• Please review our updated Terms and Rules here

Help! My Intel Inboard 386/PC is dead!

Mike1978

Experienced Member
Joined
Sep 6, 2020
Messages
120
Location
Sheffield UK
I’ve recently acquired an Intel Inboard 386/PC it’s been running great, I made a 4mb daughterboard thanks to others in this forum.

I swapped out the oscillator from 32MHz to 40Mhz and adding a socket in the process.

I put a Cyrix 486 cx486drx2 20/40 CPU on the board and a DLC 487 FPU.

It’s been running in my 5150 for the best part of a month and for a good few weeks with the new Oscillator.

I’ve been looking to upgrade my XT IDE and on running xtidecfg I hit an issue where the flash failed saying it wasn’t right. I took the cable from the inboard to access the Xtide to remove the label and check the rom type, did so and put everything back then black screen.

Noting I’d removed and unplugged the cable many times and the card as a whole also many times without issue. I know these things aren’t meant to be moved about a lot but getting the right combination and experimenting to get it right led to some tinkering and moving about. That and the limitation of 5 isa slots in the 5150.

The crux of my issue is now the inboard will no longer boot.

Replaced the original cpu and the 5150 board starts up, but put the inboard into either 5150 and a whole lot of nothing, no beeps just the psu powering up.

As there are no moving parts and only the parasitic cable to the CPU slot I figured I’d check the cable for potential faults.

I beeped out each hole in the idc connector to the dip pins that go to the cpu socket. 40 holes / pins and each matched up accordingly.

I’m at a loss as to what may have gone bad or if there’s any hope of ressurection, it’s a rare card and one I’d hope for many years of service to come.

So putting a plea out for anyone who knows how to help or what to try next.

I have swapped back the original oscillator and put a standard 16mhz Intel cpu back in the card and removed the fpu all to no avail tried with and without the daughterboard.

I have very little in terms of diagnosis beyond a standard multi-meter and probably not many more skills besides other than relying on the goodwill of others to throw me a life line.

Please help!

Mike
 
The coincidence of the problem occurring at the time of IDC cable removal/replacement, really does suggest a cable or IC socket related issue.

It has been many years since I have had something like an Inboard 386/PC in my hands, but I recall that the pins on the cable connectors are typically thinner and shorter than the pins seen on an IC. On my IBM 5150/5160 motherboards, I have had to replace the CPU socket on one motherboard, and I presently have an early 5150 motherboard that I am soon to replace the CPU socket on (when I get tired enough of having to re-seat the CPU). Might your problem be due to a deteriorated CPU socket (having some issue with the thinner/shorter pins of the IDC cable) ?
 
The cable buzzes out so I think for sure there is contact and I hadn’t removed the pin side to the dip socket when I first lost the board it was the 40pin idc side to the board. Also each and every other cpu I place in the cpu socket does seem to work fine. The dip side does appear to make good contact and I have to use a cpu lever to remove the dip side of the cable it appears secure at both ends.

But thank you for the suggestion. Perhaps I could try something to sandwich it in tighter but then it’s the same on two separate 5150 boards :(
 
Ah. I did not pick up on your "either 5150".
Sorry unable to edit my initial post now but I have two 5150’s and tried in both.

Thanks all the same, I guess then it would be over to someone with a scope or diagnostic skills / tools that I don’t possess. However I should take this opportunity to thank you for your webpage it’s been a massively useful resource throughout my ownership of my PC / XT class machines. Perhaps I should put a plea out to anyone with such tools / skills who would be open to looking over the card. I’m UK based but the postal system doesn’t restrict me to that. I wonder if the likes of someone on here who may reverse engineer such cards would be interested to take a look?
 
only got a few comments.
One i had some issues with a CPU not plugged 100% into the socket. When pussing the CPU way down , it started again..
And the fingers on the ISA part of the Inboard386 , mine gets "black" in no time. Ones in a while i pollish the fingers with alcohol and a cotten-bud.
/cimonvg
 
Thanks Cimon, I swapped out the CPU for the original in case the CPU went bad and likewise I checked the isa gold connectors but good suggestions thank you.
 
