I had once similar problem. As far as I remember, it didn't show (some?) digits in other position. I had to remove 7-segment display to find a burned track under. I guess someone had connected something in the wrong way or to wrong connector.. I don't know. Anyway, it works correctly now.

I believe it doesn't need motherboard connection to work properly. I have that same case somewhere, don't remember has it the same display though. This looks quite the same, but maybe your version is with additional config for 'hundreds'? http://www.minuszerodegrees.net/led_speed_dis … 0-%20GT-569.gif

Thanks for the feedback, I'll try and get the display off and test the relevant traces for connectivity with my multimeter.

I'm going to need to edit the title of this topic, as I think this is down to a faulty display.
I can tell the display elements are all working, and that the switch itself is working, it's just not handling the switch state, somewhere in the display.
I will take it to my friend to have it looked at, and failing that, guess I will look for another display on ebay.

Shouldn't need any signal from the motherboard. The turbo switch in one position should short white-orange, and in the other position short white-black. On the display PCB, the turbo ground and power ground should be connected. The jumpers then connect each display LED element either directly to ground (so LED on all the time), or to the 'N' pin (so LED on if 'N' is shorted to 'G', via the turbo switch), or to the 'T' pin (on if turbo switch shorts 'T' to 'G').

It looks like it's a 2.5 digit, so can show '1' for the hundreds. So it's likely that the bottom row of pins is for the tens and hundreds and the top row for the units. Bottom row you can make a reasonable guess about how it's connected up, but the top row has a couple of possible layouts. I've had a go a sketching out what I think are the possibilities, showing the two options for the Units digit. I've also guessed which way around the 'T' and 'N' positions are, but that doesn't matter too much. The important bit is that if you want a segment to light up (either always, or just in normal mode, or just in turbo mode) then there should be a jumper on the lettered pin.

Thanks so much, Snufkin, turned out it was a simple as that! I had misread the top set of jumpers upside down, and the top most layout in your diagram was correct in this case.

I feel dumb for getting it wrong, but I'm happy now that I can have this entirely useless display working for when I switch processors 😀

Nice. It's 1 faster. "Put your helmet on; we'll be reaching speeds of 3".

It's not as if there are any silkscreen markings to give a clue, so at the point (without tracing PCB traces) you've got a 50/50 guess, and if you guess wrong then the side effect is that either 'T' or 'N' gets shorted to 'G' through a misplaced jumper, rather than the switch. Which means it behaves like the switch doesn't work.

RockstarRunner wrote on 2022-01-03, 14:36:

but I'm happy now that I can have this entirely useless display working for when I switch processors 😀

If you are feeling adventurous you could hook up the second turbobutton wire to the FS2 pins on the frequency selector jumperblock on the motherboard.
Via reconfiguring the turbo led you can toggle between fsb 66MHz and fsb 100MHz. Just don't do that when the machine is running.