Re: So... Who's coming this weekend? And what should I bring?
- Mr Ian Primus <ian_primus@...> wrote:
>There's really no time to Q&A you about "the problem". Here's my unasked for advice.
> > >> I really would love to get my PDP-11/05 running
> I already know which board is bad - it's the control board -
>M7261. I also know that the computer will run with a good
>one installed. The problem I have is trying to troubleshoot
> and repair the control board.
You will need an oscilloscope, docs for BOTH boards with schematics, a hex extender. A book of TTL logic data would help (I will bring that for my use.) And (excuse me) some clue about using these. But I don't think a public email discussion, is a very good place to discuss how to repair a control board or to describe whatever difficulties you yourself may have in its repair. Certainly not in 24 hours.
So I frankly think your task, is to contact privately some of the attending persons, to see how helpful they can be IN DETAIL. Someone will have to commit to bring extender cards. Evan has committed to provide the MARCH 11/05. I think you should scrounge up the M7260 Data Path Module's and M7261"s docs, but let's see.....
may have sufficient hardware docs, otherwise look around if you don't have them already. "bring" them on a laptop.
These old cards go by on eBay but there's none there today.
You will apparently have an opportunity to compare your NONworking 11/05, to a WORKING 11/05, side by side. Take advantage of that as best you can, down to the chip and logic level.
- I don't need these for my projects, but I can bring any of the
Tektronix 485 scope (dual-trace, 350MHz analog)
HP 1650B Logic Analyzer
Advantest TR5823H Universal Counter
Data I/O 29B EPROM Programmer
Drop me a note if you want me to bring any of them.
Mike Loewen mloewen@...
Old Technology http://sturgeon.css.psu.edu/~mloewen/Oldtech/
- --- On Fri, 6/29/12, s100doctor <hjohnson@...> wrote:
> You will need an oscilloscope, docs for BOTH boards withI've already done a fair bit of troubleshooting. The only thing I lack is card extenders. I've got the schematics and the microcode listings printed out, and I did spend a fair bit of time checking the TTL associated with the nonworking function (deposit). I do not fully understand the microcoded portion of the logic, although I have verified the bits I do understand. Another set of (experienced) eyes would also definitely be beneficial here.
> schematics, a hex extender. A book of TTL logic data would
> help (I will bring that for my use.) And (excuse me) some
> clue about using these. But I don't think a public email
> discussion, is a very good place to discuss how to repair a
> control board or to describe whatever difficulties you
> yourself may have in its repair. Certainly not in 24 hours.
Lacking card extenders, I've had to extend the individual signals from specific chips to probe them. This gets tedious and is difficult. I'm pretty sure I can find the fault, given full access to the board while it's installed in the computer.
> Someone will have to commit to bring extender cards.One person has already mentioned that they'll bring some extenders.
> You will apparently have an opportunity to compare yourThat was how I was able to quickly and easily verify that my fault lies on the Control board. I've got a friend that has a working machine. We've compared some things between the machines, and done some swapping and testing, but again, lack of card extenders makes individual logic checks tedious and difficult. I DO know that my computer works properly when a functional Control board is installed. And that the symptoms move to a working computer when my faulty Control board is installed.
> NONworking 11/05, to a WORKING 11/05, side by side. Take
> advantage of that as best you can, down to the chip and
> logic level.
But yes, having card extenders, and a working control board borrowed from one of MARCH's machines would greatly help to clarify what is supposed to be happening on the board when the deposit switch is pressed.
 For those of you playing along at home, the fault is that when the deposit switch is pressed, the address increments but nothing is written to memory. Examine works. The CPU can write to memory itself - if you write a program that will alter memory when run, toggle it in with a good Control board, then swap in the faulty Control board, and run the program, it WILL write to memory. The only fault is that the deposit switch itself does not write to memory. I've tested the flip flops and gates connected directly to the deposit switch, and everything appears to function correctly. *Part* of the problem is that I don't fully understand exactly how the microcoded portion of the logic is supposed to work, however. Another difficulty is that the available schematics are of very poor quality and are difficult to read in places. Either they're bad scans, or just good scans of bad photocopies. Either way, they do add an extra level of frustration to an already
- On 06/29/2012 11:06 AM, s100doctor wrote:
> http://bitsavers.org/pdf/dec/pdp11/1105/11/05 CPU cards don't. Precisely three have hit eBay in the past 4-5
> may have sufficient hardware docs, otherwise look around if you don't have them already. "bring" them on a laptop.
> These old cards go by on eBay but there's none there today.
years. The 11/05 is one of the earliest PDP-11s, from 1972. Not many
of these systems were made (comparatively speaking), and they dried up
many years ago.
Other Unibus cards, appropriate for use in that system, aren't tough
to find. The two CPU boards, though, really need to be repaired because
finding replacements is nearly impossible now.
Dave McGuire, AK4HZ
New Kensington, PA
- Glad Ian has done a fair amount of testing and diagnosis down to the chips. But...it's too much work to diagnose chips by email. (shrug) so I respect his analysis but won't reply to it.
Reading between the lines, Ian needs a set of bus extenders to expedite his work, and maybe better schematics. Someone with 11/05 prints or photocopies of them would expedite his work.
I misplaced my DEC bus extenders, I actually looked for them. But Gesswein will bring his.