . . . . . . . . . . . . . . . 55
Hard-to-find Parts . . . . . . . . . . . . . . . . 58
MicroVAX II Battery . . . . . . . . . . . . . . . 61
RL02 Spare Lightbulbs . . . . . . . . . . . . . . 64
Old Hardware Potpourri . . . . . . . . . . . . . . 66
About the DECUServe Journal . . . . . . . . . . . 70
Contact Information . . . . . . . . . . . . . . . 70
The DECUServe Journal December, 1994 Page 3
BA23 Fire Hazard
BA23 Fire Hazard
================
We'll kick off with a thread that has become something of a
DECUServe classic. This vivid illustration of what happens when
hardware goes over to the dark side was posted by Terry in the old
SOAPBOX conference back in '88. (SOAPBOX is no longer active, but
as we never tire of pointing out, the DECUServe philosophy is never
to throw away potentially useful information -- which is what makes
this special issue possible in the first place.)
If you own or are considering buying anything involving a BA23 box
(lots of MicroVAX II and smaller PDP-11 systems), you definitely
want to read this.
Participants: Jack Crowell, Alan Frisbie, Terry Kennedy, Larry
Kilgallen, Brian McAllister.
Note 82.0, 31-May-1988
Kennedy: Burned up over BA23 fire
---------------------------------
I just had a rather interesting (but not nice) thing happen in my
office - my PDP-11/23 caught on fire! This was a result of a power
cable with an insufficient rating being used (by DEC) in the BA23
box.
A FCO was issued for this problem. So far, so good. Now, things
get worse:
o This is listed as an 'improvement' FCO - so non-contract
customers have to a) know to ask for it b) pay for it
o The part number (EQ-01427-01) is an 'unlisted' part. This
means you have to go through the fun 'product demand quote'
system (see footnote for gory details). The estimated lead
time is 90-120 days.
I find the above situation *totally* unacceptable. I am contacting
my organization's insurance company (Lloyd's) to see what liability
they feel DEC has in a case where they (DEC) state in writing that
one of their products 'burns during normal use', but will not cor-
rect the problem in a timely manner.
If you have a uVAX I or a Micro-11 in a BA23, you may want to
investigate some yourself...
* Product Demand Quote - A system whereby DEC can refuse to sell you
a product while keeping up the appearance of helping you to buy the
item in question. Basically, they get your address, the part number
you want, and the quantity. They then send you (in 7-10 days) a form
which states (usually) one or more of the following:
o Part not available - re-order next higher assembly (part=BA123
caster, higher assembly=whole BA123!)
o Part not available - no substitute available (part=BA23 power
The DECUServe Journal December, 1994 Page 4
BA23 Fire Hazard
switch)
o Re-order in multiples of six (order was for five)
o Lead time 240 days - send purchase order if ordering desired
o Part available through DEC Direct - order through them (part=
KDF11-B3 ROM upgrade)
Of course, if you get a quote that will actually sell you the part,
you may discover that the price is $115 (for *one* TU80 boot ROM),
but at least you can order it! By the way, the above has been my
actual experience with the system to date.
Note 82.1, 31-May-1988
Frisbie: We've got it now -- but you can't have it
--------------------------------------------------
>> Product Demand Quote - A system whereby DEC can refuse to sell you
>> a product while keeping up the appearance of helping you to buy the
>> item in question.
Your experience matches mine exactly. Several of us beat up
DEC about this at the Symposium in Cincinnati, so maybe there
will be improvements in this area (but don't hold your breath).
Note 82.2, 1-Jun-1988
Kilgallen: Unless, of course, you do get a fire...
--------------------------------------------------
> (but don't hold your breath).
Note 82.3, 2-Jun-1988
McAllister: Your # is FCO kit, not part
---------------------------------------
I learned about this from a report brought back from the
symposium Nashville symposium. It also turned up in DECOlog,
for those whose have the bucks to spend on it.
The part number you have listed is the one for the FCO 'kit'.
The actual part number for the cable in it was 17-01311-01. According
to the info from Nashville (from a session by Jack Toto), there
were TWO ECOs:
#BA23-A-MK003: This one (dated 26-Aug-1986) replaced the cable (wire)
in the power harness.
#BA23-A-MK004: This one (dated 23-Feb-1987) replaced the connector
which did not meet the specs either.
The first one lists a "New Power cable P/N 17-01311-01".
The second one lists a "New Harness P/N 17-01356-01".
The DECUServe Journal December, 1994 Page 5
BA23 Fire Hazard
I was told by someone who attended the session that these were
MANDATORY, safety-related ECOs; that I could get replacement
power harnesses by contacting field service. Unfortunately, I
never got around to it, even though one of my BA23s melted its
power connector (no fire, I smelled the roasting plastic before
it got that far). I had a box that was dead for other reasons
and stole the cable out of it.
For those who may be nervous about this, it only becomes a problem
if you SERIOUSLY load the box, with lots of power-hungry boards.
If you want to check, open the box up and examine the connector
on the power harness. If there is any discoloration (brown or
even black) you have a problem.
Terry: I hope this info may help you get the parts you need.
The problem may be that the ECO/FCO has been out for a while,
and DEC may feel that anybody who really needed already has it.
I find it hard to believe that they are really claiming that
parts that burn because they do not meet the specifications of
the unit are not their responsibility to fix.
PS: The original FCO was listed as "Code: F" and in the info
regarding installing it, "Type of call" was "W" and "Action
Taken" was "D". For per-call customers "Activity" was "F" and
for warranty/contract customers "Activity" was "W". I don't
really know what any of this means as nothing in the stupid
library explains it.
BTW: The "affected population" is listed as 11,000. Could be why
they weren't real eager to give it to everybody.
Note 82.4, 2-Jun-1988
Kennedy:
---------
I appreciate the info. Here's my experience:
1) I learned about the problem just now in Cinci - It was mentioned
as part of the BA23 C/D harness problem.
2) Looking in my (newly acquired) DEC-O-LOG, the type code is 'I' for
improvement - so you get to pay Field Circus $400 or so to install
a $25.00 or so part.
3) I have *never* been able to get the component parts of an FCO by
ordering the part itself. Consider the following (true) case history:
Having problems with MS: boot on 11/23+
Learn there are new CPU ROMs (23-series)
Call DEC[in]Direct - unknown part - get product demand quote
PDQ arrives in non-PDQ time - 2 weeks - marked 'part not available -
request FCO kit EQ-xxxxx-xx
Get another PDQ for FCO kit
The DECUServe Journal December, 1994 Page 6
BA23 Fire Hazard
PDQ arrives in 2 weeks (again) with price 'N' - scrawled on the
bottom is a note - 'You can order this as a kit KDF11-B3 from
DECDirect'
Called DECdirect - ordered kit (kit cost was 'N' (from above)
*minus* $20 (yes, twenty bucks cheaper)
Kit arrived in 10 days (faster than a speeding PDQ). Kit included
2 ROMs, nice manual (which continues tradition of giving out
internal DEC product support numbers to customers), and a full
set of diagnostics on RX50's.
AAAAAAAAAARRRRRRRRRRRGGGGGGGGGGGGHHHHHHHHHHH!!!!!! Why couldn't
someone tell me that in the first place?
Note 82.5, 8-Jun-1988
Kennedy: Do it yourself & save...
---------------------------------
I decided that the connector scheme used is essentially unreliable
at/near the rated current of the supply. I also discovered that there
was considerable damage/high resistance on the metal fingers that
the connector slides onto.
I badgered a friend who works in a place where they make wiring
harnesses for nuclear subs & the like to get me a few feet of
Teledyne Thermionics cable. This is a double-Teflon insulated
nickel-silver heavy-duty wire. I then brazed (not soldered) it
to the connectors. Thus, the power harness has gone from the
*least* to the *most* reliable component of the system. I wonder
what's next?
Note 82.6, 5-Dec-1988
Crowell: Do I smell smoke?
--------------------------
We found out about this by the smell-the-smoke method on a heavily
loaded PDP-11/83. The machine was not under contract at the time,and
we paid through the nose for replacement harness and connectors
that the field-service tech said were being replaced whenever someone
called in with a problem. I wonder how much equipment was being
damaged in the meantime.
The DECUServe Journal December, 1994 Page 7
Disk Equivalences
Disk Equivalences
=================
You're likely to be seeing more of this next note from the
HARDWARE_HELP conference in future issues of the Journal. The topic
is "Low-end disk drives - The Digital difference", a discussion of
"generic" alternatives to various Digital-labelled drives. This
thread has displayed remarkable longevity -- it started in early
1988 and has been active as recently as August of '94. There is
probably enough information in this one note to fill an entire issue
of the Journal by itself.
This month, we've selected some of the earliest notes, dealing
mostly with RD5x-vintage MFM disk drives, for your reading pleasure.
These drives, especially the larger sizes, are commonly seen on
MicroVAX II and smaller PDP-11 systems. The drive types discussed
here show up in used-equipment listings in both DEC and non-DEC
incarnations, so shop around -- these days, sometimes the
DEC-labelled drives are cheaper!
Participants: Jerry Ethington, Jamie Hanrahan, Terry Kennedy, Jeff
Killeen, Marc Kozam, Bill Mayhew, Gary Rice.
Note 93.0, 16-Jan-1988
Kennedy: Low-end disk devices - The Digital difference
------------------------------------------------------
The disk drives used in Digital's low-end systems are generally
slightly- or un-modified versions of devices available on the
open market. DEC's markup appears to average about 800% of the
best price you can get on the open market. For some, that is not
an unreasonable amount to pay for the security of an 'all-DEC'
system. For others (such as myself) it is entirely unnecessary.
There are several companies who will do you the 'favor' of sell-
ing you a DEC 'equal' drive for *only* a 400% mark-up.
Since the PC 'explosion', prices on these equivalent units are
very low due to the cut-throat pricing of the marketplace. I
therefore set out to determine what the differences (if any)
between the DEC device and the generic device are. This series
of notes will present my discoveries.
Please feel free to use this information in any way you choose,
but remember that neither I, Digital, nor the vendor you buy
the equipment from can be responsible for problems you may en-
counter. Having said that, let me also say that I will do my
best to help you overcome any problems, and that all of the
vendors I have used offered to give me a refund if the unit did
not work in my DEC system.
The DECUServe Journal December, 1994 Page 8
Disk Equivalences
Note 93.1, 16-Jan-1988
Kennedy: RD52
-------------
The DEC RD52 comes in two flavors - the more common one is the
Quantum Q540 drive. The second-source drive, an Atasi AT3046,
was added due to shortages of the Q540 when DEC's demand was
highest. These drives will work on RQDX1, 2, or 3 controllers
at the current revision level. There are no differences between
the DEC and non-DEC flavors which prevent operation. DEC is
alleged to require tighter tolerances on drives made for them.
There is actually a third RD52 drive type. None of these units
were ever shipped to DEC customers, and the vendor ceased pro-
duction of disk drives. The manufacturer and model was the
Evotek ET5540.
If you have one of these drives around, you can use it with
your DEC subsystem. Both drives are no longer manufactured in
any quantity, so I would advise against buying one. If you
must, however, they should be available in the $300-$400 range
from various surplus shops. I have used both models without
problems. Since this is the largest drive a RQDX1 will support,
you might want to investigate this option.
Note 93.2, 16-Jan-1988
Kennedy: RD53
-------------
The DEC RD53 is a Micropolis 1325 or 1335 drive with one mod-
ification. Earlier units were 1325s, current units are 1335s.
The modification is as follows [yes, it will probably void
your warranty]: Turn the drive over so that the logic board
is facing you. Locate the two slotted screws and loosen them.
Lift the drive logic board up gently (it is hinged). Toward
the hinge end, locate the empty location marked R7. Solder a
jumper in. You now have a RD53. You will need to move the
drive select jumper to either 3 (for DU0 or expansion cabinet)
or 4 (for DU1).
One word of caution - if you are re-using your old drive's
skid plate (the slide mount bracket), be sure to tape over
the metal on the front end of the bracket or it will short
out the drive. DEC's RD53 skid plates have foam covering the
metal.
I have purchased three of these drives from Storage Dimen-
sions in California, (408) 395-2688. The cost was under
$850 each and they arrived in two days. They offer a one
year warranty on the drive, too. They call it a 'Model AT70',
but they're probably going to know what you mean if you
ask for 'the DEC drive' by now!
Of course, you need a RQDX2 or 3 to run this drive...
The DECUServe Journal December, 1994 Page 9
Disk Equivalences
Note 93.3, 16-Jan-1988
Kennedy: RX33
-------------
The DEC RX33 is an IBM AT-type drive with some jumper changes.
DEC's drive is some variant of the TEAC FD55G. I have exper-
imented and discovered that the following drives/jumperings
work with the RQDX3:
TEAC FD-55GFV-17-U
Jumpers in:
HG
I (Roman numeral 1)
U1
U2
DC
FG
DS0 (for first RX33)
DS1 (for second RX33)
Jumpers out:
(all others)
Install the terminator, RA1, on the last (or only) drive
TEAC FD-55GFR-540-U
Jumpers in:
I (Roman numeral 1)
U0
U1
DC2
FG
D0 (for first RX33)
D1 (for second RX33)
Jumpers out:
(all others)
Install the terminator, RA1, on the last (or only) drive
Toshiba FDD 6882E1J01 (or ND-08DE-G)
Jumpers in:
LD
DC
DE
D1 (for first RX33)
D2 (for second RX33)
Jumpers out:
The DECUServe Journal December, 1994 Page 10
Disk Equivalences
(all others)
Install the terminator jumper, TM, on the last (or only) drive
I have purchased many of the TEAC 540 drives from JDR Micro-
devices in California, (800) 538-5000. They have the follow-
ing items of interest:
FD-55G High density drive - $129.95
FD-5Y 'Y' cable to power 2 drives from the BAnn box - $2.95
FD55-MHW Mounting hardware for two FD-55G drives - $2.95
FD55-FP Beige faceplate for FD-55G drive (black is std.) - $2.95
FD55-SPEC Specification for FD-55G drive - $5.00
FD55-MAINT Maintenance manual for FD-55G drives - $25.00
Therefore, you can get two RX33's for $265.80 with the Y cable and
mounting hardware - a far cry from DEC's $795.00 each...
You need an RQDX3 to run these drives, of course. You will also
need to save the skid plate from your RX50 drive.
If you order these drives elsewhere, please remember that the
mounting hole threads are *metric*. Finding the right screws
will be a lot harder than finding the drives!
Note 93.6, 1-Apr-1988
Kennedy: RD54
-------------
The DEC RD54 is a Maxtor XT-2190. I have installed one of
these with no modifications in a Micro-11 system (with a
RQDX3 as the controller, of course).
I purchased the drive from Storage Dimensions in Califor-
nia, (408) 395-2688. The cost was under $1700 and it
arrived in two days. They offer a one year warranty on the
drive, too. They call it a 'Model AT133'.
