• Please review our updated Terms and Rules here

Problem IBM AT 5170 mainboard

H

huubwen

Member
Joined
Jun 15, 2012
Messages
45
Location
Netherlands, Europe
I have bought an IBM AT 5170 mainboard with CPU and memory. The mainboards looks clean and well cared for.

Today I build it into a case and powered her up. Nothing happens. Well, the power supply spins continuously but no beeps or screen output at all. Nothing was connected to the mainboard. I have tested as well with a Hercules card en a VGA card.

The only thing that happens is that the num-lock, caps-lock and scroll-lock lights flash once when powering up.
I have also swapped the memory modules from bank 0 to 1 and 1 to 0. No change.

I grapped my multi meter and start testing the caps. I didn’t find any faulty caps.

Next I saw an object on the board that leaked badly. (see pictures) I measured its resistance. It was not conducting any power at all.

Could this be the problem? Can I remove this object and short-circuit it for testing the board?

What else could be the problem?
20120720_160153.jpg20120720_144235.jpg
 
According to my information, Y2 is a "12Mhz XTAL" crystal oscillator in which case I'd think its a fairly necessary part.

I'm sure one of the more clever guys on here will be able to tell you far more than that.
 
I'd clean the corrosion from that area first then see if that makes a difference. It could also be a bad RAM chip causing no video output or POST beeps. Even though you interchanged the memory chips, there could still be a bad chip in either bank causing the problem. To solve that, you're just going to have to remove each memory chip at a time and see if it cures the problem. I had the same no video problem on my AT. It turned out to be a bad chip in Bank 0. Time consuming, but well worth it.
 
In fact if I am reading it right, Y2 is the main clock crystal for the CPU with its output being fed through an 82284 clock driver chip, which halves that 12Mhz frequency to form the clock speed for the 286 CPU.

Personally I'd clean up the mess around it with a dry brush and see if you can see anymore about it.
 
I have cleaned up the mess. I was quite easy. It fall off in one big chunk.
Unfortunately the numbers on the crystal where worn away by the leaking. I cannot read anything off the crystal.
20120720_171408.jpg
 
Y2 present or absent won't affect the ability of the board to boot. It's a 32.768KHz crystal for the MC146818 CMOS clock/NVRAM (you can find these crystals on modern motherboards--they're not uncommon). The corrosion was probably from the battery in the 5170 leaking.

With the power on, measure the voltages at the motherboard connector. You can find the pinout here
 
Well, here are the readings. Those readings were taken with only the mainboard connected to the power supply.

P8
1 1,25
2 5,06
3 12,72
4 -12,16

P9
3 -5,11
4 5,06
5 5,06
6 5,06

I am not using the original IBM AT power suply but on old clone.
 
Last edited:
I start verifying the J18 RAM jumper setting as mentioned on the 5170 vintage-computer wiki.
I could not locate a J18 jumper on my type 2 mainboard.
I did find a J10 jumper (see picture). It is not jumpered. Should it be this way?
20120720_181332.jpg
 
huubwen said:
Well, here are the readings. Those readings were taken with only the mainboard connected to the power supply.

P8
1 1,25
2 5,06
3 12,72
4 -12,16

P9
3 -5,11
4 5,06
5 5,06
6 5,06

I am not using the original IBM AT power suply but on old clone.
Click to expand...

It looks as if your "Power good" line is not rising to the occasion. If you've got a 1K or so (anything between 220 and 2K ohms should do) resistor, try removing the PSU lead from pin 1 on P8 and jumpering pin 1/P8 to pin 2/P8 to get a logic high. I'm not certain that it'll do any good, but it stands out as an issue.
 
I found the problem. :D

Although the (clone) power supply did work properly on another XT286 board it did apparently not on the IBM AT board. The power good voltage was too low to get the cpu out of the reset state.
I used another power supply from a Pentium class system and the AT board did POST at once. It nicely counts 640 KB and then displays a whole bunch of errors. Obliviously because it had only a MDA card connected an no keyboard, floppy, battery pack etc.

Any guesses why the power supply did not work on the 5170 mainboard?
 
By "XT286" do you mean a real 5162 or just a clone? Lots of clones don't rely on the PG signal, but rather generate their own by delaying reset until +5 has been stable for an appropriate time. I'm not certain what the 5162 does.
 
strange but true. some of the IBM AT systems had a BOX in the lower drive bay
It was a load resistor
try plugging in a hard drive, you likely will not need the controller card
the clock battery area often gets crusty and needs cleaning
if it is really bad
hot water followed by shgaking blowing off
and put aside for several days
( small crvices can retain water)
the ocillator does not have anything in it to leak.
remeber the drive table in the IBM AT is very limited
even the seagate st-251 si not supported
when I was at INTEL DIV in 88, they claimed that " the last of the AT-286" did have an expanded drive table.
But I sent one out for repair and they substituted a clone phoenix bios. THIS IS likely not your current problem
there si a switch a small slide switch =on the mb.
to sent monp/color. If you only get a cursor that could be the problem.
Unlike the pc/xt a vga card should work and the cards and monitor are easier to find. Re,ener that the IBM AT supported 360 and 1.2 meg floppies but possibly not the 1.44 floppy.
the setup/.exe or setup.com from phoenix should wotrk
also remeber there si NO SETUP in bios
you must boot from a disk.
 
Well, next problem.

I plugged the original IBM MFM controller card and a VGA card into the board. I also connected a ST-225 to the MFM card and a 1,2MB floppy drive as well.
I booted the 5170 advanced diagnostics 2.07 from the floppy and configured the date/time and system parameters. I configured one fixed disk as type 2.

At boot it always displays 1782-Disk Controller Failure and sometimes 601-diskkete error. Nevertheless the floppy drive works fine. It displays no further errors.
Because I do not poses a genuine MFM control cable I use a floppy cable (straight, untwisted) and connect the drive B: part to the ST-225. The ST-225 is jumpered as drive 0.

I am sure the ST-225 is working. It comes from another working system.

The advanced diagnostics allows me to manual add option 17 (fixed disk and controller). From there I can get to the low level format menu. The drive does nothing when asked to low level format the drive.

The LED of the ST-225 does not lid up at all. Only when I choose the option to move the system. It then lids up but the advanced diagnostics displays an error that preparing the system for transport failed.

What am I missing here?
 
1782 is a 'Fixed disk controller failure' so I'm not suprised you are not able to LLF the drive. You need to try another hard drive controller.
 
You must log in or register to reply here.
Back
Top