tikoellner wrote:

What is left to be done:
- Maybe removing cache chips;

I've had a few problems recently with 486 boards refusing to boot due to faulty cache chips - hope it's something as simple as this for you too!

yawetaG wrote:

^ The way I read that, that means you either use two single-sided SIMMs in slots 0 and 2, or two double-sided SIMMs in slot 0 and 2, so two memory modules. There's no explicit mention of the board being able to boot with only one SIMM.

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memory, Pentium had to use pairs though), so I suppose it might not have been explicitly mentioned as this was standard practice back then anyway?

There will always be exceptions, but usually a single SIMM for any 486 board should suffice (unless 30p SIMM but that's a different story).

And gotta love those Engrish manuals of that era 😁

Great! Do I need to do any jumper adjustments in order to boot without cache chips?

Tetrium wrote:

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memor […]
Show full quote

yawetaG wrote:

^ The way I read that, that means you either use two single-sided SIMMs in slots 0 and 2, or two double-sided SIMMs in slot 0 and 2, so two memory modules. There's no explicit mention of the board being able to boot with only one SIMM.

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memory, Pentium had to use pairs though), so I suppose it might not have been explicitly mentioned as this was standard practice back then anyway?

There will always be exceptions, but usually a single SIMM for any 486 board should suffice (unless 30p SIMM but that's a different story).

And gotta love those Engrish manuals of that era 😁

I misunderstood Vetz' post as saying that the only information on SIMM combinations in the manual was that small piece of text, and based on that text alone I couldn't figure out whether single modules were supported. Therefore, I thought he had misread it and that this motherboard was one of the exceptions 😅 . That's all.

Simply removing cache chips did not help.

I am slowly giving up. It seems that BIOS is the last thing to try. BTW - the BIOS chip itself does not seem to have been messed with by the previous owner. I'm sorry for the silly question, but can BIOS data really get corrupted just because the motherboard was long unused?

I don't own one of these boards, but going by the manual I'd say that if you're trying to boot with an intel DX or DX2 the settings you should use are:

JP16: 1-2 & 5-6 short
JP17: 1-2 & 5-6 short
JP18: 1-2 & 5-6 short
JP19: 1-2 short
JP20: open
JP21: open
JP22: 1-2 short

I know that setting says "3x clock", but I think that is only meant for the DX4 clock, not DX or DX2.

Regarding the cache chips, I see no jumper settings in the manual for "no cache", so I don't know if removing all chips is actually an option. Do you have other chips to use in place of the current ones?

In theory it is possible to have the BIOS data to degrade and become corrupt, but I have never experienced that. Reprogramming the ROM chip will help in that case. But I would say there is higher probability that your problem is due to a short or a failed component on the board. Do you have one of those ISA POST cards to test?

yawetaG wrote:

Tetrium wrote:

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memor […]
Show full quote

yawetaG wrote:

^ The way I read that, that means you either use two single-sided SIMMs in slots 0 and 2, or two double-sided SIMMs in slot 0 and 2, so two memory modules. There's no explicit mention of the board being able to boot with only one SIMM.

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memory, Pentium had to use pairs though), so I suppose it might not have been explicitly mentioned as this was standard practice back then anyway?

There will always be exceptions, but usually a single SIMM for any 486 board should suffice (unless 30p SIMM but that's a different story).

And gotta love those Engrish manuals of that era 😁

I misunderstood Vetz' post as saying that the only information on SIMM combinations in the manual was that small piece of text, and based on that text alone I couldn't figure out whether single modules were supported. Therefore, I thought he had misread it and that this motherboard was one of the exceptions 😅 . That's all.

No worries, happens to the best of us 😁

And lets be honest, these old boards are quite a challenge to setup 🤣!

I looked over all your jumper settings. It seems OK, except the printed manual does not have a jumper set for JP20. I know some spec sheets mention to short JP20 1-2 for 2x, but perhaps try to remove that jumper.

I think I found your problem. In my notes, I have JP16 is set to 1-2 for voltages 3.45V, 3.6V, or 4V. If JP16 is set to 1-2, then you must use JP32 to set voltage between 3.45 (1-2), 3.6 (2-3), and 4V (open). However, if you are using a 5 V CPU, which you are, you must remove the jumper 1-2 from JP16.

Tetrium wrote:

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memory, Pentium had to use pairs though), so I suppose it might not have been explicitly mentioned as this was standard practice back then anyway?

Wasn't there an Intel Saturn, Saturn II, or Aries PCI 486 chipset which wanted the memory installed in pairs?

Anyway, in addition to my last post, generaly, JP20 is for setting the multiplier on 3x/4x chips, where 1-2 open = 3x and 1-2 closed = 4x.

As for the BIOS chip, I have seen it where a user tries to flash program a UV BIOS chip and this ruins the data.

