IBM 5150(A) motherboard in trouble

[predator99]

Congratulations.
I have added the failure to the list at [here].

Hi, thank you very much! Thats very interesting!

I also have a 5150 mainboard I am trying to repair since a longer time but I am totally out of ideas.

- fixed the crystal, now there is a clock signal on the ISA-bus and also on the CPU
- still no beep on power up
- POST card shows that signals and voltages are OK
- tested the BIOS with an adapter in my TLL866 - OK
- made a Supersoft-ROM with an adapter, verified pinout in my TL866 with above adaptor - no beep and no output on CGA
- tested almost all ICs with a piggyback - no success

Maybe interesting: Sometimes - especially when I touch the 8253 - there is some noise coming out of the speaker. But I dont know whats the reason and where I shall continue to look.

Any ideas?

[fs5500]

Very great.

If you can show the photo of IBM PC 5150 Rev.1 BIOS (04/24/81) and motherboard on your site, thanks.

I have most of early IBM PC 5150 motherboards (Except with Rev.1 BIOS), IBM PC 5160 (Early Rev.0(or 1) BIOS dated with 08/16/82, IBM PC 5170 (Early type with 4 bios chips)

[romanon]

It will be soon (I hope.) I also wait for CGA card repair. Then I introduce whole early 5150(A) PC

[modem7]

If you can show the photo of IBM PC 5150 Rev.1 BIOS (04/24/81) and motherboard on your site, thanks.

It is unclear as to whom you are asking the question.
If it is me:
A photo of the 04/24/81 BIOS is at [here].
A photo of the 16KB-64KB version of the 5150 motherboard is at [here].

If you are seeking a photo of the two combined, then I suggest that you look at the photos of Framer's early 5150 at [here].

[modem7]

- fixed the crystal, now there is a clock signal on the ISA-bus and also on the CPU

Good.

- POST card shows that signals and voltages are OK

The 5150 motherboard's startup sequence is at [here].

So, one check is that the POWER GOOD and RESET signals are doing what is shown at [here]. And ideally, verify that the transitioning RESET is making it from the 8284 to the CPU.

Note that simply looking for a high POWER GOOD and low RESET is a confidence thing only. To completely rule out a problem in that area, you need to see the transition.

- tested almost all ICs with a piggyback - no success

That is an unreliable testing method.

Any ideas?

Also, try swapping the CPU with a known working one.

[predator99]

Hi, many thanks for your reply! I already tested different (known working) 8088s, but no change. I also tried to replace the 8284 as it is in a socket. I just measured voltages/frequencies, but I dont own an oscilloscope:

For 8284:
Pin 9 (GND) and Pin 10 (RESET): 0.15 V
PIN 9 (GND) and Pin 18 (VCC): 4.999 V
Pin 8 (CLK): 4.773 MHz

For 8088:
GND (Pin 1 or Pin 20) and VCC (Pin 40): 4.995 V
GND (Pin 20) and Reset (Pin 21): 0.143 V
CLK (Pin 19): 4,773 MHz

For the Reset on the 8284, you can see voltage going up for a very short interval after powering up, but as said I am not able to determine the excact time.

[modem7]

For the Reset on the 8284, you can see voltage going up for a very short interval after powering up, but as said I am not able to determine the excact time.

I expect that the high time will vary from power supply to power supply, and be affected by the amount of loading on the power supply. So high time is not important (unless it is very low). The fact that you saw it "going up for a very short interval after powering up" provides some confidence.

At [here] is the sequence of the POST in the 10/27/82 dated IBM BIOS ROM. The failure of early tests simply halts the CPU. After step 9, there will be a pulse periodically coming out of channel #1 of 8253 timer chip. Because you appear to have the ability to measure frequency of TTL signals, see if you get a reading. You are expecting about 66 kHz. I am not confident that you will get that reading, because there was no output of the Supersoft diagnostic ROM.

[predator99]

Hi, many thanks for your help, very appreciated!!

I did some mesaurements on the 8253:

GND (Pin 12) and VCC (Pin 24): 4.995 V
CLK 0 (Pin 9): 1.193 MHz
CLK 1 (Pin 15): 1.193 MHz
CLK 2 (Pin 1: 1.193 MHz
OUT 0 (Pin 10): 79-90 Hz
OUT 1 (Pin 13): 117-119.3 Hz, next measurement 40-50 Hz
OUT 2 (Pin 17): 117-119.3 Hz
GATE 0 (Pin 11): 50.00 Hz
GATE 1 (Pin 14): 50.00 Hz
GATE 2 (Pin 16): 50.00 Hz (noise on speaker changing during measurement...)

...and noise coming out of the speaker (turning on and of, especially when touching the 8253), measured frequency: 50-150Hz

Most frequencies are little different when repeating the measurement, only the CLKs are very constant...

Does not seem to be OK, right..? But is the problem the 8253 itself or a supporting IC?

[predator99]

Had some time, desoldered the 8253 and put a new one on a socket. No difference, but at least the noise from the speaker is gone.

[modem7]

Most frequencies are little different when repeating the measurement, only the CLKs are very constant...

Maybe for the former, your instrument is amplifying then measuring noise.

Per my earlier post, I was not expecting channel 1 output from the 8253 because of your Supersoft ROM result. But had you an 8253 channel 1 result, your Supersoft ROM would have been under suspicion.

To progress, I think you need more test equipment (e.g. see [here]), but that introduces issues. You may outlay decent money for no repair result. I expect that the potential helpers on these forums will expect that at the least, you know how to operate such equipment. And helpers like me are prepared to offer only so much of our spare time. It is very time consuming to do deep-level diagnosis remotely (and slowly), compounded by bad measurement from things like pin confusion, probes not going through oxide build-up, misunderstanding.