Of course, you need an RQDX3 or the integrated thingie in
a MicroVAX 2000 to run this drive...
Note 93.7, 14-May-1988
Ethington: Wanted: Skid plate part #
------------------------------------
Along these lines, folks.....does anyone know the part numbers for
your friendly neighborhood field circus to sell you the plastic
skid plate and big flat grounding spring/shield or whatever? Lord
knows you can get the same disk for enormously less loot than Storage
Systems wants, but it would be nice to have the official skid plate
hardware to hold it in place instead of rubber bands or string or
whatever.....
The DECUServe Journal December, 1994 Page 11
Disk Equivalences
Note 93.10, 15-May-1988
Kozam: Skid Plate Part Number 74-25652-01
-----------------------------------------
The part number for skid plates (which I read off an RD53) is
74-25652-01 and is described as "112684 Skid" and is priced at only $ 5.00
each (that's a real surprise!).
I don't know if that includes the metal grounding plate or just
the plastic part.
Note 93.12, 18-May-1988
Rice:
------
74-25652-01 buys you the plastic "plate". Part number 74-28201-02 gets you
the metal "grounding strap". It is a bit more at $11.00 (1986 price). I
got BOTH from DECdirect by just calling and requesting these part numbers.
They were very helpful and gave no indication that I wasn't allowed to buy
these parts.
Note 93.13, 29-May-1988
Ethington: Paydirt!
-------------------
Thanks, folks! I had the part number for the plastic skid plate,
and had no difficulty at all ordering a batch - I think the price
was still around $5. I had never come up with the number for the
big spring grounding plate doodad, though - I'll try ordering some
of these critters now.
Note 93.14
Kennedy: Logic of part numbers!
-------------------------------
The spring p/n should be a -02, where the plastic is -01.
Note 93.15, 21-Jun-1988
Ethington: Stacking Hardware?????
---------------------------------
Still further along these lines.....
I understand that new mounting hardware is available to support
stacking two half-height RD3x series Winchesters in a BA23 box.
It is also apparently supported to stack a pair of RX33 half-height
floppies, allowing you to stuff 2 Winchesters and 2 floppies in
the BA23 box. Assuming one has say an official DEC RD31 with skid
The DECUServe Journal December, 1994 Page 12
Disk Equivalences
plate and such, does anyone know what parts and cables need to be
ordered to stack these varmints? I've not seen the stacking setup;
has anyone got one of these critters to see the cabling? For instance,
do you need a "Y" power cable to plug 2 Winchesters into one plug
in the BA23 box? How do the 2 floppies cable up?
Inquiring minds want to know.....
Note 93.16, 21-Jun-1988
Kennedy: Possible, proceed carefully
------------------------------------
My sales rep also 'understands' this - however, the only way we've
figured out to get the part numbers is to order an 11/53 and dissect
it. This is because the '53 is one of the only systems standard with
this 2-up stuff, and the 11/53 documentation is a few photocopied
addendum sheets which you may or may not get.
You *will* need the BA23-UC (@$200.00) upgrade kit to properly put
2 winnies in the BA23. This kit gives you an extra set of ready/write
protect buttons, and deletes the two floppy WP indicators. [Yes, you
can do it without the -UC, but it isn't very pleasant and if you get
the drive selects wrong you will blow the format off *both* drives.]
> It is also apparently supported to stack a pair of RX33 half-height
> floppies, allowing you to stuff 2 Winchesters and 2 floppies in
> the BA23 box.
Yes, that is the maximum configuration. If you go for it, be *sure*
you have the current revision of the power harness or you'll go up in
smoke!. The current harness is individual wires with white plastic con-
nectors (between the power supply and the backplane).
Anyway, to answer your question, I just went out and got the 'Y'
cable and the adapter plates from my local garage-based PC outfit. Most
places which will sell you RX33 look-alikes will gladly throw in the Y
and the plates with the order. I have ribbon-cable presses here, so I
made my own data cables up. Again, your local PC place should be able
to make them up, once you convince them it's not a standard PC part and
that you know what you're doing...
Note 93.17, 21-Jun-1988
Killeen: After Many Hrs of Wondering Why I Couldn't Write to DU1
----------------------------------------------------------------
> [Yes, you
> can do it without the -UC, but it isn't very pleasant and if you get
> the drive selects wrong you will blow the format off *both* drives.]
Yes and No - Without the second set of buttons the second internal drive
comes up R/W on an RQDX1 and RO on a RXDX2 or RQDX3. Trimarchi (no I don't
own stock in this company) sells the drives with the stacking hardware and
The DECUServe Journal December, 1994 Page 13
Disk Equivalences
a special set of cables. The cables de-activate the two front panel
disk buttons and forces the drives ready W/R all the time.
I found this one out the hard way......
Note 93.18, 22-Jun-1988
Mayhew: Drives w/stack hardware, and floating signals
-----------------------------------------------------
There is a variant of both the RD31 and RD32 drives that "comes
with" the stacking hardware, but that doesn't help a whole bunch
if you already have the drive... for those that care, though, you
should be able to find it readily (the words STK HDW or such are
in the part title) on the Estore or in the USPL.
I have reports from other folks who have put multiple Winchesters
in a BA23 without the extra switches that what happens is the
write-protect and drive-enable signal lines are left "floating".
Depending on the phase of the moon, your power supply (AC and DC),
etc., you can then have flaky problems where the drive suddenly
drops off-line, or write-protects itself, without warning. I don't
know which controller was involved, though I think it was the RQDX1.
The drives were RD52s, which are risky ANYWAY to double up in a
BA23 because of power draw, but these folks did it...
Note 93.19, 21-Jun-1988
Kennedy: Many variables...
--------------------------
> Yes and No - Without the second set of buttons the second internal drive
> comes up R/W on an RQDX1 and RO on a RXDX2 or RQDX3.
It depends on a number of things, including the revs of the RQDXn, whether
a RQDXE is also in the system, and the rev of the front panel. Also, you need
to jumper the first fixed disk as DS3 and the second as DS4, even if you have
the new front panel.
Note 93.20, 27-Jun-1988
Kennedy: RX50 drive substitute!
-------------------------------
Finally, a RX50 drive replacement!
I have managed to convince a Toshiba IBM AT-type floppy drive that it
is an RX50. I've tested extensively with both XXDP+ and RSTS/E, but that
doesn't mean that it will work with other systems - you'll have to try
it and let me know.
Here's what you need to do:
1) Get a Toshiba ND-08DE-G drive kit from your local IBM garage oper-
The DECUServe Journal December, 1994 Page 14
Disk Equivalences
ation. (The actual drive is a Toshiba FDD6882E1J01)
2) Set jumpers as follows: D1, TM, LD and DE on, all others off.
3) Cut the trace leading to pin 2 on the data connector.
4) Solder a wire from the drive side of the wire you cut in (3) to
signal ground (the large solid areas on the circuit board)
This has been tested on both the RQDX1 (M8639/M8639-YA) controller
with firmware 172E5/173E5, as well as with the RQDX2 (M8639-YB) with
firmware 188E5/189E5. With an RQDX3, you can use the same drive in
RX33 mode (see an earlier reply in this note).
For the curious: what the above has done is forced the drive into
300 RPM/low density mode, and disabled the drive ready line.
Note 93.29, 15-Sep-1989
Kennedy: Toshiba-as-RX50 doesn't work on RQDX3
----------------------------------------------
By the way, this *won't* work on an RQDX3 - it only works on the RQDX1
and RQDX2. Of course, you can use the Toshiba drive on the RQDX3 by setting
it up as an RX33 lookalike (see earlier note for jumper settings).
Note 93.32, 3-Oct-1989
Hanrahan: Control panel for drives in BA123?
--------------------------------------------
So far our adventures in drive replacement have gone well. Now I want
to ADD an RD53 lookalike to a uV2 (BA123).
I have the drive, and am ordering the skid plate and ground strap.
Cables we can make.
What about the control panel? Is there a DEC part number for this
assembly? (Presumably we need the panel and a cable to plug into
the "distribution board")
Can I use one of those PC-style Molex Y cables for power, or is there
a DEC part I should order?
Note 93.34, 3-Oct-1989
Kozam: No need for front panel
------------------------------
> What about the control panel?
If the drive is going to be on-line and write-enabled at
all times, then you have no need for a panel. I've never had any problem
with this arrangement.
The DECUServe Journal December, 1994 Page 15
Disk Equivalences
> Can I use one of those PC-style Molex Y cables for power, or is there
> a DEC part I should order?
You should have a power cable for each device slot in the BA123,
so a Y shouldn't be necessary.
Note 93.35, 5-Oct-1989
Killeen: BA123 RD Control Panels
--------------------------------
> What about the control panel? Is there a DEC part number for this
> assembly? (Presumably we need the panel and a cable to plug into
> the "distribution board")
The RD control panel for a BA123 box is part # 54-16244-02
If there is a DEC RD drive in the BA123 you should already have the
cable since it is a single cable for all four possible drives. The
cable's part # is 17-00862-01
Trimarchi sells a clone for $250 and it includes the cable if you ask
for it.
> Can I use one of those PC-style Molex Y cables for power, or is there
> a DEC part I should order?
I don't understand this question. All 5 1/4 inch mounting slots have a
power cable available. It is the standard PC type drive power cable.
The DECUServe Journal December, 1994 Page 16
MicroVAX 2000 Comm Ports
MicroVAX 2000 Comm Ports
========================
This thread is of considerably more recent vintage than the
preceding ones. Also taken from the HARDWARE_HELP conference, it
discusses the I/O ports found on the back of a MicroVAX 2000. The
primary mysteries concern the console ports and that funny
SCSI-looking thing. So what can you plug in there, anyway? Read
on....
Participants: Jeff Barton, Harry Flowers, Pierre Hahn, Jamie
Hanrahan, Pete Sivia.
Note 1767.0, 12-Oct-1994
Hahn: MicroVAX 2000 Questions
-----------------------------
We have been given some MicroVAX 2000 without any manuals, I need
to understand a few of the components, and I/Os.
There is a device which plugs into the 9 and 15 male D connectors
and has three (3) MMJ conectors labeled 1 2 3.
I am guessing that these must be for a terminal, a mouse, and ?,
but which is which.
There is a 25 pin D male connector marked with two arrows going
in opposite direction. Is this a 2nd rs232 port?
There is an additional piece on the bottom of the mVAX (about 1.5")
with a SCSI connector and a 34-pin 3-row D connector - what is it?
A power cord connector, a BNC (for Ethernet), and a small square
green button next to the 9 pin male connector marked with a
triangle - what is this, the reboot switch?
I presume I can connect a VT terminal to the proper MMJ connector
and power on to see what I have... correct?
Note 1767.1, 12-Oct-1994
Barton: 2000 answers...
-----------------------
I am recalling this from memory, I had many of these in a prior life.
Your mmj ports appear as normal terminal/printer ports...... tta0,
tta1, tta2, ..I believe.
> There is an additional piece on the bottom of the mVAX (about 1.5")
> with a SCSI connector and a 34-pin 3-row D connector - what is it?
This is for an external SCSI device. TK50, RD53, RD54 remember these?
> A power cord connector, a BNC (for Ethernet), and a small square
The DECUServe Journal December, 1994 Page 17
MicroVAX 2000 Comm Ports
> green button next to the 9 pin male connector marked with a
> triangle - what is this, the reboot switch?
The triangle button is the halt button.
> I presume I can connect a VT terminal to the proper MMJ connector
> and power on to see what I have... correct?
I believe you can connect a VT to tta0 and it will be the console. That
or use the RS232 port you mentioned. It has been a while and my mind
fails me more than it used too.
Note 1767.2, 12-Oct-1994
Flowers:
---------
On a MicroVAX 2000, the asynch ports are as follows:
MMJ1 OPA0:
MMJ2 TTA1:
MMJ3 TTA3:
RS232 TTA2: (with modem control)
Note 1767.3, 12-Oct-1994
Sivia: Not quite SCSI
---------------------
The 1.5" device is the BA40 adapter. It supported external to the CPU
cabinet devices. And it wasn't quite "SCSI" in the sense most folks
use it today. It only worked with the TK50-FA (? I think that's
right) tape drive, and the RX33 (floppy), RD32 (40MB?), RD53 (71MB) &
RD54 (159MB) disks.
We've still got a bunch of these around. If you've got specific
questions, I can probably drag up a real manual. (I've got the
VS2000's manual in hand, it was was nearly identical for all the
console tests and the like as I recall.)
Note 1767.4, 12-Oct-1994
Hanrahan:
----------
> There is a device which plugs into the 9 and 15 male D connectors
> and has three (3) MMJ conectors labeled 1 2 3.
> I am guessing that these must be for a terminal, a mouse, and ?,
> but which is which.
if you were using the unit as a VAXstation, you'd remove this device and
plug the monitor/mouse/keyboard cable into the D15 connector. This
connector provides access to OPA0 and TTA1, used for the keyboard
and mouse (they're serial, near-EIA232 devices), along with the video
The DECUServe Journal December, 1994 Page 18
MicroVAX 2000 Comm Ports
leads. At the other end of the monitor/mouse/keyboard cable are three BNC
plugs, the mini-modular jack to accomodate DEC keyboard cables, and the mini-DIN
socket to accomodate the DEC mouse.
The D9 connector remains the printer port, TTA3, and the D25
connector is TTA2, providing full modem control (even CTS, which iis odd given
that it's a DZ-like device, but it does). The box with the
MMJ sockets on it simply converts all of these to MMJ (and ignores the
video signals).
The DECUServe Journal December, 1994 Page 19
Old Q-bus Hardware
Old Q-bus Hardware
==================
Beyond the original (mainly antiquarian?) question of 18-bit Q-bus
parts from an old 11/03 in a 22-bit Q-bus world, the following
HARDWARE_HELP stream contains a couple of interesting asides.
First, you will notice a few references to "Micronotes" in the
discussion. The documents in question were produced by an outfit
within DEC called the OEM Technical Support Group, and deal with a
variety of technical topics concerning early "Micro-something"
platforms (LSI-11, MicroVAX I and II, and so on). DECUServe has
Micronotes 1 through 44 online (never throw anything away,
remember?) in a read-only Notes conference.
There is also a digression into the limitations of the 11/03's H780
power supply that could prove interesting, or maybe even useful, to
some readers.
Participants: Alan Frisbie, Terry Kennedy, Jeff Killeen, Marc
Kozam, Frank Nagy
Note 144.0, 4-Jun-1988
Killeen: 18-bit vs 22-bit Qbus Devices
--------------------------------------
I just pulled apart a PDP-11/03 system. The question is how can I tell
if a board will work on a 11/53/73/83 system. I have heard that some
PDP-11/03 boards are 18-bit addressing only. For example....
DLV11 (M7940)
RXV11 (M7946)
REV11 (M9400-YA) After I change the 9270-A to a 9270-Q backplane
Will they work on a 22-bit system?
