Help troubleshooting a surface mount LED?

[AnubisGuard]

I have a set of 29th century tricorder electronics from gmprops that I'm waiting to permanently install (once I paint the body) and lately the third of the four green chaser LEDs on the left hand side - the ones under the "ALPHA BETA GAMMA DELTA" graphics - has intermittently stopped working. It was fine for the first couple weeks I was fiddling with the board, but lately more and more frequently the board powers on with that LED dark, until I shake or hit the tricorder body. This suggests a loose connection, and indeed one side of the LED appeared poorly soldered; I added more solder last week but, after initially seeming to fix it, today it's started doing it again. I re-resoldered the connections to no avail, and am at this point worried about making things worse.

I'm not skilled with electronics, so I'm not sure how else to diagnose it. The solder joints look fine to me now; could the LED itself be bad? Is there a way to repair it without having to remove and replace it?

 

[TazMan2000]

It could be the LED or it could be the circuit that provides it power. If you are not familiar with electronics, you might have put too much heat when soldering. LEDs are sensitive.
I don't want to give you advice that would make the problem worse, but taking it to someone more familiar with electronics would be a better bet. They may need the circuit diagram.

TazMan2000

 

[xl97]

Can you post some close up pics.

Do you have a multi-meter? (Do you have a HarborFreight near by? you can get one there for free usually.)

Where are you located?

 

[AnubisGuard]

Can you post some close up pics.

Do you have a multi-meter? (Do you have a HarborFreight near by? you can get one there for free usually.)

Where are you located?

Sure. My phone doesn't do well focusing on close-up, so these aren't the best.

The LED in question is LD3; the big blobs of solder are of course the ones I added. (Now that I post the pics, I see some discoloration around the solder; I'm guessing heat damage?)

I do have a multimeter:

And I'm in Los Angeles, just north of Culver City.

I appreciate the help. Thanks.

 

[xl97]

Wow.. too far (Wisconsin...LOL.. but worth a shot I guess)

Could be heat damage.. could be a bad solder still..?

check to see what voltage you are getting at the LED? is it constant?

Then could be a faulty LED

Turn the MM to 20 DCV and see what you get.. (not sure if you are working with a +5v system or +3.3v system?

What color LED? You can try to test BEFORE and AFTER the resistor too.. to ensure your getting proper voltage before. and the proper voltage TO the LED.


If all else fails.. send it to me to attempt to fix it before you scrap it!

Even if the led is blown.. and there is some heat damage.. you might be able to jumper it from other places to power it.. (if a trace is blown or something)

 

[AnubisGuard]

If all else fails.. send it to me to attempt to fix it before you scrap it!

I'd be fine with mailing it to you, paying at least a small fee for your time, and paying to have it sent back; this is a $140 set of electronics and I'm not gonna risk messing it up even more when I don't know what I'm doing. I'm not gonna scrap this either way, I just might have to accept a replica with a bum LED if nothing else works out.

Turn the MM to 20 DCV and see what you get.. (not sure if you are working with a +5v system or +3.3v system?

I don't know. How could I tell?

What color LED? You can try to test BEFORE and AFTER the resistor too.. to ensure your getting proper voltage before. and the proper voltage TO the LED.

It's green. I assume the resistor for it is the one marked R43?

Even if the led is blown.. and there is some heat damage.. you might be able to jumper it from other places to power it.. (if a trace is blown or something)

Bear in mind it works about half the time; it doesn't come on when I initially power on the board but if I jostle the board at all, then it starts working. Hence why I thought it was bad solder joint. Would a busted trace cause that?

Also, a thought: could the charger have caused this? The board came with a charging cable but no brick and no instructions on use, so I plugged it into my fast charging brick I use for my phone. (I think it's 8w?) The problem wasn't present at first and only started after my first time charging it that way. I'd assume an overpowered charger would fry the battery, but just in case....

 

[xl97]

What batteries are you powering things with? Is that a single +3.7v li-po battery pack powering it? (unless it has a boost converter I'm guessing its a +3.3v system then)

A bad trace.. or a lifted pad....

I was not aware is had built in charging circuit on the board/prop?

Its hard to tell without having it in hand... so one can look at the labels on the chips.. and see if there is some sort of voltage regulation going on..etc..

can you reproduce the 'instability' with the board out of the case.. as is?
trying touching the led and/or resistor next it to as well.. see if pressing down on either makes the light constant..

I'll definitely take a look if needed. But to be clear.. I'ma Web Developer/Software Architect by day.. and only a 'hobby' electronics guy.

But I have done many projects around here.. as well a help others on their electronics kits or fixing stuff.

