Maybe this

S - Turbo Switch ground
T - Turbo Led +
TH - Turbo Switch High
TL - Turbo Switch Low

Or

S - Turbo Led +
T - Turbo Switch ground
TH - Turbo Switch High
TL - Turbo Switch Low

🤔

To be sure one would need to check where these connections go to 7406 inverter and beyond.

Is there a need for a universal turbo display replacement? might be able to whip something up..

I'm having a vague recollection that someone may have already though.

Sphere478 wrote on 2022-08-18, 06:36:

Is there a need for a universal turbo display replacement? might be able to whip something up..

I'm having a vague recollection that someone may have already though.

Fabricating a New Turbo Display

Seems like it 😀

Ok got it figured out. I read somewhere here that the switching is done typically by branching off the mobo's Turbo LED ground connector. So I connected the S connector of the turbo display to a branch of the turbo LED ground line, and the display was correctly switching numbers according to the mobo's turbo state, controlled by the turbo switch button.

Nice, that's two things solved. So TH,T,TL aren't needed in my setup. Though would have been nice to learn what they're for.

The actual LED digit setup is done by the big collection of jumpers. Each digit cell is represented by 4 pins laid out like interlocking T and ⊥. If no jumpers are used, the cell won't light at all. A jumper across the vertical line of the T will light the cell in both turbo on and off. A jumper on the right wing of the T will light the cell when turbo is on. Jumpering the left wing will light the cell only when turbo is off.

Also, this display uses one shared current limiting resistor, which made it a bit harder to try out lighting cells in isolation. If only one single cell is lit across the whole array, it becomes orange instead of green. I guess if you don't disconnect it quickly it will burn. So be careful. You'll need to light up more cells to avoid overcurrent.

With that figured out, I'll need to know what my 486 DX2-66MHz will slow down to when turbo is off, because you'll need to plan both turbo states together when deciding what to jumper.

What tool can give me a good approximation of Mhz when turbo is on vs off?

Cheers!

Maybe TH,T,TL are outputs for a Turbo LED since the motherboard's LED output is being used to switch the display? Might be able to connect a bi-colour LED to change colours depending on turbo selection.

If your board has a clock generator chip (it'll be a chip next to a 14.318 crystal) then you might be able to look up exactly what the frequency outputs are.

wbahnassi wrote on 2022-08-18, 15:48:

What tool can give me a good approximation of Mhz when turbo is on vs off?

Tools like speedsys, NSSI, or Norton SI can measure clock speed.

The result might be surprising, because most early 486 mainboard do not change processor clock at all in deturbo mode, but instead disable memory caches to slow down the system. Varying the processor clock is only available on later "green" mainboards with power management that usually include a "doze mode".

snufkin wrote on 2022-08-18, 17:14:

Maybe TH,T,TL are outputs for a Turbo LED since the motherboard's LED output is being used to switch the display? Might be able to connect a bi-colour LED to change colours depending on turbo selection.

YES! That's it. As a test, I took the S line from the display and connected it to the mobo's Turbo LED ground. And plugged the actual Turbo LED from the chassis to T and TH on the display. And indeed, the Turno LED now gets its state from the display, and you can reverse the state by using T and TL instead of T and TH. So yes, it can be used for a dual color LED too, but my case only has a single color LED. But yeah, very nice! And I don't need a breakout wire now from the mobo's Turbo LED connector. 👍

mkarcher wrote on 2022-08-18, 17:46:

Tools like speedsys, NSSI, or Norton SI can measure clock speed.

The result might be surprising, because most early 486 mainboard do not change processor clock at all in deturbo mode, but instead disable memory caches to slow down the system.

Right, I guess what matters is how much MHz the deturbo is effectively simulating for me. Whether it's via cache disables or clock frequency change is not important here.

I'll give a try to the tools suggested. Thanks!