Announcement

**jsog** · 07-11-2011, 03:28 PM

Re: Not 100% sure where I should have posted this :P

Picture #3 makes me think it has two LEDs for backlights. Both of them failing seems unlikely. They are probably fed from +5V or +12V with a current limiting resistor. R18 and R19, maybe. The pictures are blurry and it's hard to tell what's going on there.

Check the LEDs with the diode test on a meter, or a 5V voltage source with a 1K or 220 ohm resistor. Tracing the power for them backwards to a source might help. The solder on D2 looks nasty, maybe just lead-free.

And it could be a software thing, where the microcontroller on that board is able to turn off the backlights. If that's the problem, cutting traces and wiring them direct to the power supply would be one way to keep them on.

**Demix** · 07-11-2011, 03:36 PM

Re: Not 100% sure where I should have posted this :P

Thanks for the reply! Yeah, D2 did look like a bit of a terrible solder job so I re-did it just to be sure. I re-soldered any parts which looked to me like they could be a problem (also on the third picture you mentioned you think it has two LED's - this is the case, anywho one of the leads had no solder on it at all and was only just making a connection, so I soldered that properly as well. I will be honest with you, I havent actually used my multi-meter in quite some time so ive just removed the dust from it as we speak and im about to see what I can remember =] again, thank you so much for the help and I will reply with my results in a little while ^^

**Demix** · 07-11-2011, 05:43 PM

Re: Not 100% sure where I should have posted this :P

Right, I did my best with the multi-meter and if im honest I didnt gain all that much insight into whats wrong. All I know is that ive had the computer its connected to switched on now for about 20 mins and ive noticed that just around the LED's (the underside) It's getting extremely hot? (Yes I'm aware Resistors do create heat). I think this could be whats wrong though.. argh I wish I wasnt so rusty with this stuff. I'm pretty sure its R18 and R19 that are heating up. R18 is hot to touch, but R19, if i leave my finger on it, it begins to burn me hehe. It's just annoying that these LED's wont light up now.. I'm not sure why they just chose to stop working.. >_< Every other function of the PCB works. I can change the temps, set the time, the two connections that are visible - the longer one is the power - it's connected straight to a molex connector inside the PC, and the other is for the two fans + the cpu temp probe which is also inside the case.

P.S - I'm sorry for the double post, I didn't realise >_<

**jsog** · 07-11-2011, 06:03 PM

Re: Not 100% sure where I should have posted this :P

A better picture of the R18/R19 area would help.

First, check the LEDs. Disconnect the board from the computer and remove it. Put the meter on a "diode check" setting. Touch the probes to the leads of one of the LEDs, then reverse the probes and do it again. One way it should read nothing, the other it should show a voltage. The LED might light up dimly, depending on the type. Then check the other LED.

If the LEDs are good, then I would switch the meter to continuity and trace back from the LEDS to the resistors and then where the resistors are getting their power.

**Demix** · 07-11-2011, 06:33 PM

Re: Not 100% sure where I should have posted this :P

Well, one of the LED's gives me 156 one way and 119 the other, and the other LED is showing 41 one way and 40 the other. This is while they are in the circuit though. I'm not sure if fully removing them to test would be better. The one showing 40ish both ways is the one above the really hot R19 though. Also as requested, I've attached a few better pictures. Sorry, my phone seems to be terrible today. It just wont focus! lol

Also I've just done a continuity test and the LED's both come up as good, but then the resistors are 0.1 (im not sure if thats okay?) and then then carrying on, the smaller resisters are something like 0.18, etc.

Attached Files

**jsog** · 07-11-2011, 07:58 PM

Re: Not 100% sure where I should have posted this :P

Much better pictures! The blotchiness is due to low light, but they're good enough for now.

Here's an enhanced marked version:

Don't know why there are different resistor values to the two LEDs. Incoming power (yellow) comes from one side of D2, then goes through a zener diode. Then it splits through two resistors to the anode of each LED (purple). (Why are the resistors different values? Variable colors?)

