• Please review our updated Terms and Rules here

AST Advantage 613e (similar to 610/611) POST codes / dead board

cwathen

Experienced Member
Joined
Feb 1, 2018
Messages
268
Location
Bristol, UK
Having waited the better part of 25 years to get another AST Advantage system after my original suddenly died with a motherboard fault, the 613e which I've owned for just under a month has done the same.

When I first got it I tried to upgrade the 8MB of RAM (2x4MB 72 pin EDO SIMMs) to 16MB. I chucked in 2x 8MB that I had spare but the system wouldn't POST with the new RAM fitted. Not even any beeps. Nothing. Swapped back to the original and left things as they were until today...

...when I got out my entire collection of 72 pin SIMMs to try and find any that would work in it. Having verified that it would run with only a single bank fitted (it's a Cyrix 5x86 so 486 architecture) I went through the lot. Only 4MB modules seemed to work in it. I thought this was odd since the published specs state it will support up to 32MB RAM and there are only two slots (and no jumpers to configure the amount of RAM fitted). Decided to put it down to experience and re-fit the original RAM.

Only now the board is dead. Well, almost dead. It is beeping POST codes at me so the BIOS clearly is still running but nothing more. It is sensitive to what is fitted:

No RAM - as soon as power is applied 3 short, 5 long, a pause and 4 short
Original RAM/Other 4MB modules - a pause at power up followed by 2 long and 3 short
Modules bigger than 4MB - nothing (as before)

I have verified that the original RAM is good in another system, the problem is on the board.

I've googled various AST Bios revisions for beep codes but none seem to match this pattern. I would imagine since the BIOS still seems to be up then it is only just falling over and likely not much is wrong. But no idea where to start especially without a definite match for the beep codes are.

Anyone have any experience with these boards?
 
My first PC was an AST Advantage 611S and I still have it but I haven't powered it up in many years due to lack of space to set it up. I'm not sure if I ever had a manual (describing any beep codes) but I couldn't find any when searching for it now (I feel like there should have been a manual included with the computer but I guess it's been lost somehow if so).

So I can't help with the beep codes. Sorry.

Yes, the max amount of RAM is 32 MB - I have 2x16 MB in mine.

Not sure what could be wrong with your motherboard. Bad caps? They are pretty old now. ESD damage? Cracked the motherboard or somehow damaged/shorted the SIMM slots when inserting the RAM? Unlikely, I suppose.

BTW, have you tried clearing CMOS? Could be worth trying.

I hope you get it fixed and I REALLY hope these AST computers aren't time bombs. :(
 
Removing the CMOS battery in case it got 'confused' by the changing RAM fitted was the first thing I tried (it should still POST up to to an error message with no battery fitted).

The board is covered with identical SMD aluminium can electrolytics. I would have thought it's a little bit too old for caps plague but possible they're just warn out I suppose.

As far as I can tell the 610/611/613 are all just minor variants of the same basic machine so if you did have any documentation it probably would be the same.
 
Well this is eerie. I having a very similar issue that cwathen is with an AST Advantage 613e as of yesterday.

Just for clarity this is my testing and situation ::
No RAM - as soon as power is applied 3 short, 5 long, a pause and 4 short
Original RAM (2x 16Mb) and certain other modules - a pause at power up followed by 2 long and 3 short
Certain modules - nothing

I found no pattern in my testing regarding module sizes and whether or not it decides to do the 2 long 3 short beep code or not.

Original modules from my AST 613e and modules used for testing have also been verified as working on another 486 machine with memtest for a couple of hours each just to be certain.

Power supply has been checked, all correct voltages and tolerances. Tested with a spare PSU just to be certain, exactly the same results.

BIOS battery removed to clear CMOS. Removed again overnight, exactly the same results.

I don't have any original documentation, nor can I find much online. The only useful resource I've managed to find is a basic motherboard layout ::
https://stason.org/TULARC/pc/motherboards/A/AST-RESEARCH-INC-486-ADVANTAGE-610-611-486-202728.html


Using a ISA POST 4 bit card :: With RAM Removed - Code 54
Using a ISA POST 4 bit card :: With Original RAM (1x16mb in bank 0) - Code 23
Using a ISA POST 4 bit card :: With Original RAM (2x16mb in bank 0) - Code 23
Using a ISA POST 4 bit card :: With Original RAM (1x16mb in bank 1) - Code 54 (I know this wasn't necessary, just going over the top with testing)

I seem to remember that the AST Bios is based on an AMI bios, although some clarity here would be useful to exactly which bios it's based off.

If it is an Ami Bios then the following codes mean the following ::

23 : Base 64K memory initialized. Going to set BIOS stack and to do any setup before interrupt vector initialization.
54 : Shutdown successful, CPU in real mode. Going to disable gate A20 line and disable parity/NMI

What does interest me is that cwathen reports nothing happens with modules greater that 4mb : I'm no expert but I do know that with 30pin SIMMS the system requires and uses and extra address pin (line) when moving from 256k to 1mb, 1mb to 4mb, 4mb to 16mb SIMM modules. I wonder if the same situation might be happening here but perhaps the corresponding pin that becomes required when going above the 4mb size is damaged and we get the strange behaviour?

