• Please review our updated Terms and Rules here

Ruud's Diagnostic ROM

I replaced chips U7 and U14 it did not fix the problem.
I installed ST T74LS373B1 for U7 and Motorola SN74LS244N for U14
Failure address: o KB
Post card shows: 00 FF
I noticed that my logic probe shows nothing on pin1 of the U39 and U40 now
And the processor is quite hot
 
Failure address: o KB
That is from Ruud's Diagnostic ROM.
Is that still an an address failure. ('Testing RAM - Address') ?

Post card shows: 00 FF
Which type of POST card: ISA slot, or parallel/LPT ?

( Modern ISA types are not expected to work in an IBM 5160, even for Rudd's Diagnostic ROM. )

And the processor is quite hot
That is normal.

I noticed that my logic probe shows nothing on pin1 of the U39 and U40 now
So things have gotten worse.

Per the diagram in post #29, that is the ADDRSEL signal. Per the diagram at [here], ADDRSEL comes from the delay line. The RAS signal goes into the delay line, and the delay line produces ADDRSEL and a signal for pin 2 of U69. Does your logic probe show a RAS signal ?
 
Further to my previous post:

By the way, when you are looking for signals at certain points of the motherboard circuitry, you always need to keep in mind that some signals are only present for certain activities.

For example, RAS, CAS, and ADDRSEL are only present when the motherboard is reading or writing to motherboard RAM.
If I was to use the TEST6065 test code for ROM at [here], that code does no reading nor writing of motherboard RAM. No RAS, CAS, or ADDRSEL, will occur.

For example, if I had test code in ROM that only tested bank 0, then there will be no CAS1, CAS2, and CAS3 signals during that testing (only CAS0).

Hardware/software knowledge is required.

In your case, you are using Ruud's Diagnostic ROM, and it is getting as far as the motherboard RAM tests. Because of that, we know from [here] that the 8253 timer chip and 8237A DMA chip have been set up to do dynamic RAM refreshing in the background. If working, that refreshing process continuously reads the motherboard RAM (even when the CPU is later halted).
 
Is that still an an address failure. ('Testing RAM - Address') ?
No, it stops at Check first 2 KB of RAM now
Which type of POST card: ISA slot, or parallel/LPT ?
ISA card. Sometimes it shows 00 FF, another time 00 04, 00 F4, 00 89 but maybe it doesn't matter
So things have gotten worse.
The motherboard now behaves unpredictably.
Sometimes it displays an image on the screen and then shows a first 2 KB error.
But sometimes it doesn't turn on and then the POST card only displays "bA"
Most often it does not turn on when I move the POST card to another ISA slot or when I remove and insert the U14 chip.
(The motherboard behaves exactly the same when I remove the U14 chip...)
When the motherboard does not display anything on the screen then there is no pulsation on pin 13 in the 8253 chip

I haven't found a reason why the motherboard sometimes doesn't turn on. Perhaps there is something wrong with the U14 socket, but I'm not sure. When I remove and insert the U14 chip several times or when I remove the POST card and insert it into another slot, the board starts working.
I think it's not a problem with the U14 chip because I have 3 different ones and each of them sometimes works and sometimes doesn't.
The socket is probably soldered correctly. I measured each pin on the top and bottom of the motherboard with a mult
imeter.

When it works, the readings on the memory chips are as follows:
A0 – High
A2 – High
A1 - Pulse
A7 - Pulse
A5 - Pulse
A4 - High
A3 - Pulse
A6 - High

And on chips U39 and U 40 the readings are as follows:

U39U40
1 nothing16 HIGH1 nothing16 HIGH
2 pulse15 nothing2 pulse15 LOW
3 nothing14 pulse3 pulse14 pulse
4 HIGH13 pulse4 pulse13 pulse
5 pulse12 HIGH5 pulse12 pulse
6 pulse11 pulse6 pulse11 pulse
7 HIGH10 pulse7 pulse10 pulse
8 LOW9 HIGH8 LOW9 pulse
 
Today, when I took the U14 out and put it back in, despite many attempts, I was unable to display the image.
But when I finally removed the Ruuds ROM and inserted Landmark Supersoft, the motherboard started without any problems and displayed the 16K error.
 
