Looks similar to the types of backlight used in synths, which I'm pretty sure are electro luminescent (EL) displays, here's a short video of someone replacing one: https://www.youtube.com/watch?v=HGoZCSdIAvA

If it is an EL panel then be careful if working on it when it's powered up, they run at around 100V AC.

I guess the EL panel is connected via those two solder pads next to the 'select' button. If it's attached in a similar way to that video then maybe it's a case of removing the three buttons, unsoldering those two pads and then sliding the EL panel out. If that all works then you can measure the panel size and get something similar to fit.

If you've got a multimeter then it might be worth, carefully, trying to measure the AC voltage between those two pads. Maybe the panel itself is ok and the inverter has stopped working.

EL requires high voltage dc-dc converter to light up and is kind of blue-green glow. Yours does not have this feature, meaning it has simple LED backlight behind the LCD panel.

But you have big problem; need to de solder the panel from the circuit board to check and repair the backlight.

Cheers,

>display was really dim

how about a picture showing that?
unplug the thing and take good quality pictures of pcb at right angle so all the markings can be read

Between the display and the buttons there are two solder points which seem to be on a different pcb, likely the backlight.
What is the voltage on these ?

I'm guessing this has some led backlight and the leds are worn out.
It might be possible to replace the leds. even if they are of the kind were the leds are wirebonded to the pcb (like the black blob chips one often sees)
Scrape the wirebonded leds off and replace them with tiny smd leds.

pentiumspeed wrote on 2022-12-10, 17:45:

is kind of blue-green glow.

Blue-green a bit like this maybe (picture of a different guru panel)?

(from https://hexus.net/tech/news/peripherals/9026- … man-goes-crazy/ )

Think I've also seen different orange EL displays. And these people https://elpanelandtape.co.uk/el-backlights/ list: White, Light Blue, Green, Pink, Red, Yellow, Deep Blue and Orange. So I think it could still be an EL display. Compared to LEDs they have a limited number of hours, so if the PC the display came from was on a lot then it could definitely have faded. I think the thing to carefully check is the voltage (could be >100V AC, although low current), with meter on AC voltage, on those two pads next to the select button.

Thank you for all your replies! I will post more pics with the thing on later today, so you can see it in action. But just to clarify, when i said "dim" i didn't specifically talk about the backlight, but rather the LCD itself. The symbols on the display is more of a greyish tone, and can only be seen from specific angles. I tried massaging the display while it was on, which created a momentarily better connection, and the symbols on the screeen went from grey to black. So it's probably possible to jerry-rig it to apply more pressure on the contacts, but that's probably not the proper way to do it. With all that said, the backligt is also pretty dim, but worst case scenario, fixing the LCD with out fixing the backligt could suffice. But best case scenario, best to fix both.
And yes, it's probably just a LED backlight. If there was a 100V current, i'd probably already been electricuted by now. 😉 And the connections to the mobo is just regular USB-type pins.

Well, thats not good because from the pictures it looks like those long pins go straight to the glass? Pushing would fix zebra strip connected display.
Its possible negative voltage supply went weak and LCD driver doesnt have enough oompf to drive liquid crystals properly anymore.

rasz_pl wrote on 2022-12-11, 03:02:

Well, thats not good because from the pictures it looks like those long pins go straight to the glass? Pushing would fix zebra strip connected display.
Its possible negative voltage supply went weak and LCD driver doesnt have enough oompf to drive liquid crystals properly anymore.

Good thought ! I had an issue decades ago with a backlit LCD and it was an issue of supply not driving the crystals hard enough to block the light as intended.

While waiting for the testing pics, I thought i could just post som complimentary pics so you can see the whole of both PCBs. Cheers!

how does it connect to motherboard?
bad news is I dont see power supply anywhere, good news is it might be a generic LCD driver chip with integrated negative supply source? Im not super versed in LCD drivers.
Can you read exact marking on the chip directly behind LCD? Is it something like Holtek HT1623?
As an experiment you can try slightly pushing (with your finger) only the pins connecting glass to pcb and checking if there is any difference in the picture

rasz_pl wrote on 2022-12-11, 03:45:

how does it connect to motherboard? bad news is I dont see power supply anywhere, good news is it might be a generic LCD driver […]
Show full quote

how does it connect to motherboard?
bad news is I dont see power supply anywhere, good news is it might be a generic LCD driver chip with integrated negative supply source? Im not super versed in LCD drivers.
Can you read exact marking on the chip directly behind LCD?
As an experiment you can try slightly pushing (with your finger) only the pins connecting glass to pcb and checking if there is any difference in the picture

It connects through 4 wires for the "guru link", and i also had to connect one ground pin via the USB pins for the thing to start up, which is abit weird (no pun intended).

