• Please review our updated Terms and Rules here

5170--No Video or POST.

D

DOS lives on!!

Veteran Member
Joined
Mar 14, 2011
Messages
2,303
Location
East Tennessee
My most exciting score I got at the E-Waste event was my IBM PC AT. But of course, the one I most wanted to work doesn't work.

When turned on, the computer comes to life, but it it doesn't beep, the floppy drive doesn't engage, and nothing comes up on the screen. Also none of the keys will respond on the keyboard, which is working. (The keyboard from my XT.)
I have removed the cards one by one, turning the AT on after each removal and it still doesn't work. Also, (only on this computer), the monitor makes a high pitched squealing sound when the computer is on. It doesn't do that on any of the other computers it's connected to.

Any suggestions as where to start?
 
Well first I'd try another video card, just in case and check the voltages from the PSU.

But failing that are you able to burn EPROM's?
If so, fit a CGA or MDA card and use the Lansoft ROM images here: http://www.minuszerodegrees.net/rom/supersoft.htm (5170 at the bottom)

Certainly the quickest way to find faults as it brings up the display much earlier in the test sequence.
Could also try piggy backing memory chips in bank 0 in case one of them is faulty- I prefer Lansoft because it gives you better direction and memory piggy backing can be rather time consuming (I find it frustrating) - where as Lansoft will tell you what is wrong (unless it's something core, like the CPU).

Edit: also on minuszerodegrees check out 5170 under DRAM here http://www.minuszerodegrees.net/failure.htm - one of those cheap POST cards could tell you something
 
Check minuszerodegrees.net for further info. Make sure to pull all of the expansion cards but video and see what happens. Check if you have a Type 1 board (piggybacked ram is the tell-tale sign). If you do - congrats! - it doesn't lock you into using IBM-only floppy/HD controllers. And also congrats - you probably have something as simple as a bad RAM chip of the piggy-back variety (if you have a bad ram chip, your machine will do exactly what it's doing now - mine did; it had 3 bad chips)

Of course make sure to press in all socketed chips first, yadayada... Good luck!
 
My 5170 does have the Type 1 motherboard, and also has the piggyback RAM chips. I tried just the video card inserted, and it still has the problem.

So now I begin pulling them out one by one and hoping I find the culprit.......
 
I've already pulled out three of the four rows of chips and still the problem persists. The fourth row is not easily accessible due to the metal piece that the speaker is attached to and guides the cards. Is there a way to remove this piece?
 
I thought that you were going to measure all of the voltages (including the POWER GOOD line) !

DOS lives on!! said:
The fourth row is not easily accessible due to the metal piece that the speaker is attached to and guides the cards. Is there a way to remove this piece?
Click to expand...
No. It's part of the chassis. If you have to, remove the motherboard.

DOS lives on!! said:
I've already pulled out three of the four rows of chips and still the problem persists.
Click to expand...
On a type 1 motherboard, bank 0 is the 2 rows of RAM at the front of the machine, and bank 1 is the next 2 rows.
The type 1 motherboard will not start up if bank 0 has been removed (or has bad RAM chips).
Removal of bank 1 as a diagnostic activity is valid.

You could try swapping banks 0 and 1, but that will only work if all the chips in bank 1 are good.

I suggest that you measure the voltages soon. It's a good step 1.
 
Actually after checking the power supply check the processor. Failing to POST is failing somewhere before the bios can even get read or function. Bad RAM should give you beep codes, bad video would normally do the same.
 
Sometimes the 5170 battery packs will leak onto the motherboard.

Is the motherboard clean in the area around the keyboard connector ?
 
barythrin said:
Bad RAM should give you beep codes,
Click to expand...
On a 5170 type 1 motherboard, an early POST test is that of the first 64K of RAM (which of course resides in bank 0).
Failure of the test results in the POST sending a repeating series of codes to any POST card that is plugged in (set to port 80h).
The codes are shown in the 'Dynamic RAM Chips' section at http://www.minuszerodegrees.net/failure.htm
No beeps are issued.

Failure of RAM after 64K (a later POST test) is displayed on-screen.
 
I've had some success in reviving dead boards by removing everything removable, including the battery, power supply, all cards, everything, and letting the board set, overnight or longer(sometimes it doesn't take that long, but in some cases it did) and then putting it back together again very slowly, step by step.
 
OK, so the power supply is outputting the correct voltages. The motherboard is clean of battery leakage. When I remove Bank 1 of the piggybacked memory, the computer still has the same symtoms. I was told of a diagnostic EPROM I could burn, but I do not have an EPROM burner.

Can anyone recommend a good POST card and/or EPROM burner? Since it seems that there is a faulty RAM chip in Bank 0, I'd need the card to identify the sore thumb.
 
DOS lives on!! said:
Since it seems that there is a faulty RAM chip in Bank 0, I'd need the card to identify the sore thumb.
Click to expand...
Seems like it would be simpler to just switch the banks of ram and see if you get lucky and that 1 has good ram where 0 doesn't. From there it would be much easier to isolate the bad chips if there are any.
 
Stone's spot-on... It's a bit of a pain, but it's what you have to do. If you can, go get 2 or 3 "new" RAM pieces from ebay. Then swap those in one at time... Start with swapping in place of chip 1 in bank 0, then attempt to boot. If it still doesn't post, then put back in the original, and swap the new into chip 2 and power it up again. Wash/repeat until you get to the end of the bank, or until you get a successful boot. If you're lucky, you'll only have the one bad chip in bank 0 and it'll boot. If you're like me, you'll end up with a bad chip in bank 0 and two bad in bank 1, and you'll have some REAL fun with chip swapping :)

FWIW, while it's a pain, it's not really that time-consuming... I think that Chromedome45 and I had mine working 100% within 20 minutes, and that was with two bad chips in the same bank, which required some extra guesswork and trial/error with the chip swaps.
 
Making some progress! I found the faulty RAM chip, which was the third chip in Bank 0. Now when I start the computer, it emits two sets of two short beeps and the screen still stays blank. What do those beeps mean?
 
It means you need to get the display working, because that's where the error code is being displayed. It's a POST error. This is assuming you have a true IBM BIOS.

It's gonna be one of these:

http://www.personal.psu.edu/~sab139/postcode.htm

My guess is that it's a 2xx error, probably a 201, IOW you still have memory problems. Did you stop searching after you found that bad chip? :)
 
Last edited:
I reseated the display card and now it makes 1 long beep and two short beeps, then the second set of two beeps. And also there is this switch on the motherboard, being able to set it to C76 or SW1. What is that switch for?

It also makes the same beep code when the video card is removed.
 
You must log in or register to reply here.
Back
Top