What do you expect without CPU...?
Install one and try again...

Predator99 wrote on 2020-09-23, 10:47:

What do you expect without CPU...?
Install one and try again...

I've put a 16MHz AMD in it. Changed the oscillator and used a 12Mhz Intel. Same thing. POST card registers constant RESET signal. Something must be dead on the board I guess.

http://www.win3x.org/uh19/motherboard/result/?name=jam
I didn't find it in uh19 either but I found these boards which may be made by the same brand.

Do you have a EPROM reader ? Dumping that bios could help identifying the board if it's a Ami bios.

DAVE86 wrote on 2020-09-23, 12:52:

constant RESET signal. Something must be dead on the board I guess.

I recently troubleshot[1] a 386SX board with similar symptoms. I identified two possible sources for constant reset.

1. The "reset generator". It typically has a R/C delay generator, evaluates PWRGOOD and/or uses a comparator on key voltages (+5V/+12V). Often, the reset generator circuit is close to the CMOS NiCd battery, and traces might be etched away by a leaky battery.
2. The clock source. RESET to the CPU needs to be synchronized with the clock. If the clock signal is missing, RESET stays active.

Both causes are worth investigating. Your chipset is called CS8220, the datasheet should be retrievable using Google, as I retrieved it some time ago without problems. The 82c201 has a lot of pins involved with clock and reset, which are worth a probe. Even if you don't have a scope, use a DMM to verify that the clock pins show a voltage between 2 and 3 volts, to check for stuck clock signals.

1: or troubleshooted? Native english speakers, please advise!

I had a similar issue on a XT board I had. It ended up being the source n°2
The 14.3MHz crystal had been removed. Once I repopulated it, the board started.

This board seems to have one but maybe it's dead or something's damaged ?

These frequent motherboard identification threads make me wonder: why are pre-Pentium boards often unmarked and difficult to identify. What reason could the manufacturer have had to not just think of a type designation and write it on the board?

According to https://picclick.com/PC-AT-80286-10MHz-Mother … 2819842823.html it's a Jaton board, but no idea what model.

Alright, thanks everyone! I got the two ROM contents unified, but no hint to motherboard id or microcode . Looks like the 3.6v nimh battery was removed long ago. No damaged traces. I have to test those transistors and a lm339 and some passives. Frequencies 32.7Khz and 14.3Mhz are present. I might reverse engineer the circuits and probe the chipsets. It will be fun. I never came across a fault like this.

There's absolutely no info about the mobos history (Got it from a garbage AT case). I was surprised there is no info on it anywhere (aside that 2017 ebay auction).

@oetker maybe they didn't think it was mattering. They were making clones. Since there was no internet back then, knowing what board you had was useless anyways since you couldn't simply look for it on the web and find the manual ^^.

@DAVE86
Can I have a look ?
If you figure out something I'll add every piece of info to uh19 😀

Perhaps they wanted to be unknown.

Deksor wrote on 2020-09-23, 21:30:

Can I have a look ?
If you figure out something I'll add every piece of info to uh19 😀

I've attached the orignal ROMs. Check it out 😀

I've done some measurements. Took the board appart a little. Several things seems to have failed. Nihm battery charging voltage is high 8V -9V. There is a shorted 4148 diode. Several open resistors and some out of their values. P82C206 has missing Vcc. One of the customs PALs is super hot at all times. At this point it seem to be a lost cause...

DAVE86 wrote on 2020-09-24, 15:10:

I've done some measurements. Took the board appart a little. Several things seems to have failed. Nihm battery charging voltage is high 8V -9V. There is a shorted 4148 diode. Several open resistors and some out of their values. P82C206 has missing Vcc. One of the customs PALs is super hot at all times. At this point it seem to be a lost cause...

NiCd/NiMH charging could happen from the +12V rail using a suitable current limiting resistor. In that case, 8V-9V open circuit voltage might be OK, as long as the short circuit current is 10mA max. I doubt it, though, especially as you are talking about a dead diode, which might be involved in that logic, too. I am surprised about open resistors. The resistor values are high enough that they should not be able to be burnt out by 5V (5V at e.g. 100 Ohms makes 50mA. 50mA * 5V = 250mW (or 1/4W)). I would not be surprised by resistors connected to broken traces, though. If you measure resistors in-circuit, you might find their value to be displayed way too low, because some of the test current is not flowing through the resistor, but through other parts of the board.

Missing VCC on the '206 is strange. Boards didn't have power management to cut power from the '206 at those times. VCC on that chip should be directly connected to the AT power supply connector. This definitely is a broken trace unless your measurement is off.

I'm afraid that a hot PAL might indicate the PAL is fried, and that would be game over unless you reverse engineer the board around the PAL far enough to guess the logic. On the other hand, original PALs (not GAL, PEEL or PALCE) are PROM-based (fuse or anti-fuse technology) and thus quite robust. Maybe that thing just gets hot due to a shorted output, but the program is still intact. You might want to try whether you can read back the PAL (unlikely, but I recently had success with PALs from a OPTi local bus graphics card) if you have a programmer at hand.

Thank you very much for the BIOS upload.
I can't see any useful piece of info here, but at least it might be helpful for someone in the future 😁

@mrkarcher which opti board ?
this might be a neat feature for UH19 ? "PAL dumps" 😀

Make sure you have the proper datasheet, there are 206s in different packages (QFP and PLCC) and the pinouts aren't the same. Missing Vcc on any chip is unlikely without extensive (visible) damage to some traces (or almost impossible if the Vcc is routed through one of the inner layers, like on most of my 286 motherboards, only exception being the PCChips M209 which is dual layer).

@mkarcher - what are you using to read PALs? I have the TL866 but apparently it doesn't support them.

Edit: tried your BIOS in PCem and it doesn't want to start with it. I might be doing it wrong, though, so maybe someone else can try it too...

The CHIPS plcc packages sit on through hole adapters and I noticed some of the vias barely had any solder and were very a bit wiggly. Vcc is present now but no change in behaviour!
I've tested the passives in circuit and taken out too. Sometimes old resistors values start to drift and stretch. This one is open. Right after the 12V line. Another resistor to ground... These things make up a voltage divider for some reason. No sign of physical damage.

What could this be? U71 IC under the CHIPS P82C202
JATON-ATU71A

Also getting kind of warm.

I have a dodgy PAL reader that can run truth tables. I might try and see if that PAL is really fried.

So the psu is bad actually. One of the most low cost simple unit I've ever seen. Turns on with some 10-15W load and output voltages are crazy fluctuating.
The mobo is from a rusty AT case that had nothing but the psu and board in it.

Deksor wrote on 2020-09-24, 20:43:

@mrkarcher which opti board ?
this might be a neat feature for UH19 ? "PAL dumps" 😀

It's the no-name Taiwaneese OPTi Local Bus graphics card I mentioned over here: OPTi Local Bus variants.

quicknick wrote on 2020-09-24, 20:48:

@mkarcher - what are you using to read PALs? I have the TL866 but apparently it doesn't support them.

I use a HiLo systems ALL-03. That system uses an ISA interface card is is supplied by the PC power supply, but there are different versions that can be attached to the parallel port using an external power supply (The Sunshine Expro 60 and Sunshine Expro 80). There is a lot of information about this programmer on http://matthieu.benoit.free.fr/120.htm.