cloppy007 wrote on 2024-06-09, 20:19:

You're right. I've got 2 similar PCI/ISA post cards and one of them won't work in 5V-only PCI slots (nor ISA...). Have you... […]
Show full quote

You're right. I've got 2 similar PCI/ISA post cards and one of them won't work in 5V-only PCI slots (nor ISA...). Have you...

• Measured the VRM output? If your board has the VRM slot it will surely have 2-4 jumpers, you can get the CPU voltage from any of those. Otherwise it'll be in one of the leads of the transistor with a heatsink just next to the battery
• Verified the BIOS chip? If you have access to a programmer, try reading the BIOS chip to see if it reads something or garbage.
• Cleaned the cpu and bios sockets with IPA or contact cleaner?

I had an extra bios chip ready to go. Switching it did nothing. I cleaned the CPU socket with non-residue contact cleaner and blew it out with compressed air. I also switched two other CPUs in its place and none of it worked. Just now I checked the solder joints on the back of the motherboard where the transistor is. I don't know which pins are which, but it was giving me 1.5V to 1.6V.

I've got a new motherboard on the way. I'm not sure what else I can look at to find the issue.

Syntho wrote on 2024-06-10, 13:44:

I had an extra bios chip ready to go. Switching it did nothing. I cleaned the CPU socket with non-residue contact cleaner and blew it out with compressed air. I also switched two other CPUs in its place and none of it worked. Just now I checked the solder joints on the back of the motherboard where the transistor is. I don't know which pins are which, but it was giving me 1.5V to 1.6V.

Well... that's way too low, you're probably reading the wrong pin. I seriously recommend you to measure here:

Red multimeter probe on any of those bridged pins and black probe on any ground such as the metal surrounding the AT mounting holes (or a black-coloured wire of the AT connector).

cloppy007 wrote on 2024-06-11, 20:29:

Well... that's way too low, you're probably reading the wrong pin. I seriously recommend you to measure here:
pentium_vrm_header.png
Red multimeter probe on any of those bridged pins and black probe on any ground such as the metal surrounding the AT mounting holes (or a black-coloured wire of the AT connector).

The picture is very small, but from what you post I take it I can put my red probe on any of the 4 jumpered pins, and my black one on any ground point? I'm getting 3.3V in that area.

If I've got 3.3V there, I wonder why I'm not getting 3.3V on my POST card.

I would remove all components including video card except RAM and CPU. I would make sure the PC speaker is working (I know this sounds stupid). If still nothing I would think the motherboard or power supply was bad. If power supply is definitely good maybe bad capacitors on motherboard are bad or something more serious.

Syntho wrote on 2024-06-11, 20:38:

cloppy007 wrote on 2024-06-11, 20:29:

Well... that's way too low, you're probably reading the wrong pin. I seriously recommend you to measure here:
pentium_vrm_header.png
Red multimeter probe on any of those bridged pins and black probe on any ground such as the metal surrounding the AT mounting holes (or a black-coloured wire of the AT connector).

The picture is very small, but from what you post I take it I can put my red probe on any of the 4 jumpered pins, and my black one on any ground point? I'm getting 3.3V in that area.

If I've got 3.3V there, I wonder why I'm not getting 3.3V on my POST card.

Then it's all fine regarding CPU power. Those 3.3V only go to the CPU, and depending on the CPU it might be lower (for instance: 2.8V for the MMXs, 2.2 for some K6-2s).

You're not supposed to get 3.3V in your PCI slots. Mind that they're keyed to 5V and the PCI spec does not require 3.3V to be present for those 5V slots. I've got an Azza/Bona/Lex/Jetway P2/P3 motherboard that also does not supply 3.3V to the PCI slots. It drove me crazy for 1 months until I found out it was OK.

I'd try to follow the boot sequence but it can be time consuming. That involves scoping the address and data pins of the BIOS to see if the CPU is trying to read the BIOS. If all are low I'd go instead for the SuperIO (that SMC chip) and find its datasheet. Pin 58 should be POWERGOOD. It should be high (5V).