The markings on the LCD PCB back chip is as follows:
HOLTEK
HT1622
A517K0243

I actually tried pushing the pins connecting to the lcd while "massaging" the LCD, and that didn't do anything. It's possible the pushing has to be done from another angle.

Interesting. Are you using a PSU with both good -12v and -5v ? If using a newer PSU then that could part of the issue...
Why did you have to re-wire a ground lead ?

Horun wrote on 2022-12-11, 03:58:

Interesting. Are you using a PSU with both good -12v and -5v ? If using a newer PSU then that could part of the issue...
Why did you have to re-wire a ground lead ?

It's a pretty bog standard "modern" PSU. A Corsair RM850x. I do have an AT PSU to try, but the Abit AV8 uses an ATX connector, so i'd have to order an adapter first.
If i didn't plug in the usb ground, the unit was just dead. So i just thought they skipped the ground pin in the "guru link" connector to save money, seeing that most people probably would wire up all the connectors anyway.

DJonekill wrote on 2022-12-11, 04:04:

It's a pretty bog standard "modern" PSU. A Corsair RM850x. I do have an AT PSU to try, but the Abit AV8 uses an ATX connector, so i'd have to order an adapter first.
If i didn't plug in the usb ground, the unit was just dead. So i just thought they skipped the ground pin in the "guru link" connector to save money, seeing that most people probably would wire up all the connectors anyway.

If you PSU does not supply the old -5v and -12v standards for early ATX that could be part of the problem. Most new ones do not supply the -5v...
Nahh that does not make sense, what case are you using ? Nearly all cases of that era are metal so ground is everywhere. Maybe they grounded thru mount screws like HD's, floppies and cdroms ?
(yes I know they also ground thru PSU connection but also thru the case) just a thought...

Horun wrote on 2022-12-11, 04:14:

If you PSU does not supply the old -5v and -12v standards for early ATX that could be part of the problem. Most new ones do not […]
Show full quote

DJonekill wrote on 2022-12-11, 04:04:

It's a pretty bog standard "modern" PSU. A Corsair RM850x. I do have an AT PSU to try, but the Abit AV8 uses an ATX connector, so i'd have to order an adapter first.
If i didn't plug in the usb ground, the unit was just dead. So i just thought they skipped the ground pin in the "guru link" connector to save money, seeing that most people probably would wire up all the connectors anyway.

If you PSU does not supply the old -5v and -12v standards for early ATX that could be part of the problem. Most new ones do not supply the -5v...
Nahh that does not make sense, what case are you using ? Nearly all cases of that era are metal so ground is everywhere. Maybe they grounded thru mount screws like HD's, floppies and cdroms ?
(yes I know they also ground thru PSU connection but also thru the case) just a thought...

I didn't actually test it in the case before i concluded that it didn't work. But to be fair, my case "thermaltake swing" uses plastic sleds for the 5,25" bays, so it wouldn't have been grounded though the case anyway.

Do you know what to look for when trying to find an early ATX PSU? I just know about the AT/ATX difference. Maybe you have some models in mind? Thanks!

Just found the rail specs for the current PSU.

no need for negative voltage, Holtek makes its own bias supply
https://www.crystalfontz.com/controllers/Holtek/HT1622/119/
>Built-in internal resistor type bias generator

resistor r21 between pins 8 and 9, measure it with multimeter, in theory this is where one sets bias voltage for this controller, zero ohm is max contrast

rasz_pl wrote on 2022-12-11, 05:34:

no need for negative voltage, Holtek makes its own bias supply https://www.crystalfontz.com/controllers/Holtek/HT1622/119/ >Buil […]
Show full quote

no need for negative voltage, Holtek makes its own bias supply
https://www.crystalfontz.com/controllers/Holtek/HT1622/119/
>Built-in internal resistor type bias generator

resistor r21 between pins 8 and 9, measure it with multimeter, in theory this is where one sets bias voltage for this controller, zero ohm is max contrast

Cheers! So are you saying this could be a problem with the LCD control chip rather than the LCD connectors? Edit: sorry, i meant the resistor.

I shall measure the aforementioned pins when i test it!

I really dont know, It might also be lack of ground on the second pcb, maybe the second one also needs to be screwed into chassis and thats where it gets its ground from