In my IBM 330 P75 I discovered that there is a pair of contacts with TURBO on them (Do I try to short the two pins or do I burn everything?)
. What is it for in Pentium? I thought it existed only up to 486. Second question: what is the correct amount of RAM for a P75 running DOS and WIN95 / 98?
image of turbo pin https://drive.google.com/drive/folders/12cVgy … FCczOsbc4V10pKZ
My guesses (without checking the picture):
1. Maybe some very early Pentiums did have turbo button.
2. This is more likely, maybe some very early Pentium PCs had 486 variants with the same motherboard. I have a similar 386/486 PC with a motherboard that were shipped with both CPUs. Mine ended up being a 386DX-40 but with VLB which is more of a 486 thing.
As for RAM, 16MB is probably more period correct, but I'd go for 32MB. It never hurts to have more.
I never knew about turbo on Pentium's, but just found this:
Turbo button on Pentium systems
Then guess #2 it is.
The J-Mark J-579-A (aka Jetway) that I have here has turbo button and turbo led functionality (theoretically).
http://www.win3x.org/uh19/motherboard/show/3273
It is a late socket 7 board with Apollo VP3 chipset, 83Mhz max, and AGP. I have tried the turbo 'function' with a couple of AMD cpu's (200 and up if I remember correctly) but there was no noticable difference.
Upto Socket 7 boards exist with working Turbo switches. There maybe newer as I've seen BX boards with them, but the Turbo header doesn't function. Working headers don't all work the same way, where some just seem to halve FSB, whilst others have more significant impact and reduce speeds to 386 and slower. Even managed to get a K6-III+ to below IBM XT speeds.
The Jetway J-542C SS7 board has a turbo switch connector but the manual says its non-functional.
Is this too much voodoo?
Maybe you'd want to go as high as 64MB on the RAM, but beyond that is usually pointless with socket 5/7 chipsets as their cachable area is only 64MB, so things get slower above that. However on a 430TX machine trying to run 2000s stuff on K6-2 win98, like 2003/4 era browsers etc, I took that system up to 128MB and found it was better than the HDD thrashing all the time trying to swap. But that's not what you'd typically find on games that a P75 is fast enough to play, they probably had 16MB requirements, maybe 32MB if you had a 3D card.
Turbo buttons never seemed to me to have a well defined standard behavior, some would clock CPU to exact 4.77 of original PC/XT (More common in Turbo XTs) some would clock to the 6 or 8mhz of original AT models (Seen in 286-early 486 AT class) and some would just run the CPU synchronous with whatever your ISA bus speed was set to, anywhere from 7 to 12Mhz late 386 thru 486 and up. I have not come across one in a Pentium class machine myself, but I would guess that it either runs it at 1:1 "FSB" speed, PCI speed, or ISA speed.
Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.
I had a few socket 7 boards which did have the turbo connector. And indeed it did work somewhat. If I recall these were Vtech and PV Chips.
The turbo button usually slowed down the system quite a bit, but the Pentium 133 and 166 were nowhere near a slow 486: on the Vtech board more like a Pentium 83 according to speedsys.
So I guess it may just have altered the multiplier and fsb, maybe disabled L2 cache... It seems these turbo switched are just a little black box nobody understands completely.
Oh, btw I did just short it with a jumper, no problem at all: a simple turbo button just shorts or opens the connection as well.
Stuck at 10MHz...
I have a 430VX board where the turbo header drops FSB to 50MHz.
Also a lot of Intel’s Socket 7 boards (those that used AMIBIOS, so up to 430HX) had turbo that gave something around 386dx33 — I guess those just disabled cache.
i have tried it, seems not change anithing
AlessandroB wrote on 2021-05-11, 18:16:i have tried it, seems not change anithing
Did you try rebooting after changing it? I think I read somewhere that some chipsets (sorry, it's all a bit vague) only read the status of the turbo input on startup.
snufkin wrote on 2021-05-11, 18:25:AlessandroB wrote on 2021-05-11, 18:16:i have tried it, seems not change anithing
Did you try rebooting after changing it? I think I read somewhere that some chipsets (sorry, it's all a bit vague) only read the status of the turbo input on startup.
yes, nothing happen, probably because is still a Pentium75 installed?
I remember my P100 system had a turbo button and that slowed it down a lot - most likely by flushing and disabling caches. I now have a 386 mobo that has a BIOS option to disable/ignore the turbo function, and a good thing too because it has stupid combined header without clear markings what is on what pin and defaults to slow if the button is not connected...
