Btw I have a Digital Celebris 560 board with (the same?) VLSI chipset and it has a strange quirk. It can only support "3 sides" of memory (pairs) - so its configured like yours, a pair of single and a pair of double ranked FPM SIMMs. It's EDO tolerant too but doesn't seem to benefit from EDO RAM.

AlessandroB wrote on 2022-11-22, 12:26:

[...]

Here is the pic.
Basically i not want to exceed the maximum cacheable area, i want to maximize at the level of cache. Is better to know the amount of ram cacheable because i not have all that ram, i must to buy and is not really cheeper.BUT I have discover something interesting.. if you look the pics you see some hidden/unpopulated jumper: one for the bus 60/66 (the Socket4 produced by IBM in only one and is only at 60Mhz, never produce a 66Mhz speed) and another one for the cache (256/1M) but all chip are soldered, i not want to modify this rare Computer.

That IDT 71B74 chip is an 8kx8 (8kB) tag RAM. With 32B cache lines and WT cache I'd expect that to give 64MB cacheable, with 64B cache lines 128MB. But that's a guess based on unofficial, incomplete information. Again, I'd check what it is in real life, not make assumptions.

dionb wrote on 2022-11-23, 00:40:

AlessandroB wrote on 2022-11-22, 12:26:

[...]

Here is the pic.
Basically i not want to exceed the maximum cacheable area, i want to maximize at the level of cache. Is better to know the amount of ram cacheable because i not have all that ram, i must to buy and is not really cheeper.BUT I have discover something interesting.. if you look the pics you see some hidden/unpopulated jumper: one for the bus 60/66 (the Socket4 produced by IBM in only one and is only at 60Mhz, never produce a 66Mhz speed) and another one for the cache (256/1M) but all chip are soldered, i not want to modify this rare Computer.

That IDT 71B74 chip is an 8kx8 (8kB) tag RAM. With 32B cache lines and WT cache I'd expect that to give 64MB cacheable, with 64B cache lines 128MB. But that's a guess based on unofficial, incomplete information. Again, I'd check what it is in real life, not make assumptions.

Ok thanks, what you think about the pcb between socket and cpu??? is the first time i see one.

In this mainbord if i mix 32Mb 60ns and 32MB70ns will be ok??

You tell me - never seen one either.

As for RAM speeds, 70ns is good for 60MHz operation 66MHz needs 60ns. If you don't intend modding the board for 66MHz, both should be fine.

I have exactly the same board (just the board, no riser, no case, no PSU) and cannot get it to POST (does some beeps after while). Can you check whether your board POSTs without riser installed? Would love to get this board running and maybe find rest of the parts for it someday.

Also, can you tell me please the FRU number of the riser card?

Paadam wrote on 2022-11-27, 18:06:

I have exactly the same board (just the board, no riser, no case, no PSU) and cannot get it to POST (does some beeps after while). Can you check whether your board POSTs without riser installed? Would love to get this board running and maybe find rest of the parts for it someday.

Also, can you tell me please the FRU number of the riser card?

I just finish to check: computer post correctly without riser card (as i expect) but riser card is different from the P-60 to P-75 models, can't share. PSU have FRU that differ only by the last number but.. not know...

You usually have to know the tag size to guess the maximum amount of cache-able ram -> 2^(log2(cache size) + tag bits ).
Officially these boards support up to 128MiB according to the entry in the retro web but when looking at the info you posted you may better stay with a max. of 64MiB.

AlessandroB wrote on 2022-11-22, 10:40:

PD2JK wrote on 2022-11-22, 10:38:

This looks rather rare. You might want to be more careful with the pins. One broke off.

Yes, the photo comes from another unfortunate computer I had, the pentium60 subject of the post is perfect, the CPU has never been removed since it was installed by IBM.

Is it possible to remove the CPU from that damaged carrier PCB, and then take photos of the PCB from directly above (top and bottom), ideally so the traces and resistor values can be seen. I'm wondering if there are components under the CPU. Other than that it just looks like a load of resistors, so maybe some sort of impedance mismatch between motherboard and CPU only caught late in the design? Looks like maybe Thevenin termination (had to look that one up, it's been a while), but with long stub lengths? That'd make 64 pairs of visible resistors, which is a bit of a coincidence with the 64 bit data bus, maybe.

AlessandroB wrote on 2022-11-29, 09:59:

I just finish to check: computer post correctly without riser card (as i expect) but riser card is different from the P-60 to P-75 models, can't share. PSU have FRU that differ only by the last number but.. not know...

Thank you very much. But still, can you provide me FRU of the P60 riser?

Thanks!

AlessandroB wrote on 2022-11-21, 22:57:

Doornkaat wrote on 2022-11-21, 22:13:

You'd help us by letting us know your motherboard model or at least what chipset it uses.