 

[AnubisGuard]

What batteries are you powering things with? Is that a single +3.7v li-po battery pack powering it? (unless it has a boost converter I'm guessing its a +3.3v system then)

Correct. It's a single 300mAh battery, the part of the sticker that says what voltage is ripped off unfortunately. It's wired to the board and to a charging port that appears to be a 5.5mm charging jack. The jack cuts off the power when the cable is plugged in, regardless of if it;s hooked to a power supply or not.

I was not aware is had built in charging circuit on the board/prop?

There appears to be some kind of small circuit bard inside the battery itself, soldered to one side and inside the plastic wrap around the battery.

can you reproduce the 'instability' with the board out of the case.. as is?

Sometimes. When I took the photos above, yes, but just now when I tried to do it the LED worked fine.

Anyhow, I'm testing it with the multimeter and can't get consistent numbers off of anything. I set the meter to the unmarked "20" (the others being 20m and 20k, so I assume those are the wrong ones) and plugged the leads into the ports the instructions said to use for measuring DC voltages. I'm getting numbers anywhere from 1.53 down (up?) to 0.13 on all the LEDs I test. I must not be making good enough contact with the pads; it's very easy for those leads to drift off it's so small.

As I'm writing, I've tried it a few times and sometimes LD1, 2, and 4 are around 0.7 to 0.9 while LD3 is 1.5-ish, then LD3 is 0.9 and the others are 1.5, except for LD1 which showed 0.12 a couple times in a row.

So, no idea.

 

[xl97]

" The jack cuts off the power when the cable is plugged in, regardless of if it;s hooked to a power supply or not. "
- this is normal/typical behavior.. I just couldnt see any charging circuit from the pics.


" There appears to be some kind of small circuit bard inside the battery itself, soldered to one side and inside the plastic wrap around the battery. "
- this is probably just the 'over-charging' protection circuit.. (most re-chargeable batteries have this now a days)


Can you use the MM to just check the battery first? ensure what voltage you are getting OUT of it.. )int the rest of the circuit board).. touch the red/black wires connecting the battery pack.

 

[AnubisGuard]

Can you use the MM to just check the battery first? ensure what voltage you are getting OUT of it.. )int the rest of the circuit board).. touch the red/black wires connecting the battery pack.

3.84v.

 

[13doctorwho]

The LEDs blink in sequence right? Looking at the voltage with a meter may not show you how it's working depending on the blink rate of the LED and the sampling rate of the Multi-meter. To check the LED disconnect power and put the multimeter in diode mode. The connect the red and black leads across the LED. After that flip the leads so the red is were the black was and the black is where the red was. In one direction you will get a voltage number (and the LED may light) in the other direction you should get OL for overload. If that's what you get the LED is fine.

You can also use the multimeter to measure the resistor going into the LED. Place the leads on the metal contacts on the resistor, one on each side and set the meter to resistance. You can compare it to what you read on the resistor to one of the working "Alpha, Beta, Gamma, Delta" LEDs.

There are a lot of other things you can check, but they are a little complicated.

 

[xl97]

which is legit/fine.

I believe a fully charge li-po battery will read above 4.. (I dont have one in front of me to check.. Id' have to dig some out!..... and then charge them!) lol

 

[xl97]

Good point above..

I wasnt thinking about the sequence in real time use.

Probably not ON long enough to get a decent/stable read. (I forgot this things blinks... I dont have one!)


if you want to ship it on up here... I'll take a peek at it.. I wont do anything to it without disucssing with you forst of course

 

[AnubisGuard]

Is the LEDs blink in sequence right? Look at the voltage with a meter may not show you how it's working depending on the blink rate of the LED and the sampling rate of the Multi-meter.

Yup. It's a bit of a problem because the readings fluctuate all over the place.

 

[xl97]

updated my post while you were posting.

If you wanna send it up on.. feel free. PM me for address..

 

[AnubisGuard]

If you wanna send it up on.. feel free. PM me for address..

PM incoming.

 

[xl97]

* Responded.

 

[gmprops]

The LED is probably fine. Have you tried resoldering R43 on both sides? It is the current limiting resistor. Are all of the yellow LEDs on the back side front edge working properly? If so, then the problems is either LD3 or resistor R43. Two of the yellow LEDs on the front edge share the same driver as LD3.

I designed this circuit so if you wish to send it directly to me I will repair it for you, no charge.
Gerry

 

[AnubisGuard]

Thanks Gerry! It's actually on its way to xl97 right now, but that information will come in very handy. All the yellow LEDs are fine, so you're probably right about the resistor.

 

[gmprops]

If the yellow edge LEDs all work perfectly then the problem has to be with LD3 or R43. There is also a slight possibility that circuit trace has been damaged due to excessive heat from the soldering iron.

Here is the section of the circuit board showing the top traces in red and underside traces in blue. It should help you or xI97 troubleshoot the board. I know these boards so well that it usually only takes me about 5-10 minutes to isolate the problem.