• Please review our updated Terms and Rules here

IBM 5150 - No POST

Zh1nu

New Member
Joined
Dec 13, 2021
Messages
9
Hey y'all,
I got an IBM PC a while back that doesn't POST. I recently got a probe to check, and it seems like the RAS and CAS don't strobe at all.
The chips behind it on the schematics seem to be fine, with the RAS and CAS S138s receiving strobing on 2 of the 3 input lines, and strobing on the control lines. I doubt that two S138s would fail, and the control chip for the RAS failing wouldn't explain why the CAS doesn't strobe either.
I'd buy diagnostic ROMs but the only person who's selling them wants around 50 bucks shipped. If anyone has a source for those Landmark ROMs that isn't too expensive please send it to me, I can't find any.

What would be the best way to fix this?

Thanks,
Zhinü

View attachment 1238181
 
I got an IBM PC ...
Lots of IBM PC (IBM 5150) information at [here].

I got an IBM PC a while back that doesn't POST. I recently got a probe to check, and it seems like the RAS and CAS don't strobe at all.
The chips behind it on the schematics seem to be fine, with the RAS and CAS S138s receiving strobing on 2 of the 3 input lines, and strobing on the control lines. I doubt that two S138s would fail, and the control chip for the RAS failing wouldn't explain why the CAS doesn't strobe either.
Yes, a logic probe has limitations.
See the 'Note that because a logic probe does not convey timing information ...' paragraph of [here].

I'd buy diagnostic ROMs but the only person who's selling them wants around 50 bucks shipped. If anyone has a source for those Landmark ROMs that isn't too expensive please send it to me, I can't find any.
Whereabouts on the planet are you located?

There are members and companies at [here] that can assist.
For your IBM 5150, per the 'Use of an EPROM to substitute' section of [here], there are various EPROM/EEPROM types that can be used, some requiring an adapter.

What would be the best way to fix this?
A very common problem on the 5150 motherboard is short-circuit tantalum capacitors, but we know that that is not the problem with your motherboard, because you are seeing logic activity using a logic probe.

Also, logic activity suggests that the motherboard is getting a POWER GOOD signal from the power supply, and is starting.

1. Start with some basics:
* Do a very good visual inspection of the motherboard.
* Re-seat all socketed chips, being very careful not to bend a pin up under the chip when reinserting the chip.

2. If you perform the procedure at [here], do you hear the expected beeps ((1 long, then 2 short) ?

3. Refer to [here], which corresponds to step 9 at [here]. Using your logic probe, do you see constant activity on pin 13 of the 8253 chip ?
 
I'm located in the US, namely California.

I need to find someone who'll do the two chips that don't require adapters because someone is also scalping the adapters and I can't find Gerber files to get them made at PCBway or some other company.

Answering the questions:
1. The board seems fine, I'll take some photos to see if anything seems amiss. I've reseated most of the larger chips, but not the RAM as I've disabled the socketed banks to work with the base 32k.

2. I do not hear any beeps with the config.

3. I see a 5V line with no strobing on pin 13 of the 8253 chip. At least, I assume it is, as the chip's label has been wiped off. I've cross-checked with the schematics to find it.

As a side note does anyone know how to delete forum posts? The other one I sent was an editing accident.

IMG_6193.jpgIMG_6194.jpg
IMG_6196.jpg
 
A bit of an update. I've checked the other parts of the timer chip. Write, Read, and /CS do not strobe (they're set to high), and the data lines do strobe for about a second before getting set to 5V. A1 and A2 do strobe. I'm wondering if the 8086 is bad.
 
So, you have the 64KB-256KB version of the motherboard, fitted with the 10/27/82 dated BIOS ROM.

Note that if the shown power supply is a type that does not provide -5V, that will not stop a 64KB-256KB motherboard from starting and displaying something on a monitor.
(For the 64KB-256KB motherboard, a lack of -5V is one of the causes of a 131 error on-screen.)

... but not the RAM as I've disabled the socketed banks to work with the base 32k.
On 5150 motherboards (of both versions), all four banks are permanently enabled. The RAM related switches only inform the BIOS as to how much RAM is fitted. But you could removed the socketed RAM chips in case of some form of interference.

Bank 0 is 64 KB sized, not 32 KB.

I see a 5V line with no strobing on pin 13 of the 8253 chip.
So either the POST is not starting, or is starting but stopping earlier than the step that sets up the pulses on pin 13.

[/QUOTE]
I'm wondering if the 8086 is bad.
There are lots of possibilities. The 8088 CPU is quite reliable. You are more likely to have a faulty BIOS ROM than a faulty 8088.

The motherboard start-up sequence is described at [here].
Using your logic probe you could look for:
- POWER GOOD from the power supply being asserted
- RESET no longer being asserted
- 4.7727 MHz clock

Beyond that, something else you can do is described in the first paragraph of post #19 of [here].
Doing that, your logic probe should show:
- CPU status of 'Code access' (8088 pin 28 {S2} being high, 8088 pin 27 {S1} being low, 8088 pin 26 {S0} being low) - see [here]
- ADDRESS BUS ON AN ISA SLOT: A19 to A0 = 1111 1111 1111 1111 0000 = FFFF0 hex
- U33 BIOS ROM: Pin 20 (/CS) being low - see [here]
- U33 BIOS ROM: Address pins A12 to A0 = 1 1111 1111 0000 = 1FF0 hex
- U33 BIOS ROM: Data pins D7 to D0 = 1110 1010 = EA hex
- DATA BUS ON AN ISA SLOT: D7 to D0 = 1110 1010 = EA hex
 
- POWER GOOD is asserted
- Reset is set low after less than a second
- There is a clock on the 8088's pin

- Pin 28 is set high, however both 27 and 26 are strobing
- On startup, CPU pins from A0 to A19 are 0101 0101 ???? ???? ???? where ? is instant strobe upon power-up.
- On startup, U33 goes from high, then strobing, then back high again.
I did not check further than that.
 
The BIOS configures channel 1 (DMA refresh, output on pin 13) of the 8253 timer relatively early - the only tests that happen before initializing the timer are the CPU test (flags and registers) and the ROM checksum. If either of those tests fail, the BIOS simply halts the CPU and the system will appear completely dead. I agree with modem7 that the likely culprit is a bad ROM, although it could also be a bad address or data buffer between the CPU and the ROM. It's unlikely that the CPU itself is the problem otherwise you wouldn't be seeing the initial burst of activity on the address/data lines.

Actually reviewing the BIOS listing reveals that there are a couple other I/O operations that happen prior to the ROM checksum and 8253 init:
- write 0x00 to port 0xA0 (disable NMI)
- write 0x00 to port 0x83 (DMA page register - the 74LS670 chip)
- write 0x00 to ports 0x3D8 and 0x3B8 (disable mono/color video)
- configure the 8255 PPI chip and write 0x0F to the PPI's Port B

So you could see if there are any low-going pulses on the outputs of the LS138 at U66 - Y4 is the chip select for the DMA page register and Y5 is the chip select for the NMI register. If both of those outputs pulse once during POST but the 8253 timer isn't set up, then process of elimination would determine that the ROM checksum failure is the reason the system is halting.
 
Checking U66, neither Y4 nor Y5 put out a pulse. They both shoot up to 5V.
 
I’ve bought some diagnostic roms. I’ll be back in a week or two to see what they say.
 
Back
Top