The most frustrating thing in my case was it was working just fine with 2x 16mb (32mb total), I literally only opened the system up to test a few 8mb modules i knew worked but wanted to be certain as i was selling them to a friend.

In both our cases we get pretty much the EXACT same problem after replacing memory? I'm sure thats the area to focus on, physically as a result. My current gut feeling is that it will be the 72pin socket or solder joints at this point. I don't have time this weekend but I'll do some testing hopefully next week and look into the option of replacing the sockets or reflowing the solder joints if this looks to be the problem and report back. I'm really wanting to get to the bottom of the issue regardless, simply out of curiosity. Any other information or suggestions would be appreciated from other people!

Cheers,

DK
 
I seem to remember that the AST Bios is based on an AMI bios, although some clarity here would be useful to exactly which bios it's based off.
The original model I had was from 1996, that was definitely a rebranded AMI BIOS, apart from the AST branding it was identical.

This 613e is from 1995 and the BIOS is slightly different, but pretty sure it's stll AMI.

What does interest me is that cwathen reports nothing happens with modules greater that 4mb : I'm no expert but I do know that with 30pin SIMMS the system requires and uses and extra address pin (line) when moving from 256k to 1mb, 1mb to 4mb, 4mb to 16mb SIMM modules. I wonder if the same situation might be happening here but perhaps the corresponding pin that becomes required when going above the 4mb size is damaged and we get the strange behaviour?
That is a good point. I kindof knew something was wrong when the original attempt to go to 16MB failed since I'm pretty sure AST offered that as a stock machine (they were masters of selling the same computer under 5 or 6 different model numbers where the only differences were changes to the RAM and CPU fitted) and the board is supposed to take 32MB anyway.

I've never given much thought to how 72 pin SIMMS work. Possibly there is an extra pin required for >4MB modules not connecting which is why the upgrade failed? And now making several changes has broken the connection elsewhere?

I did check the bank 0 connector for continuity through to the other side of the board and they did check out, but then could still be a dry joint as I would just be touching the pin on the other side of the socket along with the solder which would bridge any break.

In both our cases we get pretty much the EXACT same problem after replacing memory? I'm sure thats the area to focus on, physically as a result. My current gut feeling is that it will be the 72pin socket or solder joints at this point. I don't have time this weekend but I'll do some testing hopefully next week and look into the option of replacing the sockets or reflowing the solder joints if this looks to be the problem and report back. I'm really wanting to get to the bottom of the issue regardless, simply out of curiosity. Any other information or suggestions would be appreciated from other people!
I concur, I know it's hard to properly test caps in circuit with just a multimeter but they do seem to be charging and the timing of it being fine until messing with the RAM makes me think broken/dry joints. Clearly the CPU and BIOS are working, it's just the RAM failing to work properly. Hopefully it's old enough to be pre-RoHS - I do have a cheap rework station but it doesn't heat up to the claimed temperature and struggles with leadfree. Will try a little reflow in that area. If not could just shove the whole board in the oven I suppose but the only time I've ever tried that before I ended up with melted connectors and still a dead board.
 
Last edited:
Agreed,

I'll inspect my board at some point this week when I have the time.

I've got some high end Weller equipment and hot air rework station should it be required. Regardless of continuity it could also be the 72 pin connector itself not making a good enough connection or providing enough pressure to do so.

I have seen this happen, although it was visibly obvious too, in old 30 pin SIMM systems. When youy think about it, you are relying on the metal installed in these connectors to "spring" against the module installed. If they have been in this sprung state (installed) for 20-30 years and you finally come to remove a module you sometimes can visibly see the connector is fixed in it's sprung state and doesn't work as intended anymore. When removing SIMMS in cases like this I have noticed they almost feel a bit "stuck" to the connectors pins. I can't remember the technical or engineers answer behind this but i know if given enough time in close proximity (10+years) some metals will swap electronics with each other and essential "fuse" together a bit.

I do remember that when first removing bank 0 it was a little bit tricky. When inserting other modules and reinserting the originals it was much easier so i'm really starting to think it's the connector? If you happen to remember would you say the RAM was a bit toucher to remove the first time in your 613e experience?

I've gone ahead and ordered a couple of 72-way connectors from RS and after some more thorough testing and examination I think I might be replacing the sockets, gut feeling.

Link below for easy reference ::

https://uk.rs-online.com/web/c/conn...memory-sockets/?applied-dimensions=4294868388
 
Agreed,
If they have been in this sprung state (installed) for 20-30 years and you finally come to remove a module you sometimes can visibly see the connector is fixed in it's sprung state and doesn't work as intended anymore.]
I think you're onto something there, my board visibly doesn't have the pins all retracted with the modules out of the board.
 
Back
Top