Gavrilo wrote on 2025-11-08, 11:10:

... Doesn't matter because IT IS WORKING! ...

Congrats, good to hear!

So for people who mentioned that it will probably not work with this 25MHz processor or with ordinary memory modules, you were fortunately wrong 😀 ...

Now that you have BIOS access... is there anything within BIOS that indicates what processor speed it is running?

Looking at the crystal freq. I do suspect it's going at 33mhz ... not at all surprising that 25mhz processor would come up.

If it is 33mhz, technically it's overclocked (by quite a bit). CPU speeds were often set by manufacturers through a final chip test.
The ones exhibiting problems at higher speeds but worked well at lower speeds were designated lower speed ... but there is always
a "safety margin" - I don't know what numbers were actually being used at the time, but typically a 33mhz processor had to test
good at 36 or even 38 ... so you 25mhz part may have just failed to run reliably at 38mhz.

Keep it properly cooled, with good solid power rails, and you may never have a problem.
If you do have reliability problems at some point, look within BIOS and see what options there are to
lower the speed, add wait states etc...

Thanks, I still have 4 more, so only 4 left. The question is why I have 3968, not 4096kB of RAM. There is 128kB loose somewhere.

PC@LIVE wrote on 2025-11-08, 12:03:

Excellent final result, congratulations 🍾
Now, however, given the number of empty banks, I would look for RAMs identical to the ones you have, to fill all the empty banks, the total memory would go from 4 MB to 16 MB, I don't know if they are easy to find, but maybe if you are lucky 🍀, you could find others similar ones, that is 2+1 chip, maybe the same speed, you need another 12 pieces of 1 MB.

This is very good point, will check that out. I don't wanna loose this processor. Will touch if it is hot and figure out some cooler for it. There is one passive somewhere in my box.

DaveDDS wrote on 2025-11-08, 12:11:

Congrats, good to hear! […]
Show full quote

Congrats, good to hear!

Now that you have BIOS access... is there anything within BIOS that indicates what processor speed it is running?

Looking at the crystal freq. I do suspect it's going at 33mhz ... not at all surprising that 25mhz processor would come up.

If it is 33mhz, technically it's overclocked (by quite a bit). CPU speeds were often set by manufacturers through a final chip test.
The ones exhibiting problems at higher speeds but worked well at lower speeds were designated lower speed ... but there is always
a "safety margin" - I don't know what numbers were actually being used at the time, but typically a 33mhz processor had to test
good at 36 or even 38 ... so you 25mhz part may have just failed to run reliably at 38mhz.

Keep it properly cooled, with good solid power rails, and you may never have a problem.
If you do have reliability problems at some point, look within BIOS and see what options there are to
lower the speed, add wait states etc...

Gavrilo wrote on 2025-11-08, 15:10:

Thanks, I still have 4 more, so only 4 left. The question is why I have 3968, not 4096kB of RAM. There is 128kB loose somewhere.

at least 64K is likely used for shadowing the bios rom, the other 64k is likely used to shadow the 32k video rom.

Gavrilo wrote on 2025-11-08, 15:17:

This is very good point, will check that out. I don't wanna loose this processor. Will touch if it is hot and figure out some cooler for it. There is one passive somewhere in my box.

DaveDDS wrote on 2025-11-08, 12:11:

Congrats, good to hear! […]
Show full quote

Congrats, good to hear!

Now that you have BIOS access... is there anything within BIOS that indicates what processor speed it is running?

Looking at the crystal freq. I do suspect it's going at 33mhz ... not at all surprising that 25mhz processor would come up.

If it is 33mhz, technically it's overclocked (by quite a bit). CPU speeds were often set by manufacturers through a final chip test.
The ones exhibiting problems at higher speeds but worked well at lower speeds were designated lower speed ... but there is always
a "safety margin" - I don't know what numbers were actually being used at the time, but typically a 33mhz processor had to test
good at 36 or even 38 ... so you 25mhz part may have just failed to run reliably at 38mhz.

Keep it properly cooled, with good solid power rails, and you may never have a problem.
If you do have reliability problems at some point, look within BIOS and see what options there are to
lower the speed, add wait states etc...

So for the RAM, if you will use Windows 3. X, you can add the other 4 MB you have, but I recommend using them, only if they are of the same type and speed, for example 2+1 chips of -70 ns.

Regarding the CPU speed, you can check through the CHKCPU program, you can find it together with others, in Phil's benches.
In my opinion, it should work at 25 MHz, because there is the 50 MHz quartz (next to the 66).

Can I turn the shadowing off without causing some "damage"?

Another problem might be with the Yuasa battery. I took it from another board I have, it wasn't even leaked after all those years, which is strange (Varta wouldn't last, it would be leaked already). So when I measured the charging voltage while the system was on, multimeter showed above 8V and it was still rising slowly about 0.01V. Can it explode or something?

Gavrilo wrote on 2025-11-08, 16:05:

Can I turn the shadowing off without causing some "damage"?

Another problem might be with the Yuasa battery. I took it from another board I have, it wasn't even leaked after all those years, which is strange (Varta wouldn't last, it would be leaked already). So when I measured the charging voltage while the system was on, multimeter showed above 8V and it was still rising slowly about 0.01V. Can it explode or something?

Check the bios for shadowing options. the "damage" you'll have is a slightly slower speed when turned off.

The battery voltage seems high to me. 5V I could accept, but 8? seems high. As long as it doesn't burst or starts leaking, or worst case goes nuclear and causes an international incident I woul'nd worry too much about it.

Please don't take me wrong but LoL. Thank you.

weedeewee wrote on 2025-11-08, 16:14:

As long as it doesn't burst or starts leaking, or worst case goes nuclear and causes an international incident I woul'nd worry too much about it.

