Maybe your voltage regulator is broken and when jumpered for 3v the CPU is getting no power at all?

Have you checked wether the board actually supplies 3.x volts to the CPU?
I think it is normal for both CPUs to work at 5V for some time and not necessarily fail right away. There are even 486 CPUs meant to run at 3V that are 5V tolerant. Maybe the Ti is one of those?

Just read another post mentioning broken regulators. Will definitely check the boards now.

As I recall the dx4 is supposed to work at 3,3v and obviously the Ti (Cyrix) at 3,45v. These are two different voltages and some problems with the Ti could occur if the board delivers 3,30v (or slightly less). My own and 5x86 is rated 3,45v and also refused with 3,3v but goes perfectly with 3,45v setting.

But if that was your problem, the dx4 should work too. Unless that CPU isn't supported by the BIOS, and that messes with things. But that's unlikely to me. The vrm sounds like the best bet.

You were right guys, this bad boy was the culprit. You can even see the tiny solder blob on the heatspreader.
After replacing it the board behaved like it should and the CPUs worked like a charm. I hope they didn't suffer too much damage.
You are the best, thanks!

High fives all around! 😁

Great job! Another board saved 😁

And as for the CPUs: I suppose the short time these had 5v won't have to much influence of they still work 😀

Nice work getting the board fixed. I highly doubt your cpus got damaged in anyway buy the short test

Thank you! I still can't believe that they survived a 50% overvoltage. I'll check all my 486 boards from now on.

The vastly larger lithography of these older CPUs is way more robust and durable than the CPUs we have today.