Note 144.1, 4-Jun-1988
Nagy: 18/22-bit only issue for DMA devices
------------------------------------------
As I remember, the 18-bit vs 22-bit QBus issue only affects DMA devices.
Thus the DLV11 will work in the Q22 systems since its a programmed-I/O
device. The RXV11, however, is a DMA device - however, didn't RSX (at
least) support the RXV11 on the Q22 bus by setting up a common region
in the first 256K of memory? The driver would do DMA to/from buffers
in this common region and then do memory-to-memory copies between the
user buffers and the common region. I have no idea about the REV11,
what in the world is it?
The DECUServe Journal December, 1994 Page 20
Old Q-bus Hardware
Note 144.2, 4-Jun-1988
Killeen: The Answer Is
----------------------
> what in the world is it?
The bootstrap/terminator
Note 144.3, 4-Jun-1988
Frisbie: Slight confusion
-------------------------
|| As I remember, the 18-bit vs 22-bit QBus issue only affects DMA devices.
|| Thus the DLV11 will work in the Q22 systems since it's a programmed-I/O
|| device.
Absolutely correct.
|| The RXV11, however, is a DMA device - however, didn't RSX (at
|| least) support the RXV11 on the Q22 bus by setting up a common region
|| in the first 256K of memory? The driver would do DMA to/from buffers
|| in this common region and then do memory-to-memory copies between the
|| user buffers and the common region.
You are confusing the RXV11 (M7946) with the RXV21 (M8029). The
RXV11 is the non-DMA controller for the RX01. The RXV21 is a
DMA controller for the RX02. Thus, the RXV11 will work just
fine in a 22-bit backplane. You are right about RSX (and
RT-11) with the RXV21.
|| I have no idea about the REV11, what in the world is it?
The REV11 is a terminator, bootstrap and DMA refresh board.
It cannot be used in a 22-bit system because the termination is
18-bit only. For those of you with short memories, the REV11
DMA refresh capability was for refreshing memory boards that
did not have an on-board refresh capability. The LSI-11 had a
jumper option that allowed the microcode to perform the refresh
function, but at a decrease in CPU speed. The REV11 took over
this function by doing DMA refresh cycles, allowing the CPU to
execute instructions at full (such as it was) speed. The
whole issue became moot with the demise of memory boards that
needed external refresh. Only a few early models were ever built.
The whole subject of Q22 compatible options is addressed quite
will in Micronote #5 (23-Apr-84) by Charlie Giorgetti. It
lists all processors, backplanes, memories, controllers and
bus-expansion cables that were available at that time. For
options that have "restricted compatibility", it tells *why*
they are restricted (as I did above for the REV11).
In the next note, I will extract the list of "restricted
compatibility" options.
The DECUServe Journal December, 1994 Page 21
Old Q-bus Hardware
Note 144.4, 4-Jun-1988
Frisbie: Q22 bus restricted options
-----------------------------------
This note is an extract from DEC's Micronote #5, "Q22 Compatible
Options" by Charlie Giorgetti, dated 23-Apr-84. Please refer
to the original Micronote for complete information, as this
note does not list the "fully compatible" options.
----------------------------------------------------------------
A Q22 compatible option is defined as a Q-bus option that will
work without restriction in an extended Q-bus system, that is a
22-bit Q-bus system. The list below is of options that are
not compatible in Q22 systems and the reason for the restriction.
The requirements for a device to be Q22 compatible are the following:
1) Processors, memories, and DMA devices must all be capable of
22-bit addressing.
2) Devices must use backplane pins BC1, BD1, BE1, BF1 and DC1,
DD1, DE1, DF1 for BDAL18-21 only.
Processors, memories or DMA devices which are not capable of
22-bit addressing may generate or decode erroneous addresses if
they are used in systems which implement 22-bit addressing.
Memory and DMA DMA devices which implement only 16 or 18-bit
addressing may be used in a 22-bit backplane, but the size of
the system memory must be restricted to the address range of
those devices.
NOTE
16 or 18-bit DMA devices can potentially work in Q22
systems by buffering I/O in the 16- or 18-bit address
space.
Any device which uses the backplane pins list above for
purposes other than BDAL18-21 is electrically incompatible with
the Q22 bus and may not be used without modification.
FULLY COMPATIBLE OPTIONS:
(See original Micronote)
RESTRICTED COMPATIBILITY OPTIONS:
PROCESSORS:
KDF11-A M8186 LSI-11/23 CPU
Prior to etch rev. C, 18-bit addressing only,
and use of BC1, BD1, BE1, BF1 for purposes
other than BDAL18-21.
KD11-HA M7270 LSI-11/2 CPU
16-bit addressing only, and use of BC1, BD1,
BE1, BF1 for purposes other than BDAL18-21.
The DECUServe Journal December, 1994 Page 22
Old Q-bus Hardware
KD11-F M7264 LSI-11 CPU
16-bit addressing only, and use of BC1, BD1,
BE1, BF1 for purposes other than BDAL18-21.
KXT11-A M8063 SBC-11/21 CPU
16-bit addressing only.
BACKPLANES/BOXES:
DDV11-B 6 x 9 backplane
18-bit addressing only
BA11-M H9270 4 x 4 backplane
18-bit addressing only
BA11-N H9273-A 4 x 9 backplane
18-bit addressing only
BA11-VA H9281-A,B,C 2 x n dual-height backplane (n = 4, 8, 12)
BA11-VA used the H9281-A
18-bit addressing only
VT103 4 x 4 backplane
18-bit addressing only
NOTE
It is usually possible for the adventurous hacker
to add the extra backplane wires to convert the
above backplanes to 22-bit addressing. -- Alan
MEMORIES:
MMV11-A G653 8 KB core memory
16-bit addressing, Q-bus required on C/D
backplane connectors
MRV11-AA M7942 ROM module
16-bit addressing only
MRV11-BA M8021 UV PROM-RAM
16-bit addressing only
MRV11-C M8048 PROM/ROM module
18-bit addressing only
MSV11-B M7944 8 KB bus refreshed RAM
16-bit addressing only
MVS11-C M7955 32 KB RAM
18-bit addressing only
MSV11-D,E M8044/M8045 8, 16, 32, 64 KB RAM
The DECUServe Journal December, 1994 Page 23
Old Q-bus Hardware
18-bit addressing only
MXV11-A M8047 Multifunction module
18-bit addressing only on memory, the memory
can be disabled to use just the serial lines
OPTIONS:
AAV11 A6001 D/A converter
Use of BC1 for purposes other than BDAL18
ADV11 A012 A/D converter
Use of BC1 for purposes other than BDAL18
BDV11 M8012 Bootstrap/Terminator
CS Rev. E or earlier 18 bits only
DLV11-J M8043 Serial line interface
CS Rev. E or earlier incompatible with KDF11-A
and KDF11-B [Don't ask me why, I'm just
reporting it -- Alan]
DRV11-B M7950 General purpose DMA interface (quad)
18-bit DMA addressing only
KPV11-B,C M8016-YB,YC Power-fail/line-time clock/terminator
Termination for 18 bits only
KUV11 M8018 Writable Control Store
For use with KD11-F processor only
KWV11-A M7972 Programmable real-time clock
Use of BC1 for purposes other than BDAL18
REV11 M9400 Terminator, DMA refresh, bootstrap
Bootstrap for use with KD11-F and KD11-HA
processors only. Termination for 18-bits only.
DMA refresh may be used in any system.
RKV11-D M7269 RK05 controller interface
16-bit DMA addressing only. [Can be "hot-wired"
for 18-bit DMA. See me for details. -- Alan]
RLV11 M8013 RL01/2 controller interface
M8014 18-bit DMA only, use of BC1 and BD1 for
purposes other than BDAL18 and BDAL19.
Requires two adjacent slots with C/D interconnect.
RXV21 M8029 RX02 floppy disk interface
18-bit DMA addressing only
TEV11 M9400-YB 120-Ohm bus terminator
Termination for 18 bits only
VSV11 M7064 Graphics display
The DECUServe Journal December, 1994 Page 24
Old Q-bus Hardware
18-bit DMA addressing only
BUS CABLE-CARDS:
M9400-YD Cable connector
18-bit bus only
M9400-YE Cable connector with 240-Ohm termination
18-bit bus only
M9401 Cable connector
18-bit bus only
Note 144.5, 4-Jun-1988
Killeen: Questions - More Questions
-----------------------------------
> The REV11 is a terminator, bootstrap and DMA refresh board.
> It cannot be used in a 22-bit system because the termination is
> 18-bit only. For those of you with short memories, the REV11
Thanks Alan - you saved some hardware from the trash bin.
However this now raises a question - if I replace the H9270-A backplanes
in the BA11-M boxes I have with H9270-Q backplanes how do I terminate
them? (Since the REV11 or TEV11 are 18-bit only). When DEC re-did the
H9270 did they put termination in it?
Note 144.6, 5-Jun-1988
Frisbie: The answer is in the Micronotes
----------------------------------------
>> However this now raises a question - if I replace the H9270-A backplanes
>> in the BA11-M boxes I have with H9270-Q backplanes how do I terminate
>> them? (Since the REV11 or TEV11 are 18-bit only). When DEC re-did the
>> H9270 did they put termination in it?
To the best of my knowledge, none of the H9270-x backplanes
have termination. However, the only one I have handy is
the original H9270 (no suffix).
In such a short backplane, the termination on the CPU card
should be sufficient.
This subject is addressed quite thoroughly in DEC's Micronote
#35 by Jack Toto (19-Jul-85), "Backplane Expansion/Termination".
To answer your question directly, I will quote from the
section on Single Backplane Termination Rules:
1. Then using a processor with 240 ohm termination, the
bus can accommodate up to 20 ac loads (total) before additional
The DECUServe Journal December, 1994 Page 25
Old Q-bus Hardware
termination is required. If more than 20 ac loads are
included, the far end of the bus must be terminated with
120 ohms, although termination of 240 ohms is optimum.
Following the addition of at least the minimum termination,
up to 35 ac loads may be present in a single backplane.
2. Then using a processor with 120 ohm termination, the
bus can accommodate up to 35 ac loads (total) before additional
termination is required. If more than 35 ac loads are
included, the far end of the bus must be terminated with
120 ohms. When this is done, up to 45 ac loads may
be present.
3. In all cases, the bus can accommodate up to 20 (total)
dc loads.
The processor terminations are:
KDF11-A 240 ohms 11/23 dual width
KDF11-B 120 ohms 11/23+ quad width
KDJ11-A 240 ohms 11/73 dual width
KDJ11-B 120 ohms 11/73 quad width
MicroVAX I 240 ohms
MicroVAX II 240 ohms
For many years I ran a system in a H9270 backplane with
no additional termination. There were no problems.
By the way, I hope that you realize that an H9270-A backplane
can be upgraded to an H9270-Q by adding the extra four wires
to all the slots.
Note 144.7, 5-Jun-1988
Killeen: Ain't DECUServe Great!
-------------------------------
> For many years I ran a system in a H9270 backplane with
> no additional termination. There were no problems.
Thanks again Alan! I shall now build my portable PDP-11/73 (BA11-M & RD54-DA).
> By the way, I hope that you realize that an H9270-A backplane
> can be upgraded to an H9270-Q by adding the extra four wires
> to all the slots.
For the 32 wire wraps (4X4X2) I will spring for the H9270-Q and keep the
old one as "spare". (about $175)
THANKS AGAIN ALAN!
The DECUServe Journal December, 1994 Page 26
Old Q-bus Hardware
Note 144.8, 5-Jun-1988
Kennedy: Wrapping is faster!
----------------------------
> For the 32 wire wraps (4X4X2) I will spring for the H9270-Q and keep the
> old one as "spare". (about $175)
Total effort is about 10-15 minutes to do the wrapping, compared with at
least an equal time wrestling the backplane in & out. Not to mention that
you can (usually) find wire-wrap wire a lot faster than a backplane!
P.S. - If you do wrap it, do it as two vertical rows (A/B side, C/D side),
and then connect them, rather than ABCD-ABCD-ABCD... This will reduce the
noise. I've done this and it runs fine at 18Mhz - although the J11-RQDX2
combo makes the 11/03 power supply scream!
Note 144.9, 5-Jun-1988
Killeen: Who Invests In Wire Wrap Tools Any More?
-------------------------------------------------
> Total effort is about 10-15 minutes to do the wrapping, compared with at
> least an equal time wrestling the backplane in & out. Not to mention that
> you can (usually) find wire-wrap wire a lot faster than a backplane!
Not if you don't have good wire wrap tools!
> I've done this and it runs fine at 18Mhz - although the J11-RQDX2
> combo makes the 11/03 power supply scream!
Are you telling me the 11/03 power supply is marginal? I didn't even
think about this since it weighs more than a BA23 power supply.
Note 144.10, 5-Jun-1988
Frisbie: Less power to you :-)
--------------------------------
>> Are you telling me the 11/03 power supply is marginal? I didn't even
>> think about this since it weighs more than a BA23 power supply.
If you are referring to the BA11-M box (H780A power supply):
Output Power (Combinations not to exceed 110 watts):
+5 1.5 to 18.0 amps
+12 0.25 to 3.5 amps
Note 144.11, 5-Jun-1988
Kozam: H780 (BA11-M) power rating and "screaming"
-------------------------------------------------
>Are you telling me the 11/03 power supply is marginal? I didn't even
>think about this since it weighs more than a BA23 power supply.
The DECUServe Journal December, 1994 Page 27
Old Q-bus Hardware
A quick comparison between the H7864 (BA23) power supply and H780
(BA11-M) is as follows. The H786 (BA11-N) is very similar to the H780.
H780 H7864
+5 Volts 0-18A 36A (typical)
+12 Volts 0-3.5A 7A (typical)
As you can see, pound for pound, the H7864 is a much better deal.
At the same time, I've never had difficulties with any of the power supplies
being unable to deliver enough amps for normal use. (If you had intended
to power an RD5X drive with the H780, then you'd be pushing things.)
I think that the "screaming" which was referred to has to do with
the design of the power supplies. "Switching" power supplies, such as
these often emit a high-pitched tone, which is very annoying. I've only
noticed this with the H780, especially when a heavy current is drawn.
Note 144.12, 6-Jun-1988
Kennedy: Power supply *can* be a problem
----------------------------------------
> Who has wire-wrap tools any more?
You mean you remove the Unibus NPR jumpers with your teeth? 8-)
> As you can see, pound for pound, the H7864 is a much better deal.
Yep - the 780 can (according to specs) run *either* a J-11 *or* an
RQDXn, but not both. Add memory and terminal I/O and it gets worse...
However, it has worked for me - but you may not want to play 'you bet
your boards'.
> I think that the "screaming" which was referred to has to do with
> the design of the power supplies. "Switching" power supplies, such as
> these often emit a high-pitched tone, which is very annoying. I've only
> noticed this with the H780, especially when a heavy current is drawn.
The 780 has a design such that as the load increases, the switching
frequency drops, lowering efficiency, causing increased current draw.