Gavrilo wrote on 2025-11-08, 16:41:

Please don't take me wrong but LoL. Thank you.

weedeewee wrote on 2025-11-08, 16:14:

As long as it doesn't burst or starts leaking, or worst case goes nuclear and causes an international incident I woul'nd worry too much about it.

Thanks.
But Seriously, 8V on the battery terminals is high, though I have no idea why that would be.

The charging stopped at 8.63V.

DaveDDS wrote on 2025-11-08, 12:11:

Now that you have BIOS access... is there anything within BIOS that indicates what processor speed it is running?

Can't see any information about processor frequency.

About "lower the speed, add wait states etc..." take a look at the picture.

PC@LIVE wrote on 2025-11-08, 15:49:

Regarding the CPU speed, you can check through the CHKCPU program, you can find it together with others, in Phil's benches.
In my opinion, it should work at 25 MHz, because there is the 50 MHz quartz (next to the 66).

Will check that program. So it is really possible that this board has two oscillators? In case there are different processors?

weedeewee wrote on 2025-11-08, 15:46:

at least 64K is likely used for shadowing the bios rom, the other 64k is likely used to shadow the 32k video rom.

There isn't any option like this in BIOS setup. Please see above picture.

Check your multimeter. I think the battery would have exploded by now if it was really over double the expected voltage when fully charged (probably around 4V for a 3.6V NiCd), and cheap multimeters will show high voltage readings when the battery is low. There's also another funny effect where they will show high readings if you set them to AC volts and measure DC (for example, showing 20V AC from a 9V battery).

Those Yuasa batteries seem to leak less often, but I wouldn't keep an old battery like that soldered to the board. I would rather use a brand-new soldered battery (you can still buy them, and they'll probably last long enough for retro-computing purposes) or a remote battery holder.

Gavrilo wrote on 2025-11-08, 11:10:

So for people who mentioned that it will probably not work with this 25MHz processor or with ordinary memory modules, you were fortunately wrong 😀. Thanks to everyone. Now I just have to get appropriate PC case, because this board is to big for ordinary one.

PC@LIVE said in first post that there is another oscillator of 50MHz, next to that one we all saw, so, that was not the problem anymore, especially after You said that it gives beeps without memory 😀

Glad You managed to sort it out with battery jumper!!!

But check those charging voltages again!!!

Gavrilo wrote on 2025-11-08, 17:45:

PC@LIVE wrote on 2025-11-08, 15:49:

Regarding the CPU speed, you can check through the CHKCPU program, you can find it together with others, in Phil's benches.
In my opinion, it should work at 25 MHz, because there is the 50 MHz quartz (next to the 66).

Will check that program. So it is really possible that this board has two oscillators? In case there are different processors?

But in my opinion, the other quartz could be used for something else, maybe to make the memories work asynchronously, that is, 25 MHz CPU and 33 MHz memories, but you should see if there is something about it on the manual, maybe from the line at the bottom of the BIOS screen, you could understand the brand and model of your motherboard.
Or you could download the HWINFO program, you can find it here on Vogons, maybe it could give you a lot of useful information, in addition to the CPU frequency, MB brand and model, and other info.

When using one of the benchmark programs to figure out the clockspeed the cpu is running at, be sure to add a jumper to the turbo pins on the motherboard to see what effect toggling the turbo mode on/off gives. Sometimes it switches between oscillators, on other mainboards it switches cache off, ...

enjoy.

lti wrote on 2025-11-08, 18:12:

Check your multimeter. I think the battery would have exploded by now if it was really over double the expected voltage when fully charged (probably around 4V for a 3.6V NiCd), and cheap multimeters will show high voltage readings when the battery is low. There's also another funny effect where they will show high readings if you set them to AC volts and measure DC (for example, showing 20V AC from a 9V battery).

Those Yuasa batteries seem to leak less often, but I wouldn't keep an old battery like that soldered to the board. I would rather use a brand-new soldered battery (you can still buy them, and they'll probably last long enough for retro-computing purposes) or a remote battery holder.

Thanks Iti, but now you kinda insulted me 😀, I know how to use multimeter, I wouldn't set AC scale and measure DC. I confess I am sometimes stupid and I mean it, seriously, but am not that stupid. I forgive you 😉, no hard feelings. The multimeter is cheap, but it is working properly and measures good. It has proper 9V battery, not like the newer cheap ones with 2xAAA.

About the battery, there are on the market but not NiCd anymore, only NiMH and they have different characteristics. Remote holder is definitely an option. Or something like this:
https://aukro.sk/bezvarta-nahrada-nicd-cmos-a … h-pc-7066135067. Anyway, I will remove this old one from the board, it's dead anyway, the voltage dropped below 1V.

Gavrilo wrote on 2025-11-08, 17:40:

... About "lower the speed, add wait states etc..." ...

Just to be clear, as long as it's working, I wouldn't change anything!

If it really is 33mhz, a 25mhz CPU may overclock without problems.

But if you do eventually have speed related problems, lowering speed would (obviously) help, but increasing wait states can also make a difference.
Just something to keep in mind if it turns out things aren't perfect yet.

IIRC there is software you can use to tell the actual processor speed quite accurately.

Yes, but if there is no jumper on turbo switch header or no switch set to OFF, the processor should run at 8MHz only, right? With jumper or switch set to position ON the turbo should activate 33MHz, even if there is 25MHz processor? What is not clear to me yet is the function of the oscillator, why it is 66MHz. The oscillator gives the processor 33MHz while turbo is ON?

I have 486DX2 66 MHz with turbo switch connected to the header to position 2,3 according manual but it does not make any difference. It is because the cpu frequency is fixed with appropriate jumper?