Post card shows: 00 FF
Which type of POST card: ISA slot, or parallel/LPT ?
( Modern ISA types are not expected to work in an IBM 5160, even for Rudd's Diagnostic ROM. )
ISA card.
I have yet to hear of a modern (repeat: modern) ISA POST card that works in an IBM 5160.
So even though Ruud's Diagnostic ROM outputs codes to I/O port 80 (what the ISA POST card monitors), your modern ISA POST card is not expected to show those codes.

Sometimes it shows 00 FF, another time 00 04, 00 F4, 00 89 but maybe it doesn't matter
Whatever those codes are, they are not codes output by Ruud's Diagnostic ROM. Consider them to be 'noise'.

You should now remove the ISA POST card. For diagnosing a motherboard, it is best to have as little hardware present as possible.

For the IBM 5150 and 5155 and 5160, the device shown at [here], attached to a parallel/LPT port, will display the codes output by Ruud's Diagnostic ROM.
 
So things have gotten worse.
The motherboard now behaves unpredictably.
That is a pity. It means that the motherboard is more difficult to repair.

When the motherboard does not display anything on the screen then there is no pulsation on pin 13 in the 8253 chip
All that informs you of, is that when you turned on the motherboard, either:
- Ruud's Diagnostic ROM did not execute at all; or
- Ruud's Diagnostic ROM did start to execute, but failed very early (before anything displayed on-screen).
 
Sometimes it displays an image on the screen and then shows a first 2 KB error.
....
Most often it does not turn on when I move the POST card to another ISA slot or when I remove and insert the U14 chip.
(The motherboard behaves exactly the same when I remove the U14 chip...)
Of course, unless you have 'repeatability', then an observation can be due to coincidence.

Sometimes my car starts and sometimes it does not.
Step 1: I swap out the car's TGF module for a known-good spare that I have.
Step 2: On the next attempt, the car started.
Step 3: I swapped back the original TGF module.
Step 4: On the next attempt, the car did not start.

Clearly, the above could be down to co-incidence.

Today, when I took the U14 out and put it back in, despite many attempts, I was unable to display the image.
And that could even be co-incidence. (Another chip failing at the same time as you removed/refitted U14 - it happens.)

When it works, the readings on the memory chips are as follows:
And on chips U39 and U 40 the readings are as follows:
Your motherboard now has different symptoms. Therefore, the diagnostic path is now different.

But when I finally removed the Ruuds ROM and inserted Landmark Supersoft, the motherboard started without any problems and displayed the 16K error.
Confirm for us that the motherboard is no longer intermittent; that is, with the Supersoft/Landmark ROM in place, the motherboard always starts.

( If that is the case, maybe the Rudd's Diagnostic ROM became intermittent. )
 
I have yet to hear of a modern (repeat: modern) ISA POST card that works in an IBM 5160.
So even though Ruud's Diagnostic ROM outputs codes to I/O port 80 (what the ISA POST card monitors), your modern ISA POST card is not expected to show those codes.
I used the POST card mainly to avoid turning the monitor on and off repeatedly.
I noticed that when the motherboard does not display an image, the card shows "bA" and when the motherboard starts and displays an image on the POST card, it will say bA 78 and then subsequent numbers up to 00.
Of course I understand that it is noise and I removed the card.
Confirm for us that the motherboard is no longer intermittent; that is, with the Supersoft/Landmark ROM in place, the motherboard always starts.