Once you get past the knee, the supply *will* die. The BA23's supply
has the sense to either pop the circuit breaker or burn the cable har-
ness (sorry, I couldn't resist...).
The DECUServe Journal December, 1994 Page 28
Printronix 600 Setup
Printronix 600 Setup
====================
Along with old systems go old printers, at least in the dealer
listings. Printronix model 300 and 600 printers (also sold by DEC
as the LXY family) are decent, and show up fairly regularly in the
"for sale" columns. But supposing you acquire one, how do you hook
it up? It may not always be clear, especially if documentation is
sketchy, missing, or just plain wrong. The answer you're looking
for might be in (where else?) HARDWARE_HELP.
Participants: Alan Frisbie, Jerry Hudgins, Gary Rice
Note 574.0, 31-Jan-1990
Hudgins: LXY-22/P600 Serial Setup?
----------------------------------
I've recently acquired a used Printronix P600 printer with a serial
interface, and I'm having some trouble configuring it for my system.
I'm running it as a shared printer off a DS200. The used equipment
vendor shipped me DEC's LXY-22 manuals with it, but the information
on the serial board setup is cryptic (or wrong, as in the case of
defining the XON and XOFF bit patterns). It seems to work fine
EXCEPT for the serial board configuration.
Problem is mostly flow control: how do you enable automatic XON/XOFF
flow control? I've tried various jumper settings with no luck.
It DOES send an XOFF when it's placed off-line, so it groks the
XON/XOFF codes, but it happily allows its buffer to be overwritten
without sending XOFF.
Can someone with a serial P300/P600 have a look at their serial
interface card and tell me how it's set up? Many of the jumpers
are not addressed by the LXY-22 manual at all. When they are
listed, the orientation of the jumpers is not all clear, since the
board pins are mostly unnumbered...
By the way, the DEC manual says the printer is shipped at 9600 baud
(mine arrived set up for 4800), but the board has a jumper position
for 19200. Anyone successfully running a P600 or LXY-22 at this
speed?
Note 574.1, 31-Jan-1990
Frisbie: I have the same problem
--------------------------------
One of my customers just got a used P300 and is having the
same problem. Without a datascope I can't tell if anything
is being sent, but the terminal driver certainly never sees
an XOFF. This is on an RSX-11M system with a DLV11-J
interface, so we can't use DTR or similar signals.
The vendor is no help, and has yet to supply us with the
The DECUServe Journal December, 1994 Page 29
Printronix 600 Setup
manuals, so you are at least a step ahead of us.
Please post any information you find.
Note 574.2, 1-Feb-1990
Rice:
------
If your printer can't keep up (ie needing to send XOFF) at 4800, why would
you need to set it to 19200????
Note 574.4, 1-Feb-1990
Hudgins: Why not 19200?
-----------------------
> If your printer can't keep up (ie needing to send XOFF) at 4800, why would
> you need to set it to 19200????
The P600 will MOSTLY keep ahead of 4800, and with some data, 9600.
You understand (no doubt) that this is a very data-dependent question.
It depends on line lengths; the ratio of total characters to lines.
Running at 19200 should ensure that the printer is never waiting
for data, which slows down its throughput considerably.
By the way, this is working now at 19200 (see next reply).
Note 574.5, 1-Feb-1990
Hudgins: This is the setup
--------------------------
My service guys here got this thing running by comparing the serial
board to one in a working P300. They couldn't figure out the docs,
either. Mine is configured for 19200, 7 bit, no parity. It seems
to have a 2K buffer; I don't know if this was optional.
If you pull the serial board (second from left in card cage, marked
'RS232' on back) and orient it with the components up and the fingers
on the left, it should look like this:
The DECUServe Journal December, 1994 Page 30
Printronix 600 Setup
---------------------------------------------------------------------
| |: |
| |
| |:|:||:| |
| .- -. :::::|: |
| |:||||:| |
| ... ... |
| ... ... |
| :: ... ... |
| ... |
| |
| |
| .- |: |
| _. |
| ::::|::|::: |
| - |
| |
| | -. |
| :::::::::| :::: Printronix |
---------------------------------------------------------------------
If this isn't decipherable, give me a call.
The DECUServe Journal December, 1994 Page 31
11/70 Tips
11/70 Tips
==========
For something a little bit larger, there is a stream in
HARDWARE_HELP, started long ago by Terry and updated occasionally
thereafter, containing technical advice for that old standard, the
PDP-11/70. The content is geared to those intrepid folks who turn
to self-maintenance for their 11/70. While a '70 may be a bit
bigger than most of the systems discussed in this journal, there are
still a lot of them around, and some of the information is more
broadly applicable.
Participants: Terry Kennedy, Pat Scopelliti.
Note 238.1, 29-Nov-1988
Kennedy: Current module revision levels
---------------------------------------
The following list documents the latest revision (CS level) for the
boards which comprise the 11/70 (KB11-B and KB11-C) CPU. The list is
presented in slot order, right to left, not in module number order.
M9312 - B - Boot/terminator
M787 - B - Clock
M8130 - D - DAP
M8131 - * - GRA
M8132 - C - IRC
M8123 - B - RAC (used only W/ M8138-YC, else M8133)
M8134 - E - PDR
M8135 - * - TMC
M8136 - F - UBC
M8139 - D - TIG
M8137 - A - SAP
M8138 - F - SSR (This is the -YA variation)
M8140 - B - SCC
M8142 - F - CCB
M8143 - B - ADM
M8144 - * - DTM
M8145 - C - CDP
M8141 - A - MAP
Massbus controller (RH70)
M8150 - F - MDP
M8151 - A - CST
M8152 - A - AWR
M8153 - B - BCT
M5904 - F - MBS
Memory subsystem (MK11-C*)
M8158 - C
M8160 - F
M8161 - A
The DECUServe Journal December, 1994 Page 32
11/70 Tips
M8164 - A
M8728 - C
This information is especially important as the DEC FCO documentation
(DEC-O-LOG, MDS) is *incorrect* and *incomplete* for the 11/70.
Note 238.2, 29-Nov-1988
Kennedy: Fans
-------------
One of the common problems in the 11/70 CPU is failure of one or more
of the cooling fans. In the H960-cabinet 11/70 there are 32 4" fans.
The DEC fan is a Nidec-Torin TA450, model A30122-10. These fans are
available from many of the commercial component distributors for far less
than the DEC price. An important note is to *not* accept any substitutes
for this model. This is the top airflow model in its size from *any* fan
manufacturer. Also, do not accept sleeve bearings in the fan. The model
number above specifies ball bearings. After you've changed all 32 fans,
you'll *never* want to do it again, so spend a bit more for the bearings
that last.
Note 238.3, 29-Nov-1988
Kennedy: Power supply lamp upgrades
-----------------------------------
The modular power supplies in the 11/70 have an indicator lamp which
is used to show that the power supply is functional. In older supplies,
this is a normal lamp which has long since burned out. Rather than
removing the power supply or probing with a meter to see if it's the bulb
or if the supply is *really* dead, you can replace the lamp with a LED
and a resistor. This should then last nearly forever.
Method:
1) remove power supply from frame (clip connector on top, 2 screws on top,
large knurled screw on bottom)
2) unsolder lamp adjacent to voltage adjust trimpot
3) cut etch running from upper lamp pad to power bus by capacitor mounting
screw
4) solder in a 470 ohm, 1/4 watt resistor across etch just cut. (This is
for 5 volt supplies. For 15 volt supplies, use 2.2K ohm instead).
5) solder in a red LED, observing polarity (look at capacitor for + lead).
(This is for 5 volt supplies. For 15 volt supplies, use a green LED).
6) replace power supply in frame and test. LED should light when powered
on, and may slowly fade when powered off.
Note: If the above sounds like gibberish, it's not for you...
The DECUServe Journal December, 1994 Page 33
11/70 Tips
Note 238.4, 29-Nov-1988
Kennedy: Troubleshooting RH70 / Massbus problems
------------------------------------------------
Many problems which manifest as Massbus problems may in fact be caused by
out-of-revision CPU modules. The first thing to check when Massbus problems
are suspected are the CPU module revisions, especially the cache modules.
If all the CPU modules are up to revision, look at the following items:
1) Try swapping the M5904 modules with each other.
2) Look at power supply voltages (+5). All CPU +5 supplies should be within
_.5 of each other, and set to between 5.05 and 5.20 volts.
3) Inspect Massbus cable connectors. Swap cables if possible.
4) If problem is only on one device of a single Massbus, try replacing the
Massbus transceiver modules.
5) Swap the RH70 with a different one (M8150-M1853 modules). Note that the
M8153 module has jumper options to set the interrupt and vector.
Note 238.5, 20-Dec-1988
Kennedy: And now for a question
-------------------------------
Has anyone achieved greater than 3.5 Mb on an 11/70 with 256 Kb memory
modules (MK11)? The DEC documentation states that 3.5 Mb is the limit,
but the processor technical manual gives switch settings all the way up
4 Mb (obviously bogus, where would you put the I/O page?). My operating
system SPD states that the max memory limit on this machine is 3840 Kb.
It appears the problem may be the inability to 'ignore' memory in the
MK11 box, and if I put in some 64 Kb arrays it might work...
P.S. - When I install the 2 additional 256 Kb modules ans set the
switches, the system chokes trying to boot...
Note 238.6, 7-Feb-1989
Kennedy: Power-up boot and auto-restart for 11/70
-------------------------------------------------
This reply will discuss getting power-up boot and powerfail restart to
work on 11/70 systems with MK11-C memory boxes.
In order to get power-up boot and auto-restart on your 11/70, you'll
need:
o M9312 boot/terminator module, CS B or newer
o 7010329 Backplane ECO 8 (Wrap level J or newer)
o M8130 CS C or newer
o M8136 CS C1 or F, or newer
o M8138 CS C1 or E, or newer
o Pushbutton (normally open) with Faston tabs on one end of cable
o Cable with 2 conductors, Faston tabs on each end
The DECUServe Journal December, 1994 Page 34
11/70 Tips
Ensure jumpers W1, W3 and W8 are in on the M8130 (W8 is unlabelled, and is
to the left of W7).
Connect the pushbutton's Faston tabs to TP1 and TP2 on the M9312. Order does
not matter. Use the other cable to connect M9312 TP3 to the MK11-C backplane's
J101 tab, and M9312 TP4 to MK11-C J100.
Ensure M9312 S1 position 2 is OFF.
Your 70 will now trap through 24/26 when power returns if the MK11 battery
backup is still valid. Otherwise, it will trap through 773024/773026 and ex-
ecute whatever boot options you have set up in the M9312 S1.
Note 238.7, 7-Feb-1989
Kennedy: ASCII console for 11/70
--------------------------------
You can get an inexpensive electronic console for your 11/70 by ordering
the 23-248F1-00 ROM from DECdirect at $27.00. Replace the 616 boot ROM with
the 248 (on the M9312) and set switch S1 positions 1 & 7 on, all others off.
You now have an ASCII console.
Notes:
1) The console ROM only understands upper case
2) The ROM diagnostics don't test the 11/70 cache modules
3) You should implement the pushbutton boot described in the previous
reply. You'll need it to get back to the console ROM if you hang
the CPU.
Note 238.8, 21-Sep-1990
Kennedy: More 11/70 tips
------------------------
Recently I experienced problems trying to bring up a new release of my
operating system on my 11/70. I found a number of things about the 11/70
which are not widely known while tracking down my problem. One of the
most frustrating things was that the problem would not appear under any
of the 11/70 diagnostics or system exercisers on the latest XXDP tape.
1) On the M8145 module, jumpers W1, W2, and W3 should all be in. This is
supposedly the way the module is shipped when built, but of 8 modules
I examined, none were set up this way.
2) Power supply voltages should be adjusted to within .1V of 5.0 volts.
After adjusting, examine the output of each regulator for ripple or
spikes. If spikes over .1V of centerline are found, replace the power
supply. Perform all measurements at the backplane power connectors.
3) The backplane should be logic rev "S" or higher. This revision repla-
ces long runs with twisted pair wire to reduce noise. This is very im-
portant. The backplane part number is 70-10329-00 and is DECmailerable
The DECUServe Journal December, 1994 Page 35
11/70 Tips
if at rev "L" or higher. Contact DECmailer at 800-225-5385 for more
details, or contact your Field Service provider. If the backplane is
replaced and you have an M787 module in slot 1D, you may have to remove
a wire-wrap wire (D1R2 <-> D1V2).
4) If your 11/70 is a KB11-B or was field-upgraded to a KB11-C, you may
have an M787 KW11 clock in slot 1. The M787 module must be CS "D" or
higher. Older modules may generate spurious requests at priority level
6, which can hang the machine. Quick check - DEC 8881's instead of DEC
381 IC's on module.
5) The following are the highest known revisions for a KB11-C 11/70 with
FP11-C floating point. Older rev modules may work, but it doesn't hurt
to upgrade. Note that the revisions have changed without a correspond-
ing FCO/ECO in some cases. However, some FCO/ECO's were issued but do
not appear in the current microfiche library.
Module Rev
M787 - D
M9312 - C [May be M9301-YC]
M8126 - C [FP11-C - option]
M8127 - B "
M8128 - B "
M8129 - B "
M8130 - D
M8131 - D
M8132 - D
M8123 - B
M8134 - E
M8135 - *
M8136 - F
M8139 - D
M8137 - A
M8138-YA - F
M8140 - B
M8142 - F
M8143 - B
M8144 - *
M8145 - C
M8150 - F [RH70 - option]
M5904 - F
M8151 - A
M8152 - A
M8153 - B
M7856 - K [DL11 w/ clock - used when M787 not present]
M7800 - W [DL11 w/o clock - used with M787]
6) While you have the whole CPU apart, inspect the 3 BC06R cables which con-
nect the front panel to the CPU and the 4 BC06R cables which connect the
memory box to the CPU. If any ends appear frayed or loose, replace the
cable set. Front panel cable is part BC06R-02, memory cable is part BC06R-
12. If the cable ends have electrical or friction tape wrapped around them,
remove the tape and inspect for hidden nicks, cuts, etc. If the ends still
have the factory labels (black print on white plastic tape) do not remove
The DECUServe Journal December, 1994 Page 36
11/70 Tips
unless there appears to be damage underneath. When reinstalling, be sure
to connect cables properly and seat them firmly.
Note 238.9, 28-Jul-1991
Kennedy: Unusual FPA problem
----------------------------
There are some _very_ odd relationships in the 11/70. Recently I replaced
an M8129 module (part of the FP11-C floating point processor) in order to
fix some errors I was getting in floating point math. I installed the new
module and ran 3 passes of every diagnostic.
The machine began reporting "??Swap read error" messages later in the
day, but was _not_ logging any disk errors. Again, all CPU diagnostics ran
fine, as did all disk diags and the memory tests.
I built a new disk pack and had the same problem. I then decided that al-
though "it couldn't possibly be the M8129", I'd replace it with another one
anyway. That was indeed the problem. I haven't looked in the prints yet to
find how a floating point board can cause disk read errors, but I will 8-).