I removed JP16 1-2 jumper and JP20 1-2 jumper. It still refuses to work.

The CPU gets warm, no beeps, the chpset seems to get little warm as well (not entirely sure).

feipoa wrote:

Tetrium wrote:

I tend to agree with Vetz in this case and afaicr using single SIMMs was completely standard for 486 boards (due to 32-bit memory, Pentium had to use pairs though), so I suppose it might not have been explicitly mentioned as this was standard practice back then anyway?

Wasn't there an Intel Saturn, Saturn II, or Aries PCI 486 chipset which wanted the memory installed in pairs?

This might very well be the case, but personally I think that when it comes to 486 chipsets, you're the more knowledgeable one compared to me 😉

It's possible, there's always exceptions (like that s7 VIA chipset that made it possible to install single SIMMs (32-bit) even though the CPUs that were supported by it (Pentium-class CPUs) used a 64-bit memory bus).

tikoellner wrote:

I removed JP16 1-2 jumper and JP20 1-2 jumper. It still refuses to work.

The CPU gets warm, no beeps, the chpset seems to get little warm as well (not entirely sure).

Well, at least you know the voltage is getting to the CPU! Could provide a high-res photo of the memory module you are using? Or relay the part numbers from the individual memory chips? Could you try using a 3.3, 3.45, or 3.6 V chip and jumpering it accordingly? Perhaps an AMD DX4-100 or Intel DX4-100. Could you also try using a single ISA VGA card with nothing else inserted? Have you ordered some EEPROM flash chips or have on order one that is pre-programmed?

Hey there! I will provide the data later today, after the Christmas Eve.

Earlier this day I have traveled to my shack 80km from my Kraków to get some PS/2 DIN adapter 😉.

Now, not really much happens when I connect the keyboard (I used my Model M for tests).

After I turn on the PSU, first the 'scroll lock' LED blinks, then 'num lock' blinks, and then there is no further led activity. I presume that those blinks are produced rather by the internal keyboard controller rather than any motherboard activity. At least there is power getting to keyboard socket.

I think I found something that might be the reason.

Yesterday I brought some old PSU that I know to be very reliable. I connected it to the board - everything seemed OK: the fan was spinning, the CPU was getting slowly warm, etc.

There was, however, that cought my attention: silent hissing sound coming out from the PSU. This was not present with the previous ones I have tried. I connected the PSU to my other, PCI based motherboard - and the hissing sound was gone.

I assume this could mean some short on the motherboard (maybe on a specific power line), which for any reason does not trigger any safety mechanism in the PSU.

@Feipoa:

I have reviewed the RAM I was using: it's SEC Km416c1200aj-7. So it's single sided, 4mb, FPM, 70ns.
I have also set the jumpers and installed DX4. Well, nothing happened. Same dead stuff.

What else... I have tried putting some random BIOS in the motherboard, but it acts all the same way.

Now the biggest question: should I look for the short (where to start?) or should I just dump the board altogether and save some nerves?

My idea is that there is just one liquid-based capacitatior on the board and that could be the culprit, even though id does look OK. I have a multimeter. How do I check it without removing it first?

A hissing sound coming from the PSU, but only when connected to the VL/I-486SV2GX4?

If would, for starters, check the resistance between +12 V and GND, -12 V and GND, +5 V and GND. Is there a -5 V? I don't recall, but if so, check resistance between -5 V and GND. If there is a short, you will usually get a reading of under 10 ohms. If there is a short, I would check each capacitor on the motherboard for a short using the "beep" mode on the multi-meter. If yours doesn't have such a mode, just use the resistance check mode to view the resistenance. You might also want to check the SMD caps. In my experience, the bad cap was a straight short, but I suspect it may have been a few ohms at some point.

As others have mentioned, some motherboards are fussy about RAM, even for the same configurations. Not all 32 MB double-side FPM sticks of RAM worked well in my board. If you have other FPM sticks, play around with this some.

Well, did you locate a short and try a few different RAM modules? And is it confirmed that the hissing only occurs when attached to the SV2GX4?

Hey, I'm sorry for delay.

I have checked all the lines for short, but the resistance seemed normal - never less than 10 ohms.

I have tested all the capacitators as well ("beep" mode), but there was no short either.

After reconnecting the PSU, the hissing sound is gone, but the board remains dead 🙁

I decided to take to motherboard to someone who will be able to examine it further.

ok, my next step would be to check the BIOS.

We will. I have found some unused 386 bios (UV erasable) in my drawer - as I don't have any eeproms, I will attempt to use this one. Will report back.

How are you going to erase them? Sticking it under a grow-op light? I have a bunch I'd like to erase. Perhaps there is some cheap eBay UV light which will do it?