I know nothing about this board. But since your system seems to work fine when this is out and the original CPU is in, it seems it must be something with this board. Are they any tantalum caps that might have gone bad and shorted? I know this is suggested for many issues just to be sure.
 
So far as I can see no caps have gone bad it looks in good physical shape in all honesty. But it’s a reasonable call given the age of the equipment. Normally when a tantalum goes it really goes :)

Feeling a little calmer about it still massively upset it’s stopped as it’s a proper unicorn.

It’s quite a tightly packed board.

Will add some pics as well for reference. I’ve added the socket for the oscillator and the one that came with it was put in the socket for testing. This is the original 16mhz cpu also back in it. Tried with and without FPU and changed board settings accordingly.

1B546456-345F-45C7-BCA1-102EF14FE674.jpegFCFACF0F-9D96-4D6A-8075-6F7A67437FD1.jpeg893DE1D1-268F-4447-B3BB-0310B83CE6A6.jpegF977A6AB-1677-4F05-BC07-AD4163DBF94B.jpegB4A6D3BE-C130-4C4D-BB51-7A695D387ABC.jpegBAF5C264-E9B8-4F82-BD6E-796BF04E3BE7.jpeg
 
So minor update, neither of my machines seems to work any longer with the VGA cards I’d added, I’d incidentally thought I’d fried one or both when testing my machines I wasn’t getting anything to screen when this first happened. Now I’m wondering if this is also a symptom of the issue. I say this as I have tested the vga cards elsewhere and both a trident and Ati card are working fine outside of the 5150’s. I double checked the switch settings 5&6 on for vga cards with bios 5&6 off for MDA. But curiously with the trident I get a machine beep one long two short. This without the inboard present btw.. so perhaps the inboard like the vga cards might be still functional but for whatever reason no longer in these boards.

Would it be worth me diagnosing both my boards independently before writing off the inboard ? Is there a good way to do this with the 5150, I understand there is also a diagnosis roms for these specced machines.

I guess my next question where to start?
 
Would it be worth me diagnosing both my boards independently before writing off the inboard ?
Yes, but try to be open minded. You have used the word, "boards", i.e. motherboards. There is more than a motherboard in an IBM 5150 computer.

But curiously with the trident I get a machine beep one long two short.
Do not assume that the 5150's POST is generating that beep pattern. The initialisation code within the BIOS ROM on VGA cards is quite capable of generating error beeps.

If one looks at the source code for the POST in an IBM 5150, the POST can only possibly generate {1 long beep then 2 short beeps} if the two video switches in SW1 are set to MDA or CGA. You have those switches set to EGA/VGA, and therefore, it suggests that the beeps are being issued by the Trident VGA card.

( Use [this] if you want to test the switches. )

You hear this beep pattern with the Trident fitted to either 5150 ?

Note: Actually, a bug in the IBM 5150's POST results in {1 long beep then 2 short beeps} if there is a damaged/corrupted BIOS expansion ROM at C8000. If you have something at C8000, then the situation at [here] is a possibility.

This without the inboard present btw.
Good.

Diagnose with minimum required configuration. For the VGA card problem, that minimum configuration is PSU+motherboard+VGA.

... I understand there is also a diagnosis roms for these specced machines.
You have an MDA card, and so you could try Ruud's Diagnostic ROM - see [here].
 
Yes, but try to be open minded. You have used the word, "boards", i.e. motherboards. There is more than a motherboard in an IBM 5150 computer.


Do not assume that the 5150's POST is generating that beep pattern. The initialisation code within the BIOS ROM on VGA cards is quite capable of generating error beeps.