By the way, does your POST card have leds for RESET, CLOCK, FRAME or IRDY? You should at least have reset. It should be lit on for a split second and then turn off (pulled down). If it stays high, something is preventing the CPU from starting. But then again, if the POST card relies on 3.3V to be present, I'm not sure if it'll reliable work. I think mine did.

And there's always the possibility of a cracked solder joint or broken leg or broken trace. Or perhaps the first cpu with bent pins messed up the socket contacts (those can be fixed). I've repair my share of motherboards but I'm still a newbie, I'm sorry I'm not bringing a definitive solution.

By the way, do you have a picture of the board?

The card that I bought is here: https://www.amazon.com/dp/B072QSQQKT?psc=1&re … product_details

I don't have a picture right this second but you can see if here: https://theretroweb.com/motherboards/s/azza-5is

One thing that the manufacturer told me is that I should be getting the Reset LED coming on at least for a couple little flashes when starting it up. I'm not getting that at all. When I hit the Reset button it won't flash quickly either. I'm wondering if somehow this thing's Reset circuitry is messed up. I don't know how to troubleshoot it.

Not sure this will help but here's the actual mobo:

I replaced a few more caps all near the CPU just in case. They were ok it seems, except for that one.

If I had to guess, something is wrong with the CPU slot. Last time it did this, I had to reseat the CPU. It was a last ditch thing since it wouldn't POST just like now. When I pulled the CPU, I discovered at least one broken pin (I think two?). It worked fine for months after that since the broken pins were obviously non-essential.

Fast forward to a few days ago, the thing wouldn't POST once again. After testing a few things, I decided to check on the CPU as that's what the culprit was last time. I pulled it up and I noticed probably 3 MORE broken pins, including one that was laying flat and possibly causing a crosstalk between a couple other pins. I retired that CPU, blew out the holes with contact cleaner and compressed air, and tried two different CPUs (including a NOS P133), and it still won't POST.

My CPUs are probably fine, I'm thinking something's going on with that damned CPU slot after that first CPU kept breaking down slowly but surely. I just don't know what to check for right now. There's doesn't seem to be a broken pin in the CPU slot as far as I can tell.

Hey you know you need 2 simms to boot, right? Pentiums have a 64-bit wide datapath. Each SIMM is 32-bit wide only, so unlike 386 and 486s you need an even number of SIMMs to boot. SDRAM DIMMs are 64-bit wide so you only need one.

I took them out earlier to solder something, then put one stick back in before taking a pic. I've got two sticks in there, always.

Well, that'd have been too easy. Do you think the bent pins might have shorted something? I don't think that'd kill the CPU but the chipset or anything in between is another story. The board's in a pretty good condition as far as I can tell.

As for the reset line, it should be like in this video (bottom-right led). Off->On->off.

There might be a short somewhere in the board but it might be tricky to find (if it's even there). A thermal camera would surely help. Or having a similar board and compare the power draw with no cpu inserted (if you're lucky you might tell a significant difference). But at this point if I were you I'd check the data/address legs of the BIOS chip. If nothing comes up, I would lift the socket top and inspect the contacts, but that might damage the socket itself as the plastic is probably now very brittle.

This is what I get:

There is no little blink for the Reset unfortunately. And as mentioned the 3.3V LED doesn't come on.

You can only slightly tell, but the CLK LED seems to be only half lit. The video you posted seems to indicate that the right row of LEDs going down don't light up as much, so I think it's probably ok. And as you can tell, other than 3.3V, IRDY, FRAME and RESET don't come on or flash at all. When you start it up, the LEDs are lit as you see and never change.

I've never fooled with testing a BIOS chip or socket, only copying and programming E/EEPROMS. I have two BIOS chips that were verified good a few months back and switching those for one another won't help, so I imagine it's a hardware issue at this point.