Note 238.10, 28-Jul-1991
Scopelliti: RSTS/E best PDP-11 diagnostic!
------------------------------------------
Seems like they used to use some of the floating point instructions in
RSTS/E (assuming you had an FP11 istalled..) to gain some code
efficiencies, but I thought they had removed them all. Perhaps not...
The DECUServe Journal December, 1994 Page 37
RK07 Tips
RK07 Tips
=========
Also from HARDWARE_HELP, and also dealing with the larger end of the
spectrum of "mature" hardware, is the following collection of
service tips for RK06/RK07 disk drives.
Participants: Terry Kennedy, Tom Provost
Note 418.1, 23-Jul-1989
Kennedy: RK07 - identifying proper read/write board (54-11060-YA)
-----------------------------------------------------------------
One item which can cause bizarre problems in an RK07 drive is the incorrect
read/write board. This is the board that surrounds the head opening. For the
RK06, the correct part is 54-11060 (no suffix). For the RK07, the correct part
is 54-11060-YA. Simple enough, right?
Unfortunately, the -YA suffix is frequently not marked on the etch, but is
instead rubber-stamped in a random location on the board. A sure-fire way to
determine the board type is to examine closely the components directly above
the U-shaped cutout for the heads. You should see the following:
o-====-o <- resistor (no suffix) or jumper (YA suffix)
o--XX--o <- ferrite bead
o-====-o <- resistor
o-====-o <- resistor
o--XX--o <- ferrite bead
o <- test point
An RK07 with a non-YA board will appear to work, but will be very sensitive
to media defects. A pack which works flawlessly in one drive might show 20 or
30 errors in the drive with the wrong board.
Also, if the spindle is worn, these errors may appear as random recoverable
errors (but always at the same places on the pack). Replacing the spindle with
one with less non-repetitive runout will convert these errors into *hard* err-
ors, as you now have less 'slop' on track center. Thus, you will stay over the
low-signal area.
Note 418.2, 23-Jul-1989
Kennedy: RK06/RK07 - emergency head retract failure
---------------------------------------------------
The RK06 and RK07 use batteries to perform the emergency head retract
function. If the battery pack has failed or weakened, the heads will not
retract fully in a power-fail situation.
The DECUServe Journal December, 1994 Page 38
RK07 Tips
Older RK06/7 drives have the batteries tied down underneath the card
cage. Newer drives have the battery in a clip holder under the center col-
umn of the drive.
See note 267.2 in this conference for a source for these batteries.
[Reproduced below under "Hard-to-find Parts". -- ed.]
Note 418.3, 4-Aug-1989
Provost: More RK07 Woes
-----------------------
Card cage must be lifted to perform drive alignment.
Card cage is hard to push squarely back into place.
When card cage is left ajar, backplane pins scratch on back
panel under vibration. This results in blinking fault lights during
heavy system usage. We tape our back panel just in case.
On older RK07's the boards are held in place by flexible
metal straps. These can short out the 5V on the boards. We have
also taped them. Newer RK07's have used a plastic mounting.
The DECUServe Journal December, 1994 Page 39
Reviving a PDP-11/24
Reviving a PDP-11/24
====================
Shifting our attention to the PDP-11 conference for a while, we find
the following longish gem, detailing the experiences of one
participant in resuscitating a PDP-11/24 that he had picked up at
auction. (Sounds to us like he makes a habit of this sort of
thing.)
Unlike most of the articles in this issue, this series of notes
deals with more than just hardware; software and system licensing
issues also come up. And while a Unibus-based PDP-11 may be a bit
out of range for many, in this case the price was clearly right.
Participants: Duncan Brown, Barton Bruce, Jack Crowell, Arnold De
Larisch, Calvin Diddle, Jack Harvey, Terry Kennedy, Marc Kozam, Tom
Provost, Billy Youdelman.
Note 363.0, 13-Sep-1991
Brown: 11/24 Backplane configuration help needed
------------------------------------------------
Here I go again, trying to bring auction-bought PDP's back to life by
using all my Unibus VAX knowledge...
I got an 11/24 and an RL01 and an RL02. I know I got ripped off- I
paid $3 for the whole lot. Shoot. Anyway, having FINALLY scrounged
together all the missing cables and bulkhead adapters and the like, I
want to at least make the sucker boot (the RL01 disk says "RT-11 System
Disk" on it). But it doesn't work yet.
My primary suspect is the fact that two slots in the backplane have
NOTHING in them. Am I correct in guessing that this is a Unibus-based
PDP, i.e. every slot in the machine is a Unibus slot (as opposed to the
older PDP's which had vast swaths of special-purpose slots for the
processor, memory, etc.)? If so, then there should certainly be
something in those slots!
Counting from the right, with the processor board in slot 1, it's slots
5 and 7 that are empty. Slots 6 and 8 both have NPG/Grant Continuity
cards in them. I'm betting I need two more of those, or I could wire
the NPG on the backplane pins and just use a normal Grant Continuity
card. Right? No doubt there were some fun cards in those slots that
were ripped off by the department that was scrapping it before they put
it for for auction. (These are Va state auctions, and most of the
equipment comes from UVa somewhere.)
Also, just out of curiosity, when hooking the ribbon cable from the
disk controller in the PDP to the bulkhead adapter, does pin 1 of the
cable (striped side) connect to pin 1 of the adapter, as marked on the
adapter. Since the thing's not keyed, I figure that HAS to be the
case, but you never know with DEC...
The DECUServe Journal December, 1994 Page 40
Reviving a PDP-11/24
Note 363.1
Harvey: My brain is creaking
----------------------------
The 11/24 is essentially an 11/23 cpu with a Unibus. You are probably
right in think all slots are equal. The empty slots certainly do need
to be filled with bus grant and npg connections.
I've probably plugged that drive cable into the controller a few
hundred times, but I sure can't recall which way it goes. I do recall
that nothing smokes if you plug it in backwards - from experience.
Note 363.2
Brown: Unibus it is. It lives!
-------------------------------
> -< My brain is creaking >-
Gee, is this computer THAT old?! Compared to the other ones I've
picked up at auction (11/20, 11/40, 11/45, 11/70, etc.) it's downright
MODERN. I mean, ONE board for the cpu! Memory that is stored in CHIPS
instead of teensy ferrite toroids! No paddle switches!
> The 11/24 is essentially an 11/23 cpu with a Unibus. You are probably
> right in think all slots are equal. The empty slots certainly do need
> to be filled with bus grant and npg connections.
Well, I popped the bottom off of it and studied the backplane wires
carefully, and it sure LOOKED like all the empty ones were plain old
Unibus slots. I also noticed that two of the slots with no real boards
in them already HAD the NPG jumper on them!
So, I moved the NPG/GC boards to the two empty slots with no NPG
jumper, put normal Grant Continuity boards in the slots that were
already NPG-jumpered, and fired that mother up.
Yeeha! The RUN light now stays on. If I hook up the RL01 with the
RT-11 disk in it (and striped side of cable to pin 1) and hit the boot
switch, the ready light on the drive blinks for a while and then stops,
and the RUN light still stays on. I'd say this beast lives!
Now I need to dredge up a ribbon for the $1 DECWriter II I got and stop
running blind. I'll have to see if I have the rudimentary RT-11
instruction set somewhere. If I recall correctly, there's no "logging
in" to RT-11 right, you're just talking to the machine a la a PC?
Thanks again!
The DECUServe Journal December, 1994 Page 41
Reviving a PDP-11/24
Note 363.3, 14-Sep-1991
Harvey: Right. Where do you think MS-DOS came from?
----------------------------------------------------
>you're just talking to the machine a la a PC?
Note 363.4, 15-Sep-1991
Brown: Same place as Intel processors. Where snowballs melt
------------------------------------------------------------
> -< Right. Where do you think MS-DOS came from? >-
Note 363.5, 15-Sep-1991
Brown: More config questions...
-------------------------------
Okay, let's get this topic back on track...
I've dredged up a near-death ribbon for the DECWriter II. (Appropriate
to the task, no?) I hook it up to the 11/24, boot and/or hit carriage
returns, and all I get printed out is Sanskrit.
Gee, I imagine the previous owners of this 11/24 set its console baud
rate to something other than the 110/150/300 selections available on
the DECWriter. Sigh.
Anybody have off the top of their heads the dipswitch settings on the
11/24 to set it to 300 baud console mode? I'm trying to do all of this
without borrowing anything from work, so although I could drag home a
terminal capable of 9600 baud, I'd prefer to just use what's in the
basement already. I'm assuming the switches on the 11/24 cpu board
(the ones with the ballpoint pen ink all over them...) are the ones I
need to configure.
Thanks again. Believe it or not, I consider this stuff fun. If I get
good at it, maybe I'll have to trade in my VAX GEEK license plate for
one that reads PDP GEEK. Unless that one's already taken...
Note 363.6, 16-Sep-1991
Brown: VT78 comes to the rescue...
----------------------------------
Although I'd still like the switch settings so I can hook up the
DECWriter, I noticed the VT78 (VT78?!?!) in the corner that I'd never
fired up before had a rotary switch that went up to 9600 baud. It
turns out that 4800 baud is where the PDP's set now. But the VT78
actually WORKED, and so, it turns out, does the 11/24!
It's got RT-11 Version 3B. The HELP file seems to be missing, but I
gleaned enough info from an old "DEC Software" paperback that I could
do some easy stuff. A SHOW CONFIG lists my processor board as an
11/34. The board has in the etch pattern "11/24", so does this mean
The DECUServe Journal December, 1994 Page 42
Reviving a PDP-11/24
the RT-11 disk I have was built on an 11/34? (i.e. is SH CONF. showing
the config of the system that built the disk, as opposed to the system
that's running at the moment?) It also points out what I'd already
figured out, that I have the FP-11 and the mapping hardware. How do I
tell how much memory this thing has (short of counting chips)? ...and
how much is the max? I have a BUNCH of core and MOS memory kicking
around from other systems!
The only editor in sight is TECO. Gosh, guess I need to learn THAT
now, huh? All I know is esc esc gets you out....
It has Fortran 77 and MACRO. The RL02 disk is mostly empty (but still
has fewer free blocks than any disk on our VAX system!) and seems to
contain mostly data files for whatever experiments they were running
with this system.
I guess my next bit of fun would be to try to hook up the TU16 that
came with the 11/70 system!
Note 363.7, 16-Sep-1991
Diddle: SLU's
-------------
From the Tec manual :
The PDP 11/24 CPU module contains two serial line units (SLUs).
Serial line unit 1 is used for console terminal. SLU2 is general
purpose.
Switches s1-s8 of E135 are used to select the various baud rates
generated.
SLU1 transmit and receive baud rates are selected by S1-S4 of E124.
SLU2 transmit and receive are selected by jumper leads w4,w9-w13.
Table 2-2 baud rate selection
-------------------------------------------------------------------------
Switch Pack E135
baud rate 1 5 6 7 8
baud rate 2 1 2 3 4
-------------------------------------------------------------------------
50 on on on on
75 on on on off
110 on on off on
134.5 on on off off
150 on off on on
200 continue binary count
300
600
1200
1800
2000
2400
The DECUServe Journal December, 1994 Page 43
Reviving a PDP-11/24
3600
4800
9600
19200 off off off off
Table 2-3 baud rate selection
-------------------------------------------------------------------------
Switch Pack E124
tx 2 1
rx 3 4
-------------------------------------------------------------------------
baud rate 1 on off
baud rate 2 off on
Table 2-4 baud rate selection
-------------------------------------------------------------------------
JUMPERS
TX W11 W10 W13
RX W9 W12 W4
-------------------------------------------------------------------------
baud rate 1 IN OUT OUT
baud rate 2 OUT IN OUT
19.2 BAUD OUT OUT IN
-------------------------------------------------------------------------
Table 2-5 SLU jumper lead selection
-------------------------------------------------------------------------
Jumper lead
SLU1 SLU2 Function
-------------------------------------------------------------------------
w1 None;Fixed at 1. Selects the number of stop bits
IN=one stop bit
out=two stop bits
w5 w7 IN Parity detection enabled
OUT Parity detection disabled
w6 w8 IN odd parity
OUT even Parity
Note 363.8, 16-Sep-1991
Kozam: Nostalgia
----------------
SHOW CONFIG generally makes a good guess at the CPU model by looking
at the hardware registers (i.e. it is not coded when the system was
built). I can't remember the chronology for sure, but I think that
RT-11 Version 3B came out before the 11/24 was even a glimmer in a
developers eye. What is probably happening is that SHOW CONFIG is
working through its 'processor identification' algorithm and winds up
down the path where 11/34 was, at the time, the only known possibility.
Try the command SET EDIT KED to make KED (Keypad EDitor) the default
editor. It is amazingly like EDT. I think KED was available in 3B
(look for the file KED.SAV). If that fails, try SET EDIT EDIT
to invoke a line editor (EDIT.SAV).
The DECUServe Journal December, 1994 Page 44
Reviving a PDP-11/24
Note 363.9, 16-Sep-1991
Bruce: MMU card needed for 22 bits.
-----------------------------------
I can't find the manual, but seem to remember that there is a whole
board you need to go to 22 bit addressing. With it in, you have 4
slots each which can take a 1 meg card. You can only use 1920kw on
a Unibus machine, though. Without it, you only can use 1/4 meg.
Sadly, the 24 is not an I and D space or SUP mode machine, so you are
a tad limited. You CAN put in a pretty nice M+ system even without
those features, but having them would make a BIG difference.
Of course DEC still expects you to TRANSFER the license from the
previous RT user (is that possible on 11s?) and expects to be paid
big $s for any other system you put on it at home...
Note 363.10, 16-Sep-1991
De Larisch: KT24 is PDP-11/24 Memory Management Unit (MMU)
----------------------------------------------------------
Correct ... you are limited to 128KW of memory without the KT-24 memory
management module. You are limited to 4 memory boards (using the MS11-PB
you can stuff 4MB into the box).
The 11/24 is a DOG in comparison with an 11/34. However, you can address
4 MB with the 11/24 and run RSX-11M-PLUS (just don't try to run DECnet/RSX).
Note 363.11
Brown: Thanks for all the help, fellow PDPers!
----------------------------------------------
To .7: THANKS!!! Exactly what I was looking for. Gee, with 2 serial
units built onto the cpu card, and that being the maximum RT-11 can
talk to as terminals, I wonder why they have another terminal
controller card in this thing? I guess maybe they used them for other
I/O?
to .8: Yes, in their startup routine they have a SET EDIT TECO, but I
didn't know to look for those other two. Anything more like EDT than
TECO (KED) would be a welcome sight- I'll look for it! Funny aside:
After doing a DIR *.COM and DIR *.OBJ, I did a DIR *.EXE and turned up
NOTHING. But there did seem to be a lot of "Save" versions of the
normal system files. Hmmm.... Then it dawned on me (DUHHHHHH) that
under RT-11 you can only be running one executable at once anyway, and
that could just be a copy of the memory contents SAVed to disk for
later reloading. Sheeeesh, how pampered we've become.... Good theory
about the SHOW CONFIG and the lack of an 11/24 at that point. I hadn't
thought of that!
