Does the clock advance while you're in BIOS setup? If not, the crystal (I'm assuming Y2, 32.768kHz) might be bad. Had this happen at least once.

[edited mistake, should be 32.768kHz or 32768Hz]

quicknick wrote on 2022-10-03, 11:32:

Does the clock advance while you're in BIOS setup? If not, the crystal (I'm assuming Y2, 32768kHz) might be bad. Had this happen at least once.

It does not. Interesting.
It does advance in DOS though.. I guess it is not the same.
How can i check the crystal? With oscilloscope? I have only a multimeter right now. And a logic analyzer that i never even opened 😝

I think you need an oscilloscope for that, or if your multimeter can also measure frequency it could work.
But 32.768kHz "clock crystals" are abundant and cheap, you could just replace it.

you need high impedance probe to measure crystal signal directly, so clean U26 to see the markings, look at the datasheet and measure on the output of whatever 74 series chip circuit implemented there. Also check if the C50 cap is all right. Or just replace blind 😀

and hole hell, take that motherboard out of the case and wash it 😀

rasz_pl wrote on 2022-10-04, 04:06:

you need high impedance probe to measure crystal signal directly, so clean U26 to see the markings, look at the datasheet and measure on the output of whatever 74 series chip circuit implemented there. Also check if the C50 cap is all right. Or just replace blind 😀

and hole hell, take that motherboard out of the case and wash it 😀

The chip is a MC14069UBCP - an hex inverter, and this is its datasheet:
https://pdf1.alldatasheet.com/datasheet-pdf/v … C14069UBCP.html

I see that the crystal is connected to pin 6 (OUT3) and I get continuity to pin 13(IN 6) as well.. it is normal?
The other leg of the crystal is connected to C50 which is connected to pin 11 of U26 (IN 5) on its other side. The capacitor is not short, so... good?

In any case, I don't have an Oscilloscope yet, so I cant test it. I ordered crystals, so at least I'll be able to replace it when they arrive.

BTW, since the time I took the pictures a few years ago, I cleaned the mother board as much as I could and even repainted the case.. I think it looks as good as it can right now 😀

anetanel wrote on 2022-10-04, 08:21:

I see that the crystal is connected to pin 6 (OUT3) and I get continuity to pin 13(IN 6) as well.. it is normal?
The other leg of the crystal is connected to C50 which is connected to pin 11 of U26 (IN 5) on its other side. The capacitor is not short, so... good?
IMG_7022.jpg

similar circuit https://minuszerodegrees.net/images2/5170_bat … y_circuitry.jpg

I finally bought an oscilloscope!
Now, where do I need to attach the probes to find if the crystal is bad? 😅

I think I figured the probe placement. I measured pins 11 and 10 of the IC, and got a nice sine wave of about 32.7 Khz at the input, and square wave at the output.
I guess that means that the crystal is good?

trace where the clock is going

rasz_pl wrote on 2022-10-13, 17:04:

trace where the clock is going

When compared to the diagram of the 5170 that you posted earlier, it is very similar but not exactly the same.
I replaced the pin numbers I was able to identify in the diagram.
The main two differences I spotted are:
1. The output at pin 6 is connected to the input of pin 13
2. I could not find anything that is connected in parallel to the input of pin 5.
Also, the 2 unused pins (3 and 5 in the 5170) are chained, instead of both pulled low.

Here is what I was able to draw.. sorry :\

I mean past the CD4069

rasz_pl wrote on 2022-10-13, 18:06:

I mean past the CD4069

What cd4069?
Did you mean mc14069?
If so, from whic pin? That's what I'm trying to find...
Sorry for my cluelessnes...

anetanel wrote on 2022-10-13, 18:41:

What cd4069? Did you mean mc14069?

same chip
pin 10, or 12, look at all the inverter outputs for any traces going to CHIPS SCAT 82c235 pin 140 https://www.datasheetarchive.com/pdf/download … =O&term=f82c235 page 53

or just use scope on pin 140

I think I found the issue.
I had to take the board out and carefully examine it from both sides under hard light.
output pin 12 is connected to input pin9, and the output pin 8 is connected by several vias to the 28c23g chip.
The problem is that one trace is broken due to corrosion (marked by the red arrow) 🙁
I guess I'll need to get some thin bodge wires and start soldering.

I need to invest in quality sharp probes for my multimeter... I had to improvise by connecting the probe to tweezers with an alligator clip 😜

You can _slightly_ sharpen the tip depending on quality of the probes (thickness of nickel plating over copper core).
You could use this opportunity and put motherboard in a bathtub with washing powder/dishwasher, use old toothbrush to clean all nooks an crannies, rinse well, then hair drier to push water from under chips/sockets and dry for a day near the radiator. Will let you see better other potential corroded spots.
No need for bodge wire for this one, scrape soldermask with a bit of soft sandpaper and tin the trace. Solder should patch it up on its own.

It is alive! 😀 😀
For the life of me, I was not able to bridge the gap with solder only. There was no trace copper underneath to grab it.
After more than an hour of trying, fearing I'll do more damage to the board, I cut my losses and bodged it on the back. Connected pin 8 of the IC to one of the vias after the broken trace.
Powered the board and... No more error! And the clock progresses in BIOS too!
I may replace the wire with an appropriate gauge later. All I had was pretty large, so I had to solder single stand...
Many many thanks @rasz_pl and @quicknick!!

👍