This one:

Is that a Compaq mainboard ?

Paadam wrote on 2022-11-29, 15:17:

AlessandroB wrote on 2022-11-29, 09:59:

I just finish to check: computer post correctly without riser card (as i expect) but riser card is different from the P-60 to P-75 models, can't share. PSU have FRU that differ only by the last number but.. not know...

Thank you very much. But still, can you provide me FRU of the P60 riser?

Thanks!

yes as soon as i can

smokeycockatiel wrote on 2022-11-29, 19:29:

AlessandroB wrote on 2022-11-21, 22:57:

Doornkaat wrote on 2022-11-21, 22:13:

You'd help us by letting us know your motherboard model or at least what chipset it uses.

This one:

Is that a Compaq mainboard ?

Come from an IBM PC330 P-60

Paadam wrote on 2022-11-29, 15:17:

AlessandroB wrote on 2022-11-29, 09:59:

I just finish to check: computer post correctly without riser card (as i expect) but riser card is different from the P-60 to P-75 models, can't share. PSU have FRU that differ only by the last number but.. not know...

Thank you very much. But still, can you provide me FRU of the P60 riser?

Thanks!

snufkin wrote on 2022-11-29, 14:34:

AlessandroB wrote on 2022-11-22, 10:40:

PD2JK wrote on 2022-11-22, 10:38:

This looks rather rare. You might want to be more careful with the pins. One broke off.

Yes, the photo comes from another unfortunate computer I had, the pentium60 subject of the post is perfect, the CPU has never been removed since it was installed by IBM.

Is it possible to remove the CPU from that damaged carrier PCB, and then take photos of the PCB from directly above (top and bottom), ideally so the traces and resistor values can be seen. I'm wondering if there are components under the CPU. Other than that it just looks like a load of resistors, so maybe some sort of impedance mismatch between motherboard and CPU only caught late in the design? Looks like maybe Thevenin termination (had to look that one up, it's been a while), but with long stub lengths? That'd make 64 pairs of visible resistors, which is a bit of a coincidence with the 64 bit data bus, maybe.

Here are some pictures for one of these CPUs:
https://www.ebay.com/itm/255800842643?

majestyk wrote on 2022-11-30, 08:23:

Here are some pictures for one of these CPUs:
https://www.ebay.com/itm/255800842643?

Looks like R163 is missing on that one. I still can't quite make out the values of the resistors, but can almost convince myself that they're a mix of 1k and 2.2k (I think I can see 102 and 222), which would seem odd for termination as that'd make the impedance they're trying to match somewhere around 700 ohms, which sounds high. From a quick skim of the pentium pinout then it does look like those resistors are on the data bus, and probably numbered by data pin (e.g. R0 and R100 connect to D0, R23 and R123 connect to D23, etc.). Maybe they wanted the bus to be held at 3.3V ish when not being driven? No idea why though. EMC compatibility?

Also wondering about that grid of pads with no surface traces. Could be for adding capacitance to the power planes that I guess are on internal layers? I think it's all a bit odd really, curious if there are other ideas/explanations.

AlessandroB wrote on 2022-11-30, 07:34:

Thanks a lot!! May I also ask you to take photo of the complete riser? Looks to be quite rare from FRU, connector should be similar to other risers though.

Paadam wrote on 2022-11-30, 16:06:

AlessandroB wrote on 2022-11-30, 07:34:

Thanks a lot!! May I also ask you to take photo of the complete riser? Looks to be quite rare from FRU, connector should be similar to other risers though.

no connector is different from all the one used on socket5/7. i think this computer was a little laboratory for IBM. unique solution, unique socket, very rare in every expect. i will post some pics of the complete riser tomorrow.

p.s. if you take a look also on Amibay. tnks

the two 16MB simms that I was missing have arrived, but the two mounted ones were incompatible together, so I did a bit of mixing with the 486DX4 which also had 16MBx4 and now they seem to work on both computers. I did some tests to see if the 64MB worked properly on the Pentium-60

Everithing work???

AlessandroB wrote on 2022-11-30, 16:21:

Paadam wrote on 2022-11-30, 16:06:

AlessandroB wrote on 2022-11-30, 07:34:

Thanks a lot!! May I also ask you to take photo of the complete riser? Looks to be quite rare from FRU, connector should be similar to other risers though.

no connector is different from all the one used on socket5/7. i think this computer was a little laboratory for IBM. unique solution, unique socket, very rare in every expect. i will post some pics of the complete riser tomorrow.

p.s. if you take a look also on Amibay. tnks

Did you get the chance to take photos of the complete riser front and back? It would help a lot when determining suitable candidates for modding (can route ISA lines to separate ISA riser without much effort to have working sound card).
I am considering your PC330-P75 in amibay but total price with shipping probably would be quite high.

Here the riser pics