The DECUServe Journal December, 1994 Page 45
Reviving a PDP-11/24
To .9:
> I can't find the manual, but seem to remember that there is a whole
> board you need to go to 22 bit addressing. With it in, you have 4
> slots each which can take a 1 meg card. You can only use 1920kw on
> a Unibus machine, though. Without it, you only can use 1/4 meg.
Would that be the card called a "MAP CARD" I have in the machine now?
I have a whole extra memory card in a slot too, plus whatever (if any?)
memory on the cpu card. So you mean the buzillion meg of core I have
(let's think how many boards THAT would be!) is of no use, huh? Oh
well...
> Sadly, the 24 is not an I and D space or SUP mode machine, so you are
> a tad limited. You CAN put in a pretty nice M+ system even without
> those features, but having them would make a BIG difference.
You want I and D space and SUP mode? Lemme fire up the 11/45 or the
11/70! 'Course, I'll need to get a little power upgrade first...
I HAVE RSX-11M (not M+) running on the OTHER machine I brought back to
life: an 11/40. That seems to hang booting off of its former system
disk, probably due to me not having all of the original hardware wired
in and powered. (I haven't yet installed the three 115V 30A outlets I
need for that, but plan to! I've been running it with various units
plugged into a 15A plug strip- AIEEE!!!!!) I can boot the RSX release
disk just fine (I got about 30-40 disks with this thing.) Why don't I
just build a new copy of RSX, tailored to the hardware I've actually
got installed, you ask? Because the thing runs on the dread RK05's!!!!
It has two RKO5j's and one RK05f. I'm told by those who've been there
that an EXPERIENCED person should set aside a FEW DAYS to gen an RSX
system on RK05's...
> Of course DEC still expects you to TRANSFER the license from the
> previous RT user (is that possible on 11s?) and expects to be paid
> big $s for any other system you put on it at home...
Let me see. I buy a system that hasn't been fired up, much less
upgraded, in over 10 years. I buy it for under $10. I spend HOURS
breaking it down into pieces I can carry in my truck and get down the
stairs into my basement, and reassembling it based on mostly guesswork
(like the time I plugged the TU56 cable into the wrong, i.e.
non-unibus, slot...but that's another story...) I diddle with it until
I get it limping. In some cases I even BUY NEW PARTS from DEC to get
it running, increasing my investment a hundred-fold (duct tubes for
RK05 drives- the old ones had turned to slime...). I fire it up and
run it purely for my own bemusement, and to provide fodder for this
low-volume-of-notes conference. I haven't the slightest IDEA who the
previous owner was. If I wreck a disk I can't use the system any more
because I have no backup device anywhere that functions....
And DEC wants me to get all the paperwork in order and maybe even pay
them some money?! COME AND GET ME DEC!!!! HERE I AM!!!! Heck, get
the SPA involved! Raid me! Take my computers as evidence! I could
use the free space in my basement again!
The DECUServe Journal December, 1994 Page 46
Reviving a PDP-11/24
Actually, this raises a valid point. I even remember seeing an article
a coupla few DECUScopes ago where someone mentioned the plight of
people trying to maintain historical computers (OS license per machine
was a GIVEN, why hassle people over paperwork; why bother with ANY
licenses for no-longer-produced-or-supported products; etc.)
My question is: A guy recently offered to give or sell me cheap FIVE
PDP-8's and several CRATES of extra cards and documentation. He even
mentioned something about a 16K hard drive (?!?!). If I'm crazy enough
to get this stuff (and of COURSE if I am, I'll try to get it running!)
do I have to live in fear of DEC hassles over the licenses? I can
almost see it happening...
Note 363.12, 16-Sep-1991
Brown: All control panels run the same speed...
-----------------------------------------------
> The 11/24 is a DOG in comparison with an 11/34. However, you can address
Sure, but for a dollar, the 11/24's not SO bad...
I actually got an 11/34 in a pallet of DEC stuff, but it turned out to
be EMPTY! Oh well, at least I got the control panel for my
collection. It's the good one, with all the buttons. (The
"programmer's console" if I read my aging literature correctly...)
Note 363.13, 17-Sep-1991
Bruce:
-------
> do I have to live in fear of DEC hassles over the licenses? I can
> almost see it happening...
The very real situation is that MV2s are at homes in quantity, and I am
sure some *recent* layered products are probably being run from 2 month
old CONDIST cdroms.
There are a zillion DECies who do this, but G.Q.Public is technically
illegal.
Sure there are some very prosperous businesses based out of homes, but I am
only referring to home computer hacking, and possibly some work-work carried
home.
RSX11M+ may not be hot stuff these days, but DEC still has the nerve
to send mail suggesting I buy the bargain upgrade for an 2/4/8/16 user
MV2 VMS lic to unlimited for *ONLY* $10k.
It would be really 'nice' if DEC made all this legal in some very fair way.
And I mean free or token contribution class 'fair', not 50% of 'LIST'.
These home activities are really to DEC's advantage.
The DECUServe Journal December, 1994 Page 47
Reviving a PDP-11/24
I wonder what 'arrangements' DEC has made for personally owned CPUs at homes
of now 'surplus'd ex-DECies. Do they get to keep running whatever they
previously ran at home?
Note 363.14, 17-Sep-1991
Kennedy: Their Chapter at work (Our Chapter at sleep?)
------------------------------------------------------
> It would be really 'nice' if DEC made all this legal in some very fair way.
> And I mean free or token contribution class 'fair', not 50% of 'LIST'.
> These home activities are really to DEC's advantage.
I think we all need to move to New Zealand. DECUS South Pacific Chapter
members can get _any_ PDP-11 OS license for $25 (I don't know if that's
US $ or NZ $, but it isn't much either way). Further, you are allowed to
legally keep and use the OS media and doc you acquired with the system.
Why doesn't that work in the US?
Note 363.15, 17-Sep-1991
Kennedy: Memory size
--------------------
[Let's see if this one sticks - I posted the original .1 in this stream but
it was eaten by some mysterious conference glitch. Most of the points I
raised have been covered since, but here's another one]
When you power up a '24, it will sit and vegetate for a bit (15 seconds-
2 minutes). It will then print a large, mostly "even" octal number, like
300000. This is the first non-existant memory address the CPU sees. From
this, you can determine what you have in there.
Note 363.16, 17-Sep-1991
Brown: Wow, a deeper issue than I first thought!
------------------------------------------------
RE: licensing for home computers (just to continue off-topic for a
while!)
Good points, all. I can hardly see a way DEC can stop people from
using media that CAME with a junk system they bought. Of course, our
good buddy LMF has EXPIRATION date capability, so when I pick up that
used 6560 system 30 years from now, I won't be able to run diddly-squat
on it!
I consider myself so squeaky-clean on a REAL, BUSINESS side of running
DEC computers that it would irk me to think DEC would want to nail me
for my "fun with obsolete electronics" escapades in my basement. Just
to show you my naivete, it had never occurred to me there was a
home-bootlegging issue with DEC software! It does seem like there
should be some sort of allowance made for older versions and things.
The DECUServe Journal December, 1994 Page 48
Reviving a PDP-11/24
For instance, WordPerfect allows you to donate your older versions of
their programs to schools and churches and the like! What a concept!
then again, with so many new versions of everything coming out so fast,
I can see where they would want to stop the people willing to "suffer"
with VMS 5.3 and VAX C 2.3 on their home system.... pretty complicated
all-in-all....
As a third-order aside, anybody else collect old VMS release media? I
have VMS on 8" floppies all the way back to 1.4 or so. I'd LOVE to get
a set of VMS 1.0 floppies, if anybody who still has them would be
willing to part with them. (I bet VMS 1.1 came out a week later, huh?)
I'll trade you some PDP-11 DOS DECtapes... (who remembers THAT operating
system?!)
Note 363.17, 17-Sep-1991
Brown: The original IBM PC copied that trick...
-----------------------------------------------
> When you power up a '24, it will sit and vegetate for a bit (15 seconds-
> 2 minutes). It will then print a large, mostly "even" octal number, like
> 300000. This is the first non-existant memory address the CPU sees. From
> this, you can determine what you have in there.
Oh thanks, THAT'S handy to know! Actually, the whole system vegetates
for a couple of minutes anyway, since I cut it all on with a power
strip, and the disk and VT78 take that long to come up to speed...I'll
have to try it next time, though.
Note 363.19, 26-Sep-1991
Crowell: The magic number is 8
------------------------------
> Gee, with 2 units built onto the cpu card, and that being
> the maximum RT-11 can talk to as terminals...
RT-11 supports up to 8 terminals, but only one of them (any one) can be
the system console.
> Then it dawned on me (DUHHHHHH) under RT-11 you can only be
> running one executable at once anyway ...
RT-11 can run up to 8 concurrent jobs, though as with any single-cpu
computer, only one job is running at a time. Under RT-11 you can have
one background job, one foreground job, and up to six system jobs.
Note 363.20, 26-Sep-1991
Harvey:
--------
The six system jobs are a relatively new development, aren't they?
Duncan has an old version, pre-11/24.
The DECUServe Journal December, 1994 Page 49
Reviving a PDP-11/24
In any event, even with latest stuff, there is no concept of
timesharing, is there? I.e., no login with separation of resources by
username.
Note 363.21, 27-Sep-1991
Kozam: CTS-300 is timesharing for RT
------------------------------------
While there is no built in concept of login, nor is there a build in
concept of timesharing, it IS quite possible to do BOTH of these
things, although probably not worth the effort.
The CTS-300 product layers ON TOP of RT-11. It runs as a single job
and uses RT-11's multi-terminal support to control several terminals,
allowing each terminal to run different DIBOL programs, thus creating
timesharing (I could say that it creates the illusion of timesharing,
but timesharing is an illusion in itself).
CTS = Commercial Transaction System
Note 363.22, 27-Sep-1991
Provost: TSX-PLUS => RT Timeshare
---------------------------------
> In any event, even with latest stuff, there is no concept of
> timesharing, is there? I.e., no login with separation of resources by
> username.
TSX-Plus is a general purpose time-sharing extension to RT-11.
I don't know if it is still available, but it did meet the
needs you describe for quite a while.
Note 363.23, 27-Sep-1991
Youdelman:
-----------
It's still available (from S&H Computer Systems, 615 327-3670) although
like most other PDP-11 stuff it isn't getting massive updates and lots of
new features added. On the other hand, after using it for many years, it
is now a fairly dependable operating system.
The DECUServe Journal December, 1994 Page 50
Backing Up With RSX
Backing Up With RSX
===================
To continue with a little more non-hardware information, the next
series of notes from the PDP-11 conference discusses dealing with
backups on RSX and RT-11 systems.
Participants: Rob Brown, Milton Campbell, Scott Harrod, Billy
Youdelman.
Note 410.0, 12-Apr-1994
Brown: [RSX] Backups on old RSX or RT-11 systems
------------------------------------------------
I just found out that tomorrow I have to go someplace and teach somebody
how to back up their old PDP-11 system. I haven't seen it yet, but I
suspect that it is either vintage RSX or RT.
If it's RSX, I can handle BRU or DSC. If its older than that, I will
regret giving away my PRESRV manual. Can anybody help? I think the
syntax for a disk to disk copy in PRESRV would be:
PRESRV> DK1:/VE=DK0:
to copy DK0 (yes, these are RK05s) to DK1. Does anybody remember
better than this?
Now, if it is RT, I don't have a clue (yet). How do you do disk
backups in RT-11?
Note 410.1, 12-Apr-1994
Youdelman:
-----------
The first thing to try is the "BR" command (no args). If this works,
use it. It's a third party backup and restore program that's really
easy to use, prompts you thru it, and crams the max possible on to
the least possible tape. The only non-obvious thing about it is if
after stopping somewhere into a tape you want to continue on from
there with something else it's just "-" for the position, not "-1".
If BR isn't there, here's the HELP for the RT-11 BACKUP command just
in case someone zapped it for the disk space.
BACKUP Backup/Restore files or random access devices
SYNTAX
BACKUP[/OPTION] dev:[filnam.typ] dev:[filnam.typ]
SEMANTICS
The BACKUP command transfers files or volume images to or
from multi-volume backup 'savesets'. The Backup Utility
Program (BUP), is useful for making quick archive copies
The DECUServe Journal December, 1994 Page 51
Backing Up With RSX
of RT-11 files and device volumes, and employs various data
verification techniques to ensure data integrity.
BACKUP can also display a directory showing the contents of
device volumes and savesets that it produces.
BACKUP inp:files out:[ssname]
saves file(s) to a saveset on output volume(s)
BACKUP/DEVICE inp: out:[ssname]
saves a device image to a saveset on output volume(s)
BACKUP/RESTORE inp:[ssname/SAVESET,]files out:
restores file(s) from a device image saveset on inp:
BACKUP/RESTORE/FILE inp:ssname[/SAVESET] out:[filnam.ext]
restores saveset to one file on out:
BACKUP/RESTORE/DEVICE inp:ssname out:
restores a device from a device image saveset on inp:
BACKUP/RESTORE/DEVICE/FILE inp:[ssname/SAVESET,]files out:
restores file(s) from a device image saveset on inp:
BACKUP/DIRECTORY inp:[ssname]
creates a listing of saveset sections on the input volume
BACKUP/DIRECTORY inp:[ssname]/SAVESET[,files]
creates a listing of files within a saveset
BACKUP inp:files out:logdsk/SUBSET
creates a logical disk file (logdsk.DSK) and saves
the specified input file(s) inside it
BACKUP/RESTORE inp:logdsk/SUBSET[,files] out:
retrieves file(s) from a logical disk file
BACKUP/DIRECTORY inp:logdsk/SUBSET[,files]
creates a listing of files in a logical disk file
OPTIONS
DEVICE
BACKUP or RESTORE an entire device image. When used with /RESTORE
the output volume is initialized
DIRECTORY
Creates a backup directory listing. The listing is displayed on
your terminal, unless /OUTPUT or /PRINTER is specified.
FILE
Use only with /RESTORE. BACKUP/RESTORE/FILE restores an entire
backup saveset to exactly one disk file. BACKUP/RESTORE/DEVICE/FILE
restores one or more files from a BACKUP saveset. It is equivalent
to BACKUP/RESTORE.
The DECUServe Journal December, 1994 Page 52
Backing Up With RSX
INITIALIZE
Use with backup operation to initialize an output volume
LOG
Informs of beginning of backup, restore, or verify operation.
Also displays names of files processed. (default)
NOLOG
Prevents log messages from being printed on the screen
NOQUERY
Inhibits most confirmation (Are you sure?) prompts. Allows
using BACKUP from indirect command files.
NOREWIND
Prevents initial rewind of magtape before backup or restore
operation. Valid with magtape only.