( If that is the case, maybe the Rudd's Diagnostic ROM became intermittent. )
After reading your answer, I had to check it right in the morning :)
I have swapped ROMs several times
I have used two different versions of Rudd's Diagnostic ROM 3.6 and 3.7 and neither of them works with my motherboard.
(V3.6 worked in the past and yesterday for a while)
Meanwhile, Supersoft/Landmark ROM boots every time.
 
I have used two different versions of Rudd's Diagnostic ROM 3.6 and 3.7 and neither of them works with my motherboard.
(V3.6 worked in the past and yesterday for a while)
Meanwhile, Supersoft/Landmark ROM boots every time.
A difference between Ruud's Diagnostic ROM (RDR) and the Supersoft/Landmark ROM (SLR) is:
- SLR brings up a display very quickly.
- RDR does some initial tests before it brings up the display.

Regarding RDR, those initial tests are the ones at [here] where the 'video enabled' column has 'No' in it. So if you had one of the parallel/LPT reader devices, it may have shown some checkpoints, and then stopped at one checkpoint.

By the way. I know that you have other faulty 51xx motherboards that you will possibly bring to our attention. It would be good if you purchased a parallel/LPT reader device in case you need to run one of the ROM tests at [here] or [here].
 
I have to have a think about why the Supersoft ROM is getting as far as the 16K RAM test, but Rudd's Diagnostic ROM is no longer displaying.
 
I have to have a think about why the Supersoft ROM is getting as far as the 16K RAM test, but Rudd's Diagnostic ROM is no longer displaying.
Something that would explain the situation is if there was a failure in the MDA/CGA video RAM.

Unlike the Supersoft/Landmark ROM, Rudd's Diagnostic ROM (RDR) uses the unused 96 bytes of MDA/CGA video RAM for variables and for the stack. For that reason, it is critical for RDR that those 96 bytes be good. Early, RDR tests those 96 bytes, and if any are faulty, RDR will send checkpoint 83h and then then halt (before anything is displayed).

Maybe a failure in those 96 bytes is what has happened.
 
I have released a new version of Ruud's Diagnostic ROM. It is now at version 3.8

Available at minuszerodegrees.net

Fixes:

1. 'Testing RAM - Address' test. Fixed the problem where a 4164 in a 41256 socket was not always detected.

2. On a particular IBM 5150 motherboard of mine, removing a Cassette BASIC ROM did not fail the applicable ROM test. Code that checks the ROM's was modified to cater for the fact that not all motherboard's read FFh at 'empty' addresses.

3. 'Testing RAM - Refresh' test. The delay between write and read-back was increased from 30 seconds to 90 seconds. It is due to an IBM 5160 motherboard of type 256-640KB that I own. Experimentation shows that with RAM refresh disabled, the RAM on this motherboard of mine holds its contents for between 60 and 70 seconds. 90 seconds allows for a buffer, and the fact that someone may have a motherboard that holds contents for longer.

Enhancements:

1. Cosmetic change: Test "Hot interrupts" renamed to "Hot IRQ interrupts".

2. Cosmetic change: Test "Checking interrupt 0" renamed to "Checking interrupt IRQ0".

3. Cosmetic change: Some tests that take more than a few seconds to execute, now show a count-down.

4. Cosmetic change: For some tests, the address of the RAM block under test is shown in the bottom right corner of the screen.

5. Improved documentation within the source code.
 
It would be good if you purchased a parallel/LPT reader device
I tested the motherboard with a different graphics card and it didn't change anything. Ruud's Diagnostic ROM does not display anything on the screen.
Today the parallel/LPT card arrived and I connected it to the port on the graphics card. RDR generates code 07 83
I programmed the T6077 ROM and launched it. Most answers are wrong: code 99 but sometimes code 11 appears.
There is no rule here: E.g. 82 usually returns 99 but sometimes 11.
 
It would be good if you purchased a parallel/LPT reader device ...
I tested the motherboard with a different graphics card and it didn't change anything. Ruud's Diagnostic ROM does not display anything on the screen.
Today the parallel/LPT card arrived and I connected it to the port on the graphics card. RDR generates code 07 83
Regarding [07][83].

