• Please review our updated Terms and Rules here

IBM 5150 - MDA - Parallel port Error 432

fdiskitup

fdiskitup

Experienced Member
Joined
Mar 2, 2022
Messages
247
Location
Chicagoland, USA
I just got this 5150 back to life, I'm running the IBM diagnostics disk 1.x and it gives me error "432" - all is not right with my parallel port on the display adaptor.
There is nothing connected to the Parallel port, does this test always require a loopback adaptor ?
1678330805148.png 1678330938110.png
 
This is from the April 1984 manual, which includes the 2.X diags, but it suggests that the wrap plug is mandatory. And it is included with the manual.
 

Attachments

  • A68D1367-74F4-4E56-B53B-6B21A0CBC930.jpeg
    A68D1367-74F4-4E56-B53B-6B21A0CBC930.jpeg
    2.1 MB · Views: 11
  • 09072F1E-5E1B-48F3-BEAA-D3D10283C233.jpeg
    09072F1E-5E1B-48F3-BEAA-D3D10283C233.jpeg
    1.2 MB · Views: 11
Any ideas on the pin connection to make one ?
I only found the serial loopbacks on minuszero.net .
Apparently parallel loopback is non standard, I found a checkit version and a norton version.
Also, to fix this 5150 just needed a tantalum replaced C6 on the 12v line - add it to the list. Tant was shorted not exploded!
 

Attachments

  • A744ACE4-D99D-4F19-9035-5DE7A3E40C1E.jpeg
    A744ACE4-D99D-4F19-9035-5DE7A3E40C1E.jpeg
    1.2 MB · Views: 4
Last edited:
I ran 'Advanced Diagnostics 2.20' on my IBM MDA equipped IBM 5150.
Via various navigation, I then chose the test of, 'IBM MONOCHROME DISPLAY AND PRINTER ADAPTER TEST'.
Under that, I chose the subtest of, '4 - PRINTER ADAPTER TEST'.

On-screen appeared, INSERT WRAP PLUG AND PRESS ENTER.

I attached my parallel port loopback plug then pressed ENTER. After a couple of seconds, I was returned to the menu for the IBM MONOCHROME DISPLAY AND PRINTER ADAPTER TEST. There was no indication of success.

I removed my parallel port loopback plug, then repeated the '4 - PRINTER ADAPTER TEST' test. This time, a 432 error was reported.

The parallel port loopback plug that I used was the one that I created donkey years ago for CheckIt.
Photo follows.

1678342277315.png
 
The original appear to be wired the same:


EAEC6637-C43F-4E0F-898B-F1D3B867DC72.jpeg
068A4E71-F763-49A5-8B81-E47B6CD0F40C.jpeg

I have the landmark service diag dongles too, and I’ve not tested them with the IBM software, but they appear to be wired slightly differently:

42AA8EDD-2762-4DDC-8A29-B8147D7D007C.jpeg
 
Today I built my own parallel loopback,
It works on my hp vectra so the connections are good,
but still no luck on the 5150 !973A6079-BA7C-4572-A910-47077674B051.jpeg 0656CC82-351D-447D-AA04-04058CE4CCEF.jpeg
I tried a newr disk diagnostic 2.20 from 1985, same error plus a new one error 134.
 
Years ago, I wrote my own DOS based parallel port diagnostic. For 'standard' ports (not ECP, etc.) such as what one would expect in an IBM PC. Give that a try and let us know the results. It is called PARTEST, and you can get it from [here].

( On my to-do-list is to investigate its interrupt test, which does not work on one of my multi-function cards [card problem, or my interrupt triggering code needs a tweak?]. )
 
Is there any way the 134 error (cassette port) is related to the parallel port 432 error?
Does the -5V get used on the MDA card ?

I followed the troubleshooting for the cassette port on minuszerodegrees.net. Power connector P9 is connected ok.
Looking for the -5V on pin 4 of U1 (actually easier to measure on pin 4 of the empty 8087 socket) gave me -4.1 V and -3.85V when I booted into cassette basic, and ran this
10 save “g”
20 goto 10

The solenoid does click, I let this prog run for 30 mins to try and work the solenoid.
The 134 error doesnt show up in the advanced diagnostics test 2.20 after this workout. Ill check after a cold boot tomorrow.