NOSCAN
Omits bad-block scan of output disks during backup operations
ONLY:n
Limits the number of saveset sections to be displayed during
a /DIRECTORY operation on magtape
OUTPUT:filspc
Use with /DIRECTORY to send output to a file or device
PRINTER
Use with /DIRECTORY to send output to LP:
QUERY
Asks for confirmation before performing an operation (default)
RESTORE
Restores previously backed up files or volume created by
the BACKUP Utility
REWIND
Causes rewind of tape backup volume before operation begins
(default)
SAVESET
Designates a backup saveset for a /RESTORE or /DIRECTORY
operation. The default saveset on disk backups is BACKUP.BUP.
The default saveset on magtapes is the first saveset on the
tape. When /SAVESET is specified with /DIRECTORY, a list of
files contained within the saveset is produced.
SCAN
Performs scan for bad blocks of each output disk volume before
backup (default)
SUBSET
Use this switch in place of /SAVESET to indicate that the backup
is an RT-11 logical disk file. Not valid with magtape.
SYSTEM
Permits wildcard restore of .SYS files to the SY: device
VERIFY[:ONLY]
Provides verification of data transferred to or from a device.
BACKUP/RESTORE/VERIFY:ONLY performs a comparison of the contents
of a backup saveset with an original disk volume without actually
restoring it.
EXAMPLES
BACKUP/DEVICE/INITIALIZE/NOSCAN/VERIFY SY: DU2:
BACKUP/INITIALIZE/VERIFY/NOQUERY DU0:*.* MU0:JUL10A
BACKUP/NOREWIND/VERIFY/NOQUERY DU1:*.* MU0:JUL10B
The DECUServe Journal December, 1994 Page 53
Backing Up With RSX
BACKUP/DIRECTORY MU0:
BACKUP/DIRECTORY/PRINTER MU0:JUL10A/SAVESET
BACKUP/RESTORE MU0:JUL10A/SAVESET,*.FOR DU2:
BACKUP/RESTORE/FILE MU0:JUL10A.BUP SY:JUL10A.DSK
BACKUP/VERIFY *.FOR,*.MAC,*.SAV SY:STUFF.DSK/SUBSET
BACKUP/DIRECTORY SY:STUFF.DSK/SUBSET
Note 410.2, 12-Apr-1994
Brown:
-------
Thanks, Billy.
It turns out that my visit to that site has been delayed, but I have
printed out the information you provided and will take it with me when
I do go.
Note 410.3, 13-Apr-1994
Harrod: For RT-11, COPY/DEVICE might do
---------------------------------------
If it's RT-11, and if you have two RK05's, you can do a disk to disk
copy with COPY/DEVICE. For example
COPY/DEVICE DK0: DK1:
does a block by block copy of DK0: to DK1:. When you have two removable
disks of the same type, it is a relatively easy way to backup the whole
disk.
Note 410.4
Youdelman:
-----------
> When you have two removable
> disks of the same type, it is a relatively easy way to backup the whole
> disk.
Or, if there's a big disk free and some small disks to backup, you can
COP/DEV/FILE DK0: DK1:file.nam
Note 410.5, 14-Apr-1994
Campbell: Check where "DK" is assigned
--------------------------------------
> If it's RT-11, and if you have two RK05's, you can do a disk to disk
> copy with COPY/DEVICE. For example
> COPY/DEVICE DK0: DK1:
The devices might be called "RKx". On RT-11 (as I'm sure Scott knows),
The DECUServe Journal December, 1994 Page 54
Backing Up With RSX
"DKx is the default disk and "DKx" are units of the default disk type.
Scott's command will probably work, unless the default is set to some
other disk type (as it might if it boots off the other device).
Note 410.6
Brown: Job completed successfully
---------------------------------
I finally got to see that system this week. It turned out to be RSX11M
V3.2, so I was right at home. (Well almost. In order to get it to fit
on a single RK05, the original vendor had deleted everything not
required in day to day operation. If those tools were retained
somewhere, we never found them.)
Thanks to all who contributed.
The DECUServe Journal December, 1994 Page 55
SCSI for RT-11
SCSI for RT-11
==============
Still in the PDP-11 conference, but heading back toward hardware
issues, we move on to this discussion of (old) RT-11 and SCSI.
Various emulation schemes are also considered.
Some of you may be wondering about the operating system tag in the
subject of the .0 notes in this conference. That convention came
about when the PDP-11 conference was created by merging a number of
older conferences covering such topics as RSX, RT-11, RSTS, PDP
BASIC, and Pro 300 systems. (Since this issue already has a
decidedly historical tone, we don't mind throwing an occasional bit
of DECUServe history in.)
Participants: Barton Bruce, Milton Campbell, Stuart Dole, Terry
Kennedy, Jim McGlinchey, Billy Youdelman.
Note 323.0, 9-Aug-1990
McGlinchey: [RT11] SCSI Handler for RT-11?
------------------------------------------
Anyone know of an RT-11 handler for a SCSI disk Q-Bus controller?
I have not yet chosen either the disk or the controller.
My choice will come if I can find a handler. Help in choosing
the disk/controller would also be appreciated.
(Yep, Board members gotta make a living, too)
Note 323.1, 9-Aug-1990
Kennedy: A wide choice of emulators; no need for direct SCSI
------------------------------------------------------------
Most controller provide some form of emulation, usually DU. There aren't
many non-emulating SCSI controllers out there. So, just pick a) an emulation
you can use, b) features you want/like (like caching), and c) your favorite
vendor.
Note 323.2, 9-Aug-1990
Bruce: get MU: + DU: for most flexibility
-----------------------------------------
They generally can provide DUAL emulation to let you do MU: also. You can
add a DAT or 8mm or even 9tk when needed. Some will be one the same cable,
and some will offer a second cable on same board (allows more devices total,
too).
The pain is finding anyone who does NO emulation for the VMS 5.3 world where
DEC's port/class driver structure should be most useful for weird devices.
The DECUServe Journal December, 1994 Page 56
SCSI for RT-11
Note 323.3, 10-Aug-1990
Youdelman: Third party handlers for big disks
---------------------------------------------
One thing you may run into with an SCSI disk is its size
will likely be bigger than RT-11 (unless you sysgen v5.5)
can support. There are some third party handlers which
do deal with disks >256 megs, available from:
Networking Dynamics Corp. (Dan Kingsbury)
213 668-0077
Omnex Corp. (Tim Clarke)
408 966-8400
Add On Systems
614 876-0307
Weather Bank, Inc.
801 973-3132
There is also Chester Wilson's DUCM handler available in
the DECUS Library if one is using TSX+ on a Q-Bus system..
Note 323.4, 10-Aug-1990
McGlinchey: Can RT-11 V4 do it?
-------------------------------
Ok, folks, and thanks for the help.
I've got an additional hitch in this problem. The system is
an OEM-supplied optical scanning system, and it's married
to RT-11 V4. The customer can not get the application source code
from the OEM (I hate it when this happens). I gather that
RT-11 V4 did not have a DU handler. Am I right?
Can any of you RT-11 devotees suggest what the customer might do
in this case? Replacing the system is not an option. The entire
system, of which the PDP-11 and RT-11 are in integral part,
cost nearly $1Mil.
So _now_ what are the customer's options? In the worst case,
how difficult would it be to write a DU handler for RT-11 V4,
or retrofit the V5 handler?
Note 323.5, 10-Aug-1990
Kennedy: Older emulations
-------------------------
When I did this (way back when), I used a SCSI controller that emulated
2 RL02 drives. The controller ran a Seagate ST225 (20 Mb) drive through a
SCSI (actually SASI back then) to ST506 adapter.
The DECUServe Journal December, 1994 Page 57
SCSI for RT-11
You may still be able to find such beasts, with assorted emulations. Un-
fortunately, most of the bigger disks were not supported on Q-bus, so there
are oddities about where the address extension register goes, and whether
the driver knows enough to use it. RK07 and RM03 were popular emulations.
Since DU is an auto-sizing protocol, it was quickly adopted as the way to
go for larger drives, or drives with dissimilar geometry from DEC drives.
This is further helped on Q-bus by the fact that the Q-bus DU controllers
use controller-initiated BBR (bad block replacement), so a Q-bus DU board
can do whatever it likes for bad block replacement. On the Unibus, the DU
controllers involve the host in BBR, so emulation is harder to get right.
As far as DU support on RT-11 V4, I don't know. We never had any DU gear
back then. Perhaps someone else knows.
Note 323.6, 9-Sep-1990
Campbell: Try V5 before building a handler
------------------------------------------
> I've got an additional hitch in this problem. The system is
> an OEM-supplied optical scanning system, and it's married
> to RT-11 V4. The customer can not get the application source code
> from the OEM (I hate it when this happens). I gather that
> RT-11 V4 did not have a DU handler. Am I right?
It might be worth trying RT-11 V5 to see how truly "married" the
application code is. A "normal" program would not be able to easily
detect the difference between systems. Even special purpose device
handlers should work without rebuilding them. Of course there are many
non-normal programs around.
I don't remember when DU came in, it was not in V4.0.
> So _now_ what are the customer's options? In the worst case,
> how difficult would it be to write a DU handler for RT-11 V4,
> or retrofit the V5 handler?
There were V4 handlers for DU, do it is doable, probably not for the
faint of heart. One of the early V5 handlers could probably be
retrofitted easier than the current one.
Note 323.7, 21-Sep-1990
Dole: another solution
----------------------
> Anyone know of an RT-11 handler for a SCSI disk Q-Bus controller?
Peritek makes a graphics card for the Q-bus that has a full function SCSI
interface on it. You can program it directly, and at least one person has
used it for lab i/o (with the built-in DMA chip). This may be an odd way
to go, but you don't have to fight emulation. They're at 415-531-6500.
The DECUServe Journal December, 1994 Page 58
Hard-to-find Parts
Hard-to-find Parts
==================
This note from HARDWARE_HELP starts a series of several extracts
dealing mostly with elusive part numbers. The following extract
covers cabinet latches, batteries and unit selectors for old disk
drives, and a monitor tilt/swivel base. (Your editors find the word
"eclectic" well-suited for such conditions.)
Incidentally, we can't resist mentioning that at the time these
notes were posted, Alan Frisbie's personal name in Notes was set to
"Veni, Vidi, $cmkrnli, rebooti".
Participants: Barton Bruce, Alan Conroy, Alan Frisbie, Jack Harvey,
Terry Kennedy, Bill Mayhew.
Note 267.0, 11-Jan-1989
Kennedy: Hard-to-find Parts
---------------------------
This topic will be used to post information about hard-to-find part
numbers.
Note 267.1, 11-Jan-1989
Kennedy: Hex-head cabinet latches
---------------------------------
The latches on DEC cabinets are comprised of three parts - the actual
hex-head casting (part 12-12972-06, $10.00, PDQ required*), the nylon
washer (part 12-12972-03, $5.00, PDQ required), and the clip or retain-
ing ring (part 90-08500-00, $0.63, 30 days).
*PDQ = Product Demand Quote - The part is not orderable through nor-
mal DEC channels. You *MUST* call DEC Unlisted Parts at 603-884-5001
and give them your customer information and the part numbers involved.
They will then return a quote which you must complete and return with
a Purchase Order in order to place an order.
Note 267.2, 11-Jan-1989
Kennedy: RK07 batteries
-----------------------
RK07 disk drives (and other DEC devices) use a 9.6V .250 Amp/hour NiCad
battery pack. The part number is 12-10641-01, which is only available by
PDQ (see *.1). However, you can create this part with *2* 12-10641-00
parts in series. This part is $27.00, 30 days from DEC.
However, it is also available at Radio Shack stores for about 1/4 the
price. (4.8V NiCad battery)
The DECUServe Journal December, 1994 Page 59
Hard-to-find Parts
Note 267.3
Frisbie: The big drawback to buying small parts from DEC
--------------------------------------------------------
>> They will then return a quote which you must complete and return with
>> a Purchase Order in order to place an order.
With a $50.00 MINIMUM per LINE ITEM!!! You may need only
a few of those $1.00 parts, but you will wind up with a
LOT of spares.
Note 267.4, 13-Jan-1989
Conroy: What to do with your spares...
--------------------------------------
>With a $50.00 MINIMUM per LINE ITEM!!! You may need only
>a few of those $1.00 parts, but you will wind up with a
>LOT of spares.
Sell 'em to your friends! ;-)
Note 267.7, 22-Mar-1989
Kennedy: RM02/03/05, TM03 unit select plugs
-------------------------------------------
The unit-number plugs for RM02/03/05 drives are not the same as the vanilla
CDC 9762/66 unit select plugs. The CDC part for the special DEC plugs is
82389500 (for a complete set of 0-7). As of this date, the set costs $13.00
from CDC, or $50.00 from DEC (with a 180-day lead time).
The same plugs appear to work in TM03 tape drive formatters (at least,
the '0' plug from the CDC set mentioned above makes a TM03 formatter #0).
Note 267.8, 22-Mar-1989
Bruce: hign # blue = low # black
--------------------------------
> The unit-number plugs for RM02/03/05 drives are not the same as the vanilla
>CDC 9762/66 unit select plugs. The CDC part for the special DEC plugs is
>82389500 (for a complete set of 0-7).
The vanilla CDC 9766 plugs have blue letters, the TU16, TM02/3, etc plugs
have black letters. There are only 4 bits so there are 16 combinations.
If you get a full set of 16 with a 9766, you will find you can match
your TU16 buttons, but the labeling will be off.
It has been a few years, but I think black 0-7 TU16 ones are blue 8-15,
and maybe even backwards.
The DECUServe Journal December, 1994 Page 60
Hard-to-find Parts
Note 267.12, 16-Aug-1989
Harvey: Tilt/swivel base for VR260 19" monochrome monitor
---------------------------------------------------------
From "VAX Systems and Options, 1989 April-June", p III.36,
Tilt/Swivel VR10X-AA for 19-inch monochrome monitor.
From Oct 1, 1988 (sorry, note hour) U.S. Price Book:
VR10X-AA VR100/VR260 Tilt/Swivel $295
Note 267.13, 16-Aug-1989
Mayhew: VR260 tilt/swivel not compatible with VR262, by the way
---------------------------------------------------------------
Note for the record that DEC says the VR10X-AA is NOT compatible
with the VR262, which is the replacement for the VR260. There is
no known tilt/swivel that works with the VR262.
The DECUServe Journal December, 1994 Page 61
MicroVAX II Battery
MicroVAX II Battery
===================
Here is a short episode from the HARDWARE_HELP topic titled "Looking
for a battery", concerning a recent and very successful attempt to
find a replacement for the clock battery in a MicroVAX II. As an
extra bonus, the stream includes a DECdirect (Canada) success story!
(See? See? We don't just bash 'em....)
Participants: Gus Altobello, John Briggs, Jamie Hanrahan, Terry
Kennedy, Jean-Francois Mezei.
