You won't get a video signal if the CPU is dead. Check if there are pins missing on the CPU's bottom. Could you please post a picture of the CPU and the original heatsink? I have never seen an MMX with such an enormous heatsink.

The PA-2013 is a refreshingly simple board to jumper, and the P233MMX very forgiving (it would probably boot with the P133 settings, even though it would be massively overvolted like that), so my first impression is that it might mean bad news...

Upgrading from P133 to P233MMX, the only two jumper blocks you should have touched were multiplier (from 2.0 to 1.5/3.5) and VCore (from 3.3V to 2.8V). Is that correct?

In that case, about the only setting you could have messed up to stop a good CPU booting is the voltage - if you figured the wrong end as 1-2, you'd have jumpered 3-4 instead of 7-8, giving you 2.2V, which is too low for the P233MMX to boot. In case of doubt, try jumpering both 3-4 and 7-8. That is symmetrical, so you can't go wrong, and the resulting voltage - 3.0V - is high but will not cause immediate damage.

One question though: why choose a Pentium 233MMX on a board that could easily handle a K6-2 550? You can always run the latter at 233MHz if you want that level of performance...

gompertz wrote:

I then double checked the voltage (2.8v), multiplier (3.5x), and bus speed (66x) jumpers were all correct and still no luck.

That voltage of 2.8v was core voltage, while i/o voltage was set to 3.3v, correct? The MMX cpu requires dual voltage, while the non-MMX is single voltage. Does the mainboard support dual voltage at all? If not and single voltage was set to 2.8v, nothing will happen with an MMX cpu.

derSammler wrote:

That voltage of 2.8v was core voltage, while i/o voltage was set to 3.3v, correct? The MMX cpu requires dual voltage, while the non-MMX is single voltage. Does the mainboard support dual voltage at all? If not and single voltage was set to 2.8v, nothing will happen with an MMX cpu.

PA-2013 is one of the best super 7 boards. It will rock with any socket 7 CPU except Tillamook.

Wow overwhelming replies. Strangely the voltage jumper was already set to 2.8v from the past owner with no issues. Definitely have it right jumper as the pins are numbered on the silkscreen of the mobo. The PA-2013 supports dual rail voltage and 233MMX is spec'd as supported in the manual.

I actually found another thread of a guy with a PA-2013 having the same issues with his AGP card (granted mine is PCI) - black screen, doesn't awake monitor; he said disabling "spread spectrum" in the BIOS fixed the problem. I will try this when home.

Thread: FIC PA-2013 Ver 2.1 and Voodoo3 3000 AGP

Spread spectrum is a feature designed to reduce EMI in the operating frequency (66MHz bus speed in this case) by messing around with the exact frequency. If you're indoors and using a metal case, it shouldn't matter and you can safely disable it. Then again, if you're so close to the limit at stock settings that it makes a difference is also bad news, particularly as nothing should have changed in this respect vs the P133...

dionb wrote:

One question though: why choose a Pentium 233MMX on a board that could easily handle a K6-2 550? You can always run the latter at 233MHz if you want that level of performance...

I've struggled with the idea of keeping things "old school slow" vs "performance old school", and knowing where the draw the line in the build for a socket 7. I think if I was going to do the AMD route I may as well just do a Slot 1 build and have a whole different era of machine (i.e. early 2000 build). Right now I feel mine is a good 1995-1998 build.

dionb wrote:

Spread spectrum is a feature designed to reduce EMI in the operating frequency (66MHz bus speed in this case) by messing around with the exact frequency. If you're indoors and using a metal case, it shouldn't matter and you can safely disable it. Then again, if you're so close to the limit at stock settings that it makes a difference is also bad news, particularly as nothing should have changed in this respect vs the P133...

True. I think my test is going to be disabling the spread spectrum and also changing floppy disk to be part of the boot sequence so that I can actually have an audible indication if anything is going on during boot. I put in the topic it won't post, but hell maybe it is and I just ain't seeing it. Won't be able to test until tonight however.

Do you have a POST diagnostic card? If so, what is it saying?

Otherwise, choose the most conservative settings possible, so in this case 66MHz FSB and 2.0x multiplier. If it's going to work at all, it will then.

gompertz wrote:

Wow overwhelming replies. Strangely the voltage jumper was already set to 2.8v from the past owner with no issues. Definitely have it right jumper as the pins are numbered on the silkscreen of the mobo. The PA-2013 supports dual rail voltage and 233MMX is spec'd as supported in the manual.

I actually found another thread of a guy with a PA-2013 having the same issues with his AGP card (granted mine is PCI) - black screen, doesn't awake monitor; he said disabling "spread spectrum" in the BIOS fixed the problem. I will try this when home.

Thread: FIC PA-2013 Ver 2.1 and Voodoo3 3000 AGP

That's an interesting heatsink. Must be designed for some proprietary system that uses passive cooling.
I just remembered that I had a TX97 board that worked with plain old Pentium but not MMX. It turned out to have an MOSFET detached from the board, and I reckon that MOSFET is on the dual plane power route. I couldn't remember if I tried it with a Pentium at 2.8V, but it might be worthwhile to check the integrity of the boards' components.