The parallel/LPT reader device shows the second last code received on the left, and the last code received on the right. So, 07 followed by 83. If we look at the list at [here], checkpoint 07 is sent at step 16, and checkpoint 83 is sent at step 17.

So, either:
- The graphics card is neither MDA nor CGA; or
- The graphics card is MDA or CGA, but is faulty; or
- The graphics card is MDA or CGA, and is functional, but for some reason, RDR cannot successfully write-then-read-back test data to the card's video RAM.

Related to this, I was recently testing a new version of RDR. I swapped the 8088 CPU for a V20. Sometimes RDR would start and sometimes not, and when it did start, sometimes strange characters would appear on the screen. When it did not start, the parallel/LPT reader device revealed a failure to successfully write-then-read-back test data to the card's video RAM. The particular MDA card I was using became unstable when a V20 was used. I had to continue the V20 tests using a different MDA card.

I have looked at the source code for the Supersoft/Landmark diagnostic, but I understand that it assumes that an MDA/CGA card is present, and outputs to it.
 
I programmed the T6077 ROM and launched it. Most answers are wrong: code 99 but sometimes code 11 appears.
There is no rule here: E.g. 82 usually returns 99 but sometimes 11.
TEST6077: "A crude ROM addressing test. Tests 128 addresses, spanned between the motherboard addresses of F8000 and FFFFF, the 32 KB sized address range that the first BIOS ROM (socket U18) of the IBM 5160 occupies."

This failure will result in the IBM BIOS ROM not executing.

Presumably, RDR (and the Supersoft/Landmark diagnostic) are executing because, to execute, RDR requires only that a part of that address range be functional, per the diagram at [here], and the fault affects other address ranges. When TEST6077 was running, the FE and FF subtests were probably always passing (11).

There is no rule here: E.g. 82 usually returns 99 but sometimes 11.
Unstable. That will make tracking down the problem much harder.
 
Getting back to this:

- The problem appears to be getting worse. E.g. Ruud's Diagnostic ROM (RDR) was previously displaying something, but no longer.

- The motherboard may have multiple problems.

- The SuperSoft/Landmark Diagnostic ROM (SLDR) always starts and displays something.

- Presently, RDR is not displaying anything, but the parallel/LPT reader device indicates that RDR is starting, but having a problem when testing MDA/CGA video RAM. This may be due to the faulty motherboard.

- The fact that SLDR and RDR are executing suggests that the motherboard problem is not data related.

- Earlier, the hypothesis was a problem on the motherboard's RAM data bus, but failure of TEST6077 suggests a wider addressing issue. Putting that together with the fact that SLDR and RDR are executing, suggests the possibility of an address bit/s between A13 to A19 being stuck high.

- The motherboard is unstable (e.g. the 82 subtest [address F8200] of TEST6077 sometimes fails and sometimes passes).

- It is unknown if the 8088 CPU (socketed) has been swapped out for another. (I.e. Quickly eliminate the CPU.)


Refer to the diagram at [here]. I suggest replacing:
- 8088 CPU
- Chip U7
- Chip U14
 
- The SuperSoft/Landmark Diagnostic ROM (SLDR) always starts and displays something.
My original version did the same. But then I decided to initialize only that board of which I found the memory. If I may suggest: maybe It is better to initialize both boards and to display anything in the first one-time part and to continue in the loop with that board whose memory you found.
 
My original version did the same. But then I decided to initialize only that board of which I found the memory. If I may suggest: maybe It is better to initialize both boards and to display anything in the first one-time part and to continue in the loop with that board whose memory you found.
It sounds like the way to go. I will put some thought into it.

( I have only looked at parts of the source code for the SuperSoft/Landmark ROM. I wonder what it is doing video wise. And for that matter, where it stores variables, and when does it start using the variables. )
 
Back
Top