A second line of inquiry is into the parallel port on the multifuntion fastcard IV - is this causing a conflict LPT1=LPT2?
 
fdiskitup said:
Is there any way the 134 error (cassette port) is related to the parallel port 432 error?
Click to expand...
You mean "131"?
Not that I can see.

fdiskitup said:
Does the -5V get used on the MDA card ?
Click to expand...
No.

fdiskitup said:
A second line of inquiry is into the parallel port on the multifuntion fastcard IV - is this causing a conflict LPT1=LPT2?
Click to expand...
Probably better to refer to these ports by their base I/O addresses, e.g. 3BC, because what DOS refers to as LPT1 can be referred to as LPT2 elsewhere.

fdiskitup said:
Looking for the -5V on pin 4 of U1 (actually easier to measure on pin 4 of the empty 8087 socket)
Click to expand...
The reason for measuring on U1 is to cater for an open trace on the PCB, or bad solder joint. I.e. U1 is where it counts.
 
fdiskitup said:
A second line of inquiry is into the parallel port on the multifuntion fastcard IV - is this causing a conflict LPT1=LPT2?
Click to expand...
modem7 said:
Probably better to refer to these ports by their base I/O addresses, e.g. 3BC, because what DOS refers to as LPT1 can be referred to as LPT2 elsewhere.
Click to expand...
Adding to that, the IBM diagnostic is doing things at the hardware level. It knows that the parallel port on an IBM MDA card starts at I/O address 3BC, and so the registers at there are what the diagnostic is targeting.

fdiskitup said:
A second line of inquiry is into the parallel port on the multifuntion fastcard IV - is this causing a conflict LPT1=LPT2?
Click to expand...
Presumably, you are simply going to remove the FastCard as an experiment.

IBM won't provide specifics about the 432 error . Their view is, 'There is an error - replace the card with new.' It is expected that the tool that I pointed to in post #8 should show that a particular sub test is failing. Then reference to the circuit diagram, and the use of a logic probe or oscilloscope, should point to the problem component on the card.
 
Uggh, it’s never simple is it.
I eliminated the Fastcard IV from our inquiries, the fault code still appears when it is not in the machine. But also when it is not in the machine 64k is not enough to run PARTEST….

LPT1=03BC (fails code 432)
LPT2=0278 (fastcardIV, passed the “partest”)

Running the PARTEST program, on lpt1/03BC with the loopback gives me some fails.

958A2606-CABB-46C1-97E5-4775F9D247FB.jpeg
 
fdiskitup said:
Running the PARTEST program, on lpt1/03BC with the loopback gives me some fails.
Click to expand...
'Toggle of I/P line BUSY' test

The loopback plug loops the SELECT INPUT line (pin 17) to the BUSY line (pin 11). The 'Toggle of I/P line BUSY' test toggles the SELECT INPUT line, expecting (because of the loopback wire between pins 17 and 11) to see the BUSY line toggle. We can see from the failure of the earlier 'Toggle of O/P line SELECT' test that the card's SELECT INPUT line cannot be toggled, and so that is a probable cause of the failure of the 'Toggle of I/P line BUSY' test.

'Toggle of O/P line SELECT' test

So this failure is what is to be targeted.

Note that on the DB25 connector, there is a SELECT line (an input to the card) and a SELECT INPUT line (an output of the card). Confusing. I am writing of the latter. As confirmation, a look at the source code for my PARTEST shows that I am toggling bit 3 of port 3BE, which corresponds to pin 17 on the DB25 connector, SELECT INPUT.

The circuitry unique to the SELECT INPUT line is shown at [here].
Q1. Do have a logic probe or oscilloscope?
Q2. If the loopback plug is removed, does the 'Toggle of O/P line SELECT' test still fail?
 
modem7 said:
The circuitry unique to the SELECT INPUT line is shown at [here].
Q1. Do have a logic probe or oscilloscope?
Click to expand...
yes, both. (and a XGecu rom programmer that will "check" 74LS chips and a solder sucker to extract the IC's).
modem7 said:
Q2. If the loopback plug is removed, does the 'Toggle of O/P line SELECT' test still fail?
Click to expand...
selecting the "nothing connected test" - yes! Toggle of O/P line select --> FAIL
selecting the with loopback test - yes ! but also Shorts test FAIL - data lines affect I/P busy.
so lets see if there is a short on the port to ground.
735822CF-AF35-49F4-9A71-261D3A275F31.jpeg
 
Pin 17 of the parallel port connects to Vcc on every chip on the board.... I dont think thats normal ? always high ?
(with the board out of the computer)
 
Last edited:
fdiskitup said:
Pin 17 of the parallel port connects to Vcc on every chip on the board.... I dont think thats normal ?
Click to expand...
Zero, or a few ohms (i.e. not continuity mode). And that is with the loopback plug removed.

Looking at the diagram at [here], it seems unlikely that the pull-up resistor will have gone short. Could be a problem in U37 or U38.
 
You must log in or register to reply here.
Back
Top