• Please review our updated Terms and Rules here

Turbo XT clone - dead motherboard

dieymir

Experienced Member
Joined
Jan 11, 2017
Messages
245
Location
Spain
Hi,

I have one of those Taiwanese XT clone boards like the one discussed in this Russian forum

http://www.phantom.sannata.ru/forum//index.php?t=26329

But mine looks much better :)

IMG_152654530174F.jpg

despite of its gorgeous appearance it does not boot :( Appareance it's not everything

I have followed minuszerodegrees Minimum Diagnostic Configuration (is this trademaked or patented? ;)) with the following results:
- There are no beeps
- PSU fan works (I know for sure that the PSU it's OK since I got it from a working system)
- I do not have a multimeter to check voltages but some chips on the board get warm included the 8088-1 so I think there are not shorted capacitors
- I have tested several RAM chips on bank 0 (I left populated only this bank) unsuccessfuly
- I have noticed that the the DM74S74N flip flop in U39 gets extremely hot It burns when touched and I was afraid that it would catch fire

Se the photo:

IMG_152654535503F.jpg

Any ideas ?? Could this be the root cause of the MB failure?

Thanks

BTW the MB has a stick over the BIOS chip that read TD3.91, there is this BIOS dump on minuszerodegrees:

BIOS source Motherboard: Unbranded XT clone (CPU: 8088 @ 4.77 MHz)
Supplier giobbi at the Vintage Computer Forums
BIOS chip type 2764
BIOS contains the strings "86 © TD3.86 ID: 75102637"
"COPAM © 1985"
"Author: Thomas Lao"
Comment BIOS chip is labelled: TD 03.86
Download Clone XT BIOS - TD3.86 ID 75102637.bin
and in the Russian forum they have posted version TD3.93 I repost it here in case there is someone interested (I have tested it on PCem and boots fine)

View attachment td393.zip
 
Last edited:
Unfortunately, replacing ICs randomly will not be a very fruitful experience. You'll need some tools; even a logic probe and cheap multimeter would be great improvements at this point.

Otherwise, you're in the position of one trying to repair a vintage auto and having only a rusty screwdriver.
 
If you have access to an EPROM burner you could try the 5160 Supersoft Diagnostics ROM.
I think they may work on a Non-IBM board.

http://minuszerodegrees.net/supersoft_landmark/Supersoft%20Landmark%20ROM.htm

I've had several boards that appeared totally dead, and SuperSoft ROMs at least identified
the area of the error. The chip getting that hot doesn't sound good, I would check the PS
voltages and see if they are correct.
 
- There are no beeps
- PSU fan works (I know for sure that the PSU it's OK since I got it from a working system)
- I do not have a multimeter to check voltages but some chips on the board get warm included the 8088-1 so I think there are not shorted capacitors
- I have tested several RAM chips on bank 0 (I left populated only this bank) unsuccessfuly
- I have noticed that the the DM74S74N flip flop in U39 gets extremely hot It burns when touched and I was afraid that it would catch fire


First, as mentioned, trying to troubleshoot without some basic tools is pretty much futile. A half-decent set of basic tools shouldn't set you back more than $10 (US). A cheap DVM (sometimes free with Harbor Freight coupons) would be another $5-$10. It could still tell you voltages and check for shorted components.

Having said that, (In reverse order of your diagnosis) I'd say that given your description, the 74S74 is likely bad. Problem being is that you don't know if it's just the IC or something driving it too hard (adjacent filter cap shorted, shorted trace/pins, etc). How did you check the RAM chips? I'd remove the P/S from the unit if you're sure it's OK...at least until you've troubleshot the board first. And obviously you won't get any beeps until you get some semblance of a boot going (i.e.---repairing the MoBo)
 
One other thing you can do is resetting the system while pin IOCHRDY has been tied to GND. Have a look for the pinout of the ISA bus on my site. Then measure the voltages at the data bus and address bus. Generally everything above 2.5 Volt is considered (H) or 1, below 0.7 is (L) or 0. If in between, then you have a problem. The address bus should be outputting the address FFFF0 and the data bus should read EA. FFFF0 is the first address the 8088 CPU outputs after a reset and EA stands for "jump far" (unless you have a weird BIOS). If neither of the buses outputs these values, you have a serious problem. If the address bus outputs FFFF0 but the data bus outputs something else, check what the EPROM outputs. If it outputs EA, then one of the buffers between the EPROM and ISA bus is broken. If it outputs FF most probably it is not selected and thus start checking the logic between the CPU and the EPROM.

If the output is OK, then something else is going on. I have built a debugger for the PC based on this design and it enables me to step through the BIOS. Best is to use a BIOS that comes with the source code as well. If at a certain point the BIOs does THIS and you expected THAT, you probably have found a possible reason why the board won't start. Now you have to find the reason why the PC does THIS instead of THAT.

Good luck!
 
Yes you are right, I'll get first a basic set of diagnosing tools. Now another question, What is that basic set of tools?:
- A multimeter
- ...

Thanks
 
I have followed minuszerodegrees Minimum Diagnostic Configuration (is this trademaked or patented? ;))
Patent pending :)

... with the following results:
Keep in mind that the procedure was written for the IBM 5150/5155/5160's. Applicability/suitability for a clone will depend on how close the clone is of the IBM 5150/5155/5160 (including BIOS).

- I have noticed that the the DM74S74N flip flop in U39 gets extremely hot It burns when touched and I was afraid that it would catch fire
"It burns when touched". For a 7474, that is very very suspicious. If it was my motherboard, and I measured the +5V within spec, I would be replacing the 7474.
 
"It burns when touched". For a 7474, that is very very suspicious. If it was my motherboard, and I measured the +5V within spec, I would be replacing the 7474.
Yes, as soon as I get a multimeter I'll check this.
 
Back
Top