At this point, it might be easiest just to cut the leads to the LEDs close to the board, then solder wires to those leads, add a resistor and some heatshrink and wire it straight to the incoming power. There are open VCC and GND holes in that empty connector location on the board.

Attached Files

**jsog** · 07-11-2011, 08:00 PM

Re: Not 100% sure where I should have posted this :P

Somehow I made a duplicate post, please ignore this one.

**Demix** · 07-11-2011, 08:49 PM

Re: Not 100% sure where I should have posted this :P

Hehe xD It happens

And thank you again for your help! I think both LED's are actually clear o.O I was looking at the LCD and its tinted blue, so the white light goes through it. hmm, I see what you mean, that would be a lot easier! as i have no wire or resistors for that matter, im going to take a trip to maplins in the morning and pick some up. Sorry to be a pain, but what resistance value do you think would be suitable for this? - Also I think I may pick up a new pair of LED's while im there as this case is around 3 years old, and when it was first in use (It belonged to my brother) the thing never worked for him at all. It's only when I decided to hijack his old case and mess around with it that the LED's kicked in in the first place, so its a likely bet that this case may have been used for display purposes or something so the LED's have seen some heavy use in the past - that and the fact that my brother blew his old computer up at least 3 times with various issues. He kept frying his PSU's lol

**jsog** · 07-11-2011, 09:29 PM

Re: Not 100% sure where I should have posted this :P

LEDs rarely wear out. They'll fail if they get too much current. The resistor limits the current. The value of the resistor depends on the supply voltage and the current the LED needs. Red LEDs are the most efficient (lowest current needed), blue and white are the least efficient. There are LED resistor calculators on the web.

**Demix** · 07-12-2011, 09:21 AM

Re: Not 100% sure where I should have posted this :P

Right I took it to maplins and one of the guys took a quick look for me. He confirmed the LED's were fine like you said they would be, and he said the resistors were okay, but when it came to D2 he said he couldnt get a clear reading and it kept flaking out. He gave me a replacement Diode to try and said I should start there. I took D2 out and out of curiosity i powered the panel up, it came up as before with the data on the screen etc but no backlight, so im defo on the right track here. I did something probably viewed as pretty stupid but I couldnt think of another way to test it. I got a bit of wire just to bridge the gap where the diode was (I know the diode is there for a reason) but anyways I got nothing from that.. so now i have this vaguely similar diode that im going to try and solder on and hope for the best. any other suggestions? - oh btw even without the diode there, the Resistors near the LED's are still getting quite hot o.O I didnt think they were currently even receiving any power. Thanks - Ryan

Edit:

I just tried the new diode and im still getting nothing, also its odd. when i touch the multi meter from the far end of D2 to the zener diode, the multi meter goes off instead of showing a reading. I really wish I could get to the bottom of this.

**jsog** · 07-12-2011, 09:43 AM

Re: Not 100% sure where I should have posted this :P

Can't really tell what D2 is doing. It may just be to keep the battery from being charged by the power supply. Don't jumper stuff with wires--the designers were shooting for absolute lowest cost, and they still put a diode there.

The R18/R19 area should only warm up if the LEDs are on, and not an awful lot then. The LEDs on the other side should be the only relatively high-current thing there. And still shouldn't get hot, maybe just warm. Which seems to say the LEDs are on. There may be some failure mode where they've shorted internally, and the resistor keeps them from melting. A normal LED with no resistor gets very bright and then goes out when the internal connection melts. (And the plastic case usually goes BINK! and cracks. Uh... or so I've heard...)

If it were me, I'd separate the leads and test them with a separate power supply. Looks like they're molded into that block of plastic for the backlight. Depending on motivation, new LEDs could be taped/glued in place, or holes could be drilled. Results vary on that optical plastic and drilling into things next to fragile electronics.