If one looks at the source code for the POST in an IBM 5150, the POST can only possibly generate {1 long beep then 2 short beeps} if the two video switches in SW1 are set to MDA or CGA. You have those switches set to EGA/VGA, and therefore, it suggests that the beeps are being issued by the Trident VGA card.

( Use [this] if you want to test the switches. )

You hear this beep pattern with the Trident fitted to either 5150 ?

Note: Actually, a bug in the IBM 5150's POST results in {1 long beep then 2 short beeps} if there is a damaged/corrupted BIOS expansion ROM at C8000. If you have something at C8000, then the situation at [here] is a possibility.


Good.

Diagnose with minimum required configuration. For the VGA card problem, that minimum configuration is PSU+motherboard+VGA.


You have an MDA card, and so you could try Ruud's Diagnostic ROM - see [here].
Thanks @modem7

So I’ve a few things to report. I don’t currently own the diagnostic Rom but a couple of people have offered to sort me one if needed I’d started to look at the other one hosted on your site (Landmark) I hadn’t seen the Ruud’s one before.

I do have access to an IBM diagnostic disk, something I bought along with another system interestingly. This didn’t detect any faults likewise Checkit v3 also didn’t find anything. Both seemed age appropriate. I since tried it with a paired down system also without the additional cards seen on the 5151 below. But can run the tests a few times if needed. It’s weird in that it doesn’t report anything to screen so much as flash a quick code and returns to menu. I assume though if it fails it would say so. There is logging options so I may write it to drive B, I have spare 320kb disks I can format.
34ED8290-D3C3-432D-9EC9-02DC3D91CF64.jpegCDE72503-261A-4B8E-8328-44794945A011.jpeg8696835D-8097-462D-9B49-565709DB8E50.jpeg
Apparently this came from an ex IBM employee from Italy of all places when I bought an IBM p70

I tried the vga and nothing else (except keyboard and cpu) no output to screen but the familiar beep I mentioned before, it’s not consistent though first boot tends to give nothing at all, then subsequent boots I get 1 long, 2 short for both 5150 boards or if I switch the Rom from u33 between boards.

Note the Trident (TVGA9000I - manual on your site) was being used in this machine without issue prior as well as the Ati card (Mach 8) that I have. I do have other video cards including a CGA card.

For the cassette basic switch tests do I change these while powered on? I wasn’t certain having read the page and would this potentially result in loss of screen to the mda or are these values just read on boot?

Happy to give this a go of course. Thank you for your support. Happy to report back and will see if I can get the Ruud rom sorted if you think it trumps the floppy diagnostics?
 
Ok interesting apparently this Trident card doesn’t work natively on the 5150. (see thread below) I’m wondering now if by virtue of adding the inboard that this worked at all:


Also when I last tested and saw VGA working (The Ati Mach 8), I had in an Orchid Tiny Turbo switched to the 286. Perhaps this is why it worked. Natively with an 8088 / v20 the board might not support this particular card?

I’ll revisit and try with the Tiny Turbo or its card modding the Trident as suggested on that post.

So perhaps is back to being the inboard that’s failed overall.
 
Ok here’s an update. It does appear by virtue of having a cpu upgrade it’s allowing the 5150 to host a vga card that would otherwise seem incompatible (this was news to me) however the Tiny Turbo I have is less than reliable at booting initially into 286 mode, maybe the card has issues I don’t know.

The Trident card still not working, perhaps the inboard did something sufficiently different to see that work??

When it works, it works good.. but the ati card boots and is displaying the video bios and working as an output device with the Tiny Turbo in place. This likely true also for the Inboard 386/PC

11AD6096-9981-485C-91D6-CD26F16AFFEF.jpeg

So it is now leading back to my initial assumption that the issue is indeed with the inboard.

The vga it seems natively is not supported. Weirdly though the Tiny Turbo was still routing to the 8088 on it when the above was tested, perhaps the minor delay in initialising is enough to get the video bios up?

Anyway it’s a nuance I’ve obviously discovered by misadventure

So back next to diagnosing the inboard I guess.
 
My heart aches for you, for sure the Inboard is a unicorn. Here is what I would try:

Remove everything from the 5150 except the motherboard and an MDA card if you have it and try the inboard all by itself
This one isn't easy, but if you remove your motherboard from the computer, you can test power to each and every chip on the card. You can find spec sheets for most of them. I have fixed a lot of boards this way, one chip isn't getting power and you find a broken trace or bad resistor and boom you are back in business
Try a different computer all together
Get a new power supply (especially if you are testing with the same power supply, if not, might not be worth it.)

I have heard that the GAL (PAL??) chips on these boards go pretty frequently. Sadly this is not a high ownership board and there are no other projects beyond the 4 meg upgrade that I am aware of. Sorry this happened, I would be devastated.

Chris
 
My heart aches for you, for sure the Inboard is a unicorn. Here is what I would try:

Remove everything from the 5150 except the motherboard and an MDA card if you have it and try the inboard all by itself
This one isn't easy, but if you remove your motherboard from the computer, you can test power to each and every chip on the card. You can find spec sheets for most of them. I have fixed a lot of boards this way, one chip isn't getting power and you find a broken trace or bad resistor and boom you are back in business
Try a different computer all together
Get a new power supply (especially if you are testing with the same power supply, if not, might not be worth it.)

I have heard that the GAL (PAL??) chips on these boards go pretty frequently. Sadly this is not a high ownership board and there are no other projects beyond the 4 meg upgrade that I am aware of. Sorry this happened, I would be devastated.

Chris
Thanks Chris, there is member from the site that has kindly offered to try and diagnose it. Just time and patience. Like you suggested I did try with little more than an mda card but hadn’t taken the board from the machine. I’m not sure where one test the chips for power. I guess I’ve much more to learn. Thanks for the advice and kind words. Hopefully one day I’ll find another unicorn if not get lucky and resurrect my old card.
Regards

Mike
 
I’ve recently acquired an Intel Inboard 386/PC it’s been running great, I made a 4mb daughterboard thanks to others in this forum.

I swapped out the oscillator from 32MHz to 40Mhz and adding a socket in the process.

I put a Cyrix 486 cx486drx2 20/40 CPU on the board and a DLC 487 FPU.

It’s been running in my 5150 for the best part of a month and for a good few weeks with the new Oscillator.

I’ve been looking to upgrade my XT IDE and on running xtidecfg I hit an issue where the flash failed saying it wasn’t right. I took the cable from the inboard to access the Xtide to remove the label and check the rom type, did so and put everything back then black screen.

Noting I’d removed and unplugged the cable many times and the card as a whole also many times without issue. I know these things aren’t meant to be moved about a lot but getting the right combination and experimenting to get it right led to some tinkering and moving about. That and the limitation of 5 isa slots in the 5150.

The crux of my issue is now the inboard will no longer boot.

Replaced the original cpu and the 5150 board starts up, but put the inboard into either 5150 and a whole lot of nothing, no beeps just the psu powering up.

As there are no moving parts and only the parasitic cable to the CPU slot I figured I’d check the cable for potential faults.

I beeped out each hole in the idc connector to the dip pins that go to the cpu socket. 40 holes / pins and each matched up accordingly.

I’m at a loss as to what may have gone bad or if there’s any hope of ressurection, it’s a rare card and one I’d hope for many years of service to come.

So putting a plea out for anyone who knows how to help or what to try next.

I have swapped back the original oscillator and put a standard 16mhz Intel cpu back in the card and removed the fpu all to no avail tried with and without the daughterboard.

I have very little in terms of diagnosis beyond a standard multi-meter and probably not many more skills besides other than relying on the goodwill of others to throw me a life line.

Please help!

Mike
Hi hope you were able to fix your valuable inboard, would you please tell me how did you make the 4MB daughter board. I’m very interested and I have done PCBs before. These daughter boards has been beyond impossible to find. Thanks
 
Back
Top