Have you tried populating the two RAM sockets nearest to the PCI-slots?

The chipset probably doesn´t support EDO, so make sure you´re using FFM RAM.

@majestyk:
Yes, I tried that. Do you think it's RAM problem? CPU is cold which is odd.

EDO should be supported according to trw, edo also mentioned in manual.
cpu is working since you get a beep. no cpu would be no beep.
two identical memory modules seem to be in first two simm slots (bank 0) according to manual found on trw

Can you try with other memory and/or other cpu ?

Sadly, I don't have another CPU, I will try to find other RAM.

FIC mainboards are a bit unpredictable when it comes to RAM, I have some models where the chipset does support EDO, but you need to use FPM - probably they didn´t bother adapting BIOS.
I have some FIC boards from that era that only support very few RAM sticks, so it´s worth trying some alternatives. The RAM slots should be investigated carefully for corrosion, dirt and bent contacts. And I think these long beeps indicate RAM (or maybe cache) issues...

I've got another SIMM memory but it's still beeping. Can this be a CPU problem?

Should this be empty?

informatyk wrote on 2022-11-02, 13:56:

20221102_144850.jpg 20221102_144902.jpg I've got another SIMM memory but it's still beeping. Can this be a CPU problem? […]
Show full quote

20221102_144850.jpg
20221102_144902.jpg
I've got another SIMM memory but it's still beeping. Can this be a CPU problem?

Should this be empty?
20221102_144920.jpg

Yes. The 8 chips are 256kb cache (32 KB each), the single chip below is 32 KB of TAG RAM. That's a common configuration. Did you check the motherboard manual to see if your memory configuration is supported (SIMM memory size, number and position of modules)? Is this a single 8 MB module? According to the manual, this is not supported. You would need to use two modules per bank.

informatyk wrote on 2022-11-02, 13:56:

20221102_144850.jpg 20221102_144902.jpg I've got another SIMM memory but it's still beeping. Can this be a CPU problem? […]
Show full quote

20221102_144850.jpg
20221102_144902.jpg
I've got another SIMM memory but it's still beeping. Can this be a CPU problem?

Should this be empty?
20221102_144920.jpg

Hello,

yes, the one socket is also empty in the manual on retroweb.
But what is really weird, that the data srams are 20ns only. I have never seen 20ns ones on a Pentium mainboard.
I think that will only work with 50MHz, it will not work on 60 or 66MHz.
Since you have a Pentium 75 (50 * 1.5) it should work.
Did you check, if 50MHz and multiplier 1.5 is configured correctly?
Check that. If both was correct, you could try removing all cache chips, then this cannot be the problem anymore.

Did you check, if 50MHz and multiplier 1.5 is configured correctly?

How to check that?

Are jumpers correct?

Jumpers settings from manual.

The jumpering for cache voltage is wrong.
You have 5V (Toshiba) SRAMS, so the jumpering mustn´t be the one for 3.3V but for "Mixed Mode":

JW 1,2,3,4 must be closed, JW 5,6,7,8 must be open.

(At the moment you have JW 5,6,7,8 closed so the SRAMS are at 3.3V which will at least leave them unstable.)

Sounds like a Memory module issue.
She does not like it so she is Whining.

@majestyk: Where do you find out that memory are 5v? I changed jumpers to JW1, JW2, JW3, JW4 and still the same sound. Tried bank 7,8 and 1,2.

majestyk wrote on 2022-11-13, 12:08:

The jumpering for cache voltage is wrong. You have 5V (Toshiba) SRAMS, so the jumpering mustn´t be the one for 3.3V but for "Mix […]
Show full quote

The jumpering for cache voltage is wrong.
You have 5V (Toshiba) SRAMS, so the jumpering mustn´t be the one for 3.3V but for "Mixed Mode":

cachemix.JPG

JW 1,2,3,4 must be closed, JW 5,6,7,8 must be open.

cachemix2.JPG

(At the moment you have JW 5,6,7,8 closed so the SRAMS are at 3.3V which will at least leave them unstable.)

tc55328a is 5v.
tc55v328p is 3v3.

informatyk wrote on 2022-11-13, 14:44:

@majestyk: Where do you find out that memory are 5v? I changed jumpers to JW1, JW2, JW3, JW4 and still the same sound. Tried bank 7,8 and 1,2.

Leave those jumpers as they were set. They were correct at 3v3.

weedeewee wrote on 2022-11-13, 15:20:

Leave those jumpers as they were set. They were correct at 3v3.

Did I break cache by changing jumpers?

informatyk wrote on 2022-11-13, 14:44:

@majestyk: Where do you find out that memory are 5v? I changed jumpers to JW1, JW2, JW3, JW4 and still the same sound. Tried bank 7,8 and 1,2.

Here´s a datasheet for the TC55V328BJ:

https://tvsat.com.pl/PDF/T/TC55V328_tos.pdf

It´s 3.3V but was never made in a - stone old - 20nS Version.

Then there´s the TC55V328AP:

https://datasheetspdf.com/pdf-file/1092275/To … /TC55328AP-25/1

It´s 5V and it was made in 20nS and slower.

Now I turn the tables and ask you "Where do you find out that mamory TC55V328BP are 3.3V?"

They might be but I can´t remember ever seeing 20nS DIL SRAM that was not 5V.

If there´s a convention that "A" or "B" or "J" or "P" denote the voltage the case would be clear. Maybe someone has a source for that.

majestyk wrote on 2022-11-13, 17:58:

Now I turn the tables and ask you "Where do you find out that mamory TC55V328BP are 3.3V?"

It's listed in the " American Megatrends Apollo Pentium PCI ISA motherboard " manual, page 77, section 6 upgrading cache memory, found on manualzz website.

edit: google hit "American Megatrends Apollo S728 User's Guide"
edit2: also you're assuming that since the 20ns aren't listed in the datasheet for the tc55v328 that the chips might be relabeled and are actually tc55328.
anyway op tried them on 5v, it didn't change the behaviour and now the question is, if changing it to 5v might have damaged the 3v3 cache .

Damn it - I must admit I havent´t read that manual for a while 😉

Edit: TO is encouraged to PM his Paypal-address in this case 😀
(...or call the police of course)

By the way for the 3.3V Toshiba it´s:

So 5V for a short period is not the end of the world.

After all this is not war - everybody is trying to help here and find the needed information (at least I thought so).