**Demix** · 07-12-2011, 10:01 AM

Re: Not 100% sure where I should have posted this :P

It's either down to the LED's or the Zener Diode now then I guess. I really cant see it being anything else. Also those LED's come out easily. they aren't molded, they simply sit in small holes already made for them. I lifted them out (bending them straight) to get a better look at them. My brother did manage to short his PC several times as i said, so it may be the case. I also read that this can happen when a short occurs. The zener diode will go, messing the entire thing up.

Also, i may try disconnecting the LED's and test them properly, but would a 6 volt big rectangle battery suffice? It's the only real source I have at the moment, other than just tiny batteries, AA etc lol

Just tested the LED's outside of circuit using diode test and nada.. these babys are dead i think D:

one of them gives 350ish one way and 150 reversed, and the other gives a complete blank.. is it possible that 1 of them went and caused the other to stop working?

**jsog** · 07-12-2011, 12:18 PM

Re: Not 100% sure where I should have posted this :P

I thought the Maplins guy had tested the LEDs. Since they're removable, that makes this a lot easier. If you have access to old scrap computers, the front panel LEDs already have wires attached. Usually red, green, or yellow.

A 6V battery of any size is fine for testing. Use a resistor in series with the LED, anything from 220 to 1K. If the LED is very bright, don't leave it connected long.

The zener could be burnt out. Check the voltage between the GND hole in the missing connector and the orange side of the zener.

**Demix** · 07-13-2011, 07:06 AM

Re: Not 100% sure where I should have posted this :P

Right well I tried what you said and tested from the ground to the orange end of the Zener and for a split second it gives me something like - 1.977 then it just goes to -1. This happened for both ends of the Zener. Also i tried it the other way around and got no result back at all. This is really baffling me haha >_<

Edit: I'm an idiot lol, voltage from ground to orange end = 2.35, however voltage from ground to the other end of the Zener is also 2.35

**jsog** · 07-13-2011, 09:19 AM

Re: Not 100% sure where I should have posted this :P

2.35V isn't enough to light a white LED. With power off, you can use the meter's continuity function to trace back from the black side of the zener to find out where the power is originating.

Me, I'd take 12V from a female drive Molex connector like used for addon fans, go through a 100 or 150 ohm 1/2 watt resistor and the two LEDs in series. Maybe use a potentiometer for brightness adjustment.

**Demix** · 07-13-2011, 04:29 PM

Re: Not 100% sure where I should have posted this :P

Woop! Thank you so much for all your help, jsod =] I stuck some new LED's in from maplins (just two random ones) and the results can be seen here ^^ I think they were in series and one of them went D: all of this just for a shoddy LED! Thanks again ^^

hmm I just realised I seem to have a final problem. The battery which is supposed to be only in use to save the time, seems to be powering the LED's a little when the power to the pc goes off =[ im thinking this may be because i replaced D2 with an equiv which had higher resistance o.O

Attached Files

**Th3_uN1Qu3** · 07-13-2011, 04:37 PM

Re: Not 100% sure where I should have posted this :P

It's not coming from the battery. There's probably a little leakage current somewhere, high brightness LEDs will emit a small amount of light at ridiculously small currents.

You can confirm it is not the battery by unplugging the power cord to the computer. In a few seconds the LEDs should fade away.

**Demix** · 07-13-2011, 04:39 PM

Re: Not 100% sure where I should have posted this :P

I'm afraid it is the battery >_< I did unplug it from the computer, and then waited.. and after a short while i took the battery out and boom, the lights went out. I put the battery back and the LED's come back on lol

**Th3_uN1Qu3** · 07-13-2011, 04:41 PM

Re: Not 100% sure where I should have posted this :P

Leave it be. It's not like they'll drain the battery or anything.

Announcement

Not 100% sure where I should have posted this :P

Not 100% sure where I should have posted this :P

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment