Did not manage to fix it so far...
But next one:
After some cleaning it looks better:
Seems to be alive, but the delay line in the bottom left is loose, almost all pins broken. Need a replacement...
BIOS:
wrote:This indicates an error accessing/adressing the Video memory? Tried several VGAs, same result. To my surprise, the board is up after finishing the POST. I can switch the keyboard LEDs and a reboot with Ctrl-Alt-Del works.
My guess would be a problem with one of the address lines on the ISA bus. The way these mobos are wired it's very unlikely to be a low byte data line issue since that would upset the big chips with DMA controllers and whatnot. And the lower address bits seem to work OK-ish for the same reason but it could be a line being stuck for example.
First, test the video card in an 8-bit slot if you haven't already. This board seems to have two, and conveniently almost at the opposide sides so try both. If that doesn't work then it's going to be more work for you. Get a meter with good, fast beep on short and check all the address lines on the upper 8-bit slot. Here's a pinout: https://en.wikipedia.org/wiki/File:XT_Bus_pins.svg
It's easy, just one side to test, and not even all of it. For each address line check:
- that is not shorted to ground
- that is not shorted to +5V
- that is not shorted to any other address lines
- that there is connection to one of the pins of ST62BC004
Now, the exact pin on '004 doesn't matter as long as there is a connection somewhere. Take your time with it, you don't want to miss any of these tests. Also, inspect all of the slots and make sure there isn't a bent contact anywhere that would short the ISA A and B sides - that can be verified with the meter as well but a good look should suffice.
Like to give an intermediate feedback...cannot continue with beep-testing as somebeody else here is annoyed 😉
I tested again a CGA and VGA in different slots - same results.
So I tested all Pins in the 8-bit ISA Slot 1 against Slot 8 - all connected.
But there is no conection of all adress lines to the ST62BC004. But no surprise: I followed some traces and some go the area around the CPU. So there cant be a direct connection to the ST62BC004...
EDIT: OK, tracked the adress lines as far as possible. Some end in the ST62BC002 the remaining in the St62BC003.
From a optical point of view all perfect, but will verify.
wrote:But there is no conection of all adress lines to the ST62BC004.
Oops, sorry about that. I looked at the wrong bus and gave you data lines to check. So, actually it's like this:
Addresses A1-A12 go to 003 chip, and A13-A19 to 002. I haven't yet found which chip decodes A0.
For now though check the A13-A19 lines first, and then move to /MEMR (B12) and /MEMW (B11) which should be connected to 001, most likely through 30 ohm resistors. Let's get this out of the way first.
OK, did all tests again:
A1-A12 go to 003 chip
A13-A19 go to 002 chip
A0 to 001
MEMR and MEMW to 002
EDIT: Exchanged the A and B sides in the ISA shortage test...
So only connections are:
B31=B1=B10 (GND)
B29=B3 (+5V)
Nothing else...
What also might be inteersting: The status of the LEDs on the Post-card: All lights on, except the IRDY, which is blinking during POST sequence and turns off at the end.
Therefore the Reset-LED is "on" all the time. Think this should not be correct. Even more strange that the board powers up in this status.
wrote:What also might be inteersting: The status of the LEDs on the Post-card: All lights on, except the IRDY, which is blinking during POST sequence and turns off at the end.
Therefore the Reset-LED is "on" all the time. Think this should not be correct. Even more strange that the board powers up in this status.
Oh, a clue! Do you have a reset switch connected? Try without it. And measure voltage across the pins you'd connect it to, it should be +5V.
Whatever the result, also measure pins 5 and 6 of the 4069 chip, and then 5 and 6 of the 7406. Both are next to 8042 keyboard controller chip.
That reset problem sounds kind of familiar to me 😁.
Remember that there is a hidden conection under the power connector, mine was corroed and it was causing the reset state problems. Is taht tiny almos invisible corred line (connector removed).
The trace under the power-connector is OK. Measured on the bottom side.
No, I dont have a reset switch connected. But the switch works, when shorting it restarts POST-sequence, with Reset-LED on the POST-card lighting all the time. There is 5 V between the Pins
I measured the following voltages against ground:
7406:
14 5V
13 2,275
12 1,946
11 1,841
10 2,300
9 0,302
8 0,118
7 0 V
6 5,093
5 0,390
4 5,094
3 0,111
2 5,094
1 0,396
4069:
14 5,079
13 2,612
12 2,477
11 5,051
10 0,398
9 0
8 5,084
7 0
6 0,392
5 5,088
4 2,476
3 2,612
2 2,612
1 2,438
Next one: "Youth Keep Enterprise Co., Ltd." 301-02.
Was totally dirty. Removed all ICs and did a complete shower. But still working.
Before cleaning:
After cleaning:
Bios: 3.03YK2 Award
Compared the 7406 and 4069 against the not working board (left values), but dont see a fundamental differnce:
7406:
14 5V 5,086
13 2,275 1,807
12 1,946 2,185
11 1,841 1,908
10 2,300 2,250
9 0,302 4,830
8 0,118 0,082
7 0 V 0
6 5,093 5,081
5 0,390 0,125
4 5,094 5,082
3 0,111 0,030
2 5,094 5,081
1 0,396 0,126
4069:
14 5,079 5,077
13 2,612 2,573
12 2,477 2,501
11 5,051 5,079
10 0,398 0,126
9 0 0
8 5,084 5,077
7 0 0
6 0,392 0,125
5 5,088 5,075
4 2,476 2,505
3 2,612 2,575
2 2,612 2,573
1 2,438 2,460
Somehow I missed your first reply. Anyway, the values look good so I wonder what is up with the LED. Just to make sure I did not misunderstand, the test card RESET LED is on at all times and even when the reset switch is connected and used it doesn't change the LED state?
Next round of tests then:
- Check if there is continuity between reset switch header (the pin that is not ground) and pin 42 of the 001 chip. This is mostly a sanity test but take a note of the resistance value, should be below 1 ohm or low ohms at worst.
- Check resistance between 001 pin 17 and +5V, you should be getting very close to 1k. Check the voltage on this pin when the systems runs, and when reset switch is pressed (you might need extra pair of hands for this).
- Check if there is continuity between 001 pin 17 and ISA slot pin B2.
The 001 chip is a bit weird in that the pin count starts in the middle, next to the dot, and then goes CCW. There are silkscreen numbers in the corners that should help you.
EDIT: Oh and I dunk dirty PCBs in water (with some dishwashing liquid) all the time, haven't had a problem yet - the important part is to make sure they dry out before reapplying power. Scrubbing them while wet with a nylon brush is also not a problem unless you get a PCB in such a sorry state that the soldermask is peeling off on it's own.
Thanks for your new advise 😀 See my results in bold below
Just to make sure I did not misunderstand, the test card RESET LED is on at all times and even when the reset switch is connected and used it doesn't change the LED state?
Correct. Its on all the time. Also not blinking when pressing reset. However, when pressing reset POST sequence starts from beginning.
Next round of tests then:
- Check if there is continuity between reset switch header (the pin that is not ground) and pin 42 of the 001 chip. This is mostly a sanity test but take a note of the resistance value, should be below 1 ohm or low ohms at worst.
0,4 Ohm
- Check resistance between 001 pin 17 and +5V, you should be getting very close to 1k.
0,997 kOhm
Check the voltage on this pin when the systems runs, and when reset switch is pressed (you might need extra pair of hands for this).
0,191 V when running
5,107 V when pressing reset
- Check if there is continuity between 001 pin 17 and ISA slot pin B2.
No.
Btw, I had found datasheets for the 005 + 006, but not for the others...
Everything is OK then except this: No continuity between 001 pin 17 and ISA slot pin B2. Well that'a your problem right there. This might be interpreted by the video card as reset being asserted all the time so it's just not working.
This is sort of a global reset signal, generated by the 001 chip, it goes to other mainboard chips and should also go to ISA slots pin B2. You've already checked all the ISA slots against each other and found no broken connections (though you might want to retest B2 again now) so I wonder what happened to this signal. I suppose the only option is to try and track the copper trace.
First check continuity between other mobo chips. Now, the scan is a bit hazy so I'm not always able to tell the pin number, if you don't get a connection then try to sweep around the chip and see if there's a beep somewhere else. Pin 17 of 001 should go to:
- pin 40 of 003
- pin 19 of 005
- pin 53 of 002
- pin 92 of 006
The 1k pull-up on this signal is provided by "RA5" which should be a resistor pack below the 006 chip. Try to find which pin is it and solder a wire from there to one of the B2 pins (be extra careful not to mix the ISA slot sides!). Try booting with VGA card again. The point is it's still preferable to find the broken connection but first I'd like to check if that solves the actual problem or not.
Deunan you are the master!!! 😎
I noticed on the working board that the ISA reset pin is directly connected to pin 92 of the 006. And this connection was missing on the damaged board. But no chance to track the trace, the starting end ending points are hidden here.
As I have no soldering iron available here I made a quick&dirty fix with a wire and a needle. Power up..Reset LED is off...and works!!!!
Great!!! 😎 😎 😎
Many thanks to you, learned a lot again. And also many thanks to MMaximus for this board!
wrote:Deunan you are the master!!! 😎
...
Many thanks to you, learned a lot again. And also many thanks to MMaximus for this board!
That's amazing! Deunan, your expertise is legendary 😲 I've been following this thread and the other Kenitec Suntac thread and am really impressed that you managed to essentially tele-repair these boards.
Predator99, I'm really happy that this boards finally works! Maybe I need to send you my other Suntac boards as well so you can fix them. 🤣 I would do it myself, unfortunately I have literally no idea what you guys are talking about most of the time. I really want to learn at some stage as I have a good dozen of motherboards that won't start, but I'm such a newbie with electronics that I wouldn't know where to begin.
Well it's much easier if you have schematics. Plus we got lucky and all the "big" chips are still functioning properly.
Starting your electronics repair adventure with PC boards is probably not the best idea, but then again if you're willing to learn - why not. These 286 and even early 386 mobos and not yet that complicated.
You will need some tools though. Universal meter (preferably with fast beep on short) and I recommend a test card as well. With a test card you can select a mobo that's not completly dead, this raises the chances it can be repaired and it'll take less work in general.
wow, that was fast! another board saved from the trash! Congratultions predator99, and of course Deunan!
wrote:2019-1-26 @ 11:35 Did not manage to fix it so far... […]
2019-1-26 @ 11:35
Did not manage to fix it so far...But next one:
After some cleaning it looks better:
Seems to be alive, but the delay line in the bottom left is loose, almost all pins broken. Need a replacement...
BIOS:
I now replaced the damaged delay line with a socket and put a modified SN74LS14N on it.
Re: Delay line modern replacement?
With empty delay line socket the board is dead with display "--" on the POST card.
With the SN74LS14N it powers up. But there seems to be still a problem with the RAM. Supersoft dispays garbage CMOS data (not critical I think) and RAM failing bits 0,1,3,7. To my surprise I can exchange the RAMs in any way, its always the same scheme. I also removed all RAMs and still get the same result. The ICs are known working in the other Suntac-Board...
A problem with RAM that doesn't change when moving chips around would suggest the delay line substitution isn't working properly. I still think a HCT chip would do a better job than LS.