Note 1280.2, 27-Sep-1994
Mezei: Microvax II clock battery
--------------------------------
I recently rebooted my all mighty Microvax II, and noticed that the
clock magically moved back some 8 minutes. The time was pretty well
correct prior to the reboot. (caused by power failure lasting about one
hour).
Sine the beast was built in 1987, I assume that it is time to change
the battery. Can someone tell me what the part number is, or is that a
rechargeable battery that should never need to be changed ?
I have made a call to DEC DIRECT but don't expect they will call me
back with the answer.
Note 1280.3, 27-Sep-1994
Kennedy: Explanation
--------------------
Not necessarily. The time on a running VAX is pre-loaded from the TOY clock
(the battery one) on startup, but then is updated by the precision timer. On
a normal shutdown, this time is written back to the TOY clock. I believe SET
TIME (no arguments) also does this. On a power failure, the TOY isn't updated,
so it's got all the accumulated clock drift since the last shutdown. Jamie
Hanrahan is the expert on all of this...
If the battery failed, you'd get other problems, like not remembering the
language setup.
> I have made a call to DEC DIRECT but don't expect they will call me
> back with the answer.
That is the wrong place to call. You need Part Number Assistance,
1-603-884-5000.
The DECUServe Journal December, 1994 Page 62
MicroVAX II Battery
Note 1280.4, 27-Sep-1994
Mezei: Weak battery = imprecise clock ????
------------------------------------------
Well, my experience with battery operated clocks is that when batteries
starts to get weak, the clock starts to drift faster and faster.
So, after that power failure, I saw the huge time difference and
thought that the TOY battery is getting weak and should be replaced
soon. After all, it is been there since 1987 and I figure that if a
battery lasts 7 years, it is pretty good.
Well, if DEC DIRECT (Canada) isn't smart enough to re-direct me to that
part-assistance number Terry was kind enough to list, then there is
more more pet peeve I can have against them :-) :-) :-) :-)
Note 1280.5, 27-Sep-1994
Hanrahan: Dunno if I'd say I"m the expert, but...
-------------------------------------------------
> Well, my experience with battery operated clocks is that when batteries
> starts to get weak, the clock starts to drift faster and faster.
irrelevant. The clock only runs on the battery for the brief time during a
power failure. What you're seeing is the drift between the TOY clock and
system time that occurred *while the VAX was running*.
> After all, it is been there since 1987 and I figure that if a battery
lasts
> 7 years, it is pretty good.
OTOH, if nothing is broke, why fix it? As Terry said, if your battery had gone
dead during the outage, your system would have gone through the language
prompt sequence when it was powered back on.
Those batteries always looked to me to be standard AA NiCads cells.
Note 1280.6, 28-Sep-1994
Kennedy:
--------
> Those batteries always looked to me to be standard AA NiCads cells.
They're in a 2-pack with the nice power leads and connector. If JF wants
the part number, I'll dig it out tomorrow. Or, he can use the part number
hotline.
The DECUServe Journal December, 1994 Page 63
MicroVAX II Battery
Note 1280.7, 28-Sep-1994
Briggs: For what it's worth
---------------------------
> ...is written back to the TOY clock. I believe SET
> TIME (no arguments) also does this.
$ SET TIME with no arguments sets the system time from the TOY clock
$ SET TIME="''f$time()'" sets the TOY clock from the system time
Note 1280.8, 28-Sep-1994
Altobello: You hit the HALT button, didn't you?!? :-)
------------------------------------------------------
JF, did you do a "proper" shutdown and reboot, or did you hit the halt
button and type "B"?
Running SYS$SYSTEM:SHUTDOWN.COM executes the $SET TIME mentioned
earlier that sets the current time into the TOY clock. If you did a
normal shutdown I would've expected the TOY clock to have the proper
time, regardless of any drift it may've experienced.
Note 1280.9, 28-Sep-1994
Mezei: Miracle: Got the part number !
-------------------------------------
Sorry, it was a power failure. I was in the kitchen at the time, and
didn't have the time to run down to the basement and do the "SET TIME"
command before the power ran out completely ....
Miracle: Digital actually called me back today. I think it is the first
time that Digital Direct actually calls me back. I should celebrate !
(Or should I worry because they have so few customers that they can now
spend time on such small customers :-)
Part number for the Microvax 2 battery is:
12-19245-01 and retails for CAD$19.18 (about USD13.00 I guess).
The reason I brought it up is that I never had that problem (not such a
time difference) before, and figured that the battery clock was
starting to be quite slow. (I think I had been up for about 30 days
before that power failure).
The DECUServe Journal December, 1994 Page 64
RL02 Spare Lightbulbs
RL02 Spare Lightbulbs
=====================
This next stream from HARDWARE_HELP is about as specialized as
things can get. The topic is... spare light bulbs for RL02 disk
drives. Ordering, proper voltages, GE part numbers, the works. We
suspect this is all most readers will ever need to know about RL02
light bulbs.
Participants: Ed Cetron, Pierre Hahn, Terry Kennedy, Jeff Killeen,
Marc Kozam
Note 147.0, 6-Jun-1988
Killeen: RL02 Parts
-------------------
Does anyone have any ideas on how to get light bulbs for RL02's ?
Note 147.1, 7-Jun-1988
Kennedy: Send me MAIL
---------------------
> Does anyone have any ideas on how to get light bulbs for RL02's ?
Well, you call Unlisted Parts, find out you have to order 144 of them, and
that the lead time is 120 days...
Or, send me a mail message stating how many of the 143 I'm not using
you want...
P.S. - RL02 and RK07 bulbs are *not* the same!
Note 147.2, 7-Jun-1988
Kozam: RL02 Bulbs == GE 73 Type
-------------------------------
> Does anyone have any ideas on how to get light bulbs for RL02's ?
RL02 bulbs are all GE 73 type (yes, that's General Electric). Not
too long ago, I purchased a package of 10 from a local electrical supplier.
The cost was $ 3.00 for the package. I don't recall if it was special order.
Note 147.3
Hahn: get higher voltage bulb
-----------------------------
I have now forgotten where after blowing 5 bulbs we purchased what
I believe are higher voltage bulbs (they are dimmer) to put into
RL01/2. It was a radio store. If you want I will pull out one
of them and look.
The DECUServe Journal December, 1994 Page 65
RL02 Spare Lightbulbs
Note 147.4, 7-Jun-1988
Kennedy: Careful!
-----------------
> ... we purchased what I believe are higher voltage bulbs...
Careful - there are two types of drivers used for the RL02 front panel
lamps. Your lamps probably are similar to those used in the RK07 and will
blow the other driver type used in the RL02.
Note 147.5, 11-Jun-1988
Cetron: I see the light...
--------------------------
My field service rep will supply me with as many as I need -
and usually many more - he HATES to come all the way just to replace
a 'bloody' light bulb.... I think I have 40 spares... Do you want
one Jeff....
The DECUServe Journal December, 1994 Page 66
Old Hardware Potpourri
Old Hardware Potpourri
======================
It's an interesting feeling to come across a note titled
"Information on Old Hardware" and discover that it was started in
1987. (Indeed, this was one of the very first notes in the
now-familiar HARDWARE_HELP conference.) Most of what we've been
discussing as "old hardware" in this special issue of the Journal
was state-of-the-art back then... and (for some of us, anyway) it
doesn't seem like that long ago.
This brightly-colored mix of tidbits seems like a good way to wrap
up the Journal's first of what we hope will be many "special
issues." The notes discuss a mysterious Q-bus board, an RX02 belt,
and settings for a DMF-32. Your faithful editors hope you've
enjoyed this romp through the dusty archives; please let us know how
you liked it, and what you'd like to see more of. Contact
information is at the end of every issue.
P.S.: As far as we can tell, the question about the "PDP Maxima"
never got answered; a ForWords search of all the conferences on
DECUServe turns up nothing (except for several "mea maxima culpa"
notes). Anyone care to take a crack at it?
Participants: Barton Bruce, Alan Frisbie, Pierre Hahn, Terry
Kennedy, Larry Kilgallen, Shane MacDougall, Gary Maxwell, Bill
Mayhew, Kerry Powell.
Note 7.0, 25-Mar-1987
Maxwell: Information on Old Hardware
------------------------------------
I hope I am not being too presumptuous in writing a note asking
for information on ancient hardware that some people are still
using somewhere in the Universe.
Does anyone know about a "PDP Maxima" computer system that uses
the CSP Language? They apparently exist in Europe.
Does anyone have documentation for a DKC11-AB (a W9512 module)?
This is an old bus foundation module. I have heard from Bruce Mitchell
that there is nothing more written about it other than an overview
document. I guess this was a piece of WYSIWYG hardware.
Note 7.3, 28-Mar-1987
Frisbie: Answer for DCK11-AB question
-------------------------------------
The W9512 is simply the bare dual-width Q-bus wire-wrap
prototype module. The data sheet I have is dated March 1978
and shows a photograph, an outline drawing and a listing of
the VCC and GND pins. All this should be obvious if you
have the board, but I can send you a copy if you need it.
The DECUServe Journal December, 1994 Page 67
Old Hardware Potpourri
The DCK11-AB (not DKC), on the other hand, is a W9512 module
plus the following Q-bus interface chips:
DC003 (1) Interrupt logic
DC004 (1) Protocol logic
DC005 (4) Bus Transceivers
DC006 (2) Word Count / Bus Address
DC010 (1) Bus Master logic
This information is from Fall 1978 copy of Digital Catalog Sales,
pages 58-61. I can send you a copy of this if you wish. I also
have (somewhere under the many stacks of paper in my office) a
copy of the Chipkit User's Guide for the above chips. I can send
you a copy of this 16-page guide, plus a later micro-note on
extending the DMA logic from 16- to 18-bit addressing (remember,
this was back in the dark ages of the Q-bus).
Note 7.4, 8-Jun-1988
Hahn: RX02 belts
----------------
Need part number for RX02 belt, my drive failed; the belt was laying
in the bottom of the tray! Where do I get a few? Pierre
Note 7.5, 8-Jun-1988
Mayhew: RX02 belt part number and price
---------------------------------------
According to my DEC Self-Maintenance Services book, part number
12-14005-00 is $7, assuming you have a 60Hz drive. Otherwise, you
want the -01 variant, same price. Should be orderable through
DECdirect.
Note 7.9, 2-Jul-1991
MacDougall: 11/730 DMF-32 pin settings
--------------------------------------
Does anyone know the correct DIP switch settings on the 3 switchbanks
on the 11/730 to make the ports 2 thru 7 modems. I have made ports 0
and 1 modems, but would like to designate the remaining ports also.
Note 7.10, 2-Jul-1991
Kilgallen: Impossible?
----------------------
I thought only the first two could provide modem control.
The DECUServe Journal December, 1994 Page 68
Old Hardware Potpourri
Note 7.11, 2-Jul-1991
Kennedy: That's correct
-----------------------
> I thought only the first two could provide modem control.
Note 7.12, 3-Jul-1991
Bruce:
------
Supposedly full modem support is there in the chips DEC installed
on the controller, but they saved $s by not having enough ribbon cable
wires to carry the signals to to bulkhead! Guess who should have
been included in one of their lay-offs...
Note 7.13, 3-Jul-1991
Kennedy: In defense of the DMF-32
---------------------------------
You could say the same thing about the DLV11-J, you know 8-)
The 2661 UARTs support modem control, however there is no provision on
the board for it. If you look at the component side of the DMF, you will
see two vertical rows of UARTs on the bottom right-hand side. Look at all
of the right-hand row and the bottom two of the left-hand row. You'll see
etch connecting pins 16 (CD), 17 (CTS) and 22 (DSR) to pin 4 (GND), which
permanently makes those signals "active".
The last time I looked at the DMF microcode, I don't think it tried to
set/read the modem states for those ports, even if the host asked it to.
Remember that the DMF was introduced as a combo board to reduce board
count in bounded systems (11/730). It replaced the set of LP11, DZ11,
. I believe DEC thought that if you wanted lots of modem
control ports that you'd get a DH or DZ. The idea was to reduce the cost
of smaller systems by reducing board count.
Note 7.14, 10-Nov-1991
Powell: The ABLE VMZ32 had modem control on all lines.
------------------------------------------------------
If you really want a lot of terminal ports with modem control, the
ABLE-VMZ32 provides 16 ports, all with modem control, on a single hex
(unibus) board (it emulates the async part of _two_ DMF32s).
They should be readily available for little $ on the used market. ( I
even have a couple gathering dust)
The DECUServe Journal December, 1994 Page 69
Old Hardware Potpourri
Note 7.15, 11-Nov-1991
Bruce: Emulex as well as ABLE and others
----------------------------------------
Or several EMULEX models. Some do just 16 lines. One model does
up to 64 ports using smart external panels. Changing its proms
gave you DH, DV, or DMF32 emulation. I think in DMF32 mode, the
older Ubus card only supports 3 panels for 48 ports.
Many were sold as PDP11 DHs, but new proms are $250 and convert
you to another emulation. With the same smart distribution panels
(at least the newer models) - and that is where the money is - you
can get a different Q bus card for about $1k that will run them
as 64 DHV ports (8 x DHV), or a newer Unibus card that does 128
DMF32 ASYNC ports all with modem control. They even support 422
(yes folks, BALANCED) from the panel, as well as out of band flow
control (a la DS200) without confusing the terminal driver.
All of that from a single slot in the host bus.
The DECUServe Journal December, 1994 Page 70
About the DECUServe Journal
About the DECUServe Journal
===========================
Publication Information
Topic threads in the DEC Notes conferences on DECUServe are selected
for publication on the basis of strong technical content and/or
interest to a wide audience. They are submitted to the editor from
various sources, including DECUServe Moderators, Executive Committee
members, and other volunteers. Suggestions for inclusion are
enthusiastically solicited. Articles selected for publication are
edited on an OpenVMS VAX system in TPU and then formatted with
Digital Standard Runoff.
What's a DECUServe, Anyway?
DECUServe is an on-line conferencing system which is available 24
hours a day, seven days a week. System backups are done on Fridays
from 7:00am to approximately 9:00am Eastern (Daylight/Standard)
Time. Scheduled downtimes for equipment or software maintenance are
announced in advance. The system can be reached worldwide, and
membership is by individual subscription only (no group or company
accounts). The current annual subscription fee is US$75.00.
On-line subscription information is available in the U.S. by
dialing 1-800-521-8950 and logging in with username INFORMATION.
DECUServe and the INFORMATION account can also be reached on the
Internet via telnet connection to decuserve.decus.org or e-mail to
information@decuserve.decus.org.
Contact Information
===================
The editors of the DECUServe Journal are Brian and Sherrie McMahon.
They can be reached by any of the following means:
mcmahon_bdecuserve.decus.org
mcmahon_sdecuserve.decus.org
mcmahonbdecus.org
griffithdecus.org
mcmahonac.grin.edu
griffithac.grin.edu
+1 515 269 4901 (normal office hours, U.S. Central time)
+1 515 269 4936 FAX