Author Topic: RED LED Shorted on??  (Read 4771 times)

Offline richardb

  • Full Member
  • ***
  • Posts: 83
RED LED Shorted on??
« on: December 05, 2010, »
I built 4 Aethers at the same time, and "cooked" the all on the griddle at the same time.  (I have the same model as RJ).
However, that didn't go very well as there was insufficient heat to reflow.  So, I had to remove 2 of them and let them reflow and the swap them out...
I then "glued" the LEDs onto the board and tested for short between the ground pad for the LED and all 6 of its legs.  All was good.
I then completed assembly of the thru-hole parts.
At test, 1 of the units will "flash" all white and then the LED with the Temp sensor stays full brigth RED.  Since this is the LAST LED in the series circuit for the LEDs, the short should be easy to find...  Alas it is not...
From top to bottom with the board oriented with text upright, the 3rd pin shorts to ground with a resistance of 1.5 Ohm (My short lead reading is .1 Ohm), or with the 4700 uF cap remvoed about 12.6 Ohm.
This causes the tab on the LED with the temp sensor to read a short as a capacitor charges and then an open.  Its neighbor to the left also reads a short with a slight capactive drop from first contact of the leads to stability of about 500 ms and about 12.6  Ohms.  This reading is polarity insensitive, where as the other will dischage and recharge with lead reversal.  
When you power up and down the board, the RED Led fades as the capacitor discharges.. Finally becoming a faint red glow for many seconds.
I have inspected the VIA's at the pads of the LED and where they attach to the other side, and do not see soldier seepage shorting to ground on that side.
I am concern the this may be a reflow issue where some solder has not flowed, but all looks shiny and spot checks for leg shorts tunrs up nill
The constant current sections are from Left to Right Green, Red, Blue, so the middle one is the red circuit, and I have inspected it also turning up nothing.

My second has a similar problem, except the top row and the left bottom RED ligth up.  The discharge is more rapid, with little to no fade.

THoughts???

Offline PJNMCT

  • Patron Member
  • Sr. Member
  • ****
  • Posts: 821
Re: RED LED Shorted on??
« Reply #1 on: December 11, 2010, »
You are not allowed to view links. Register or Login
I built 4 Aethers at the same time, and "cooked" the all on the griddle at the same time.  (I have the same model as RJ).
However, that didn't go very well as there was insufficient heat to reflow.  So, I had to remove 2 of them and let them reflow and the swap them out...
I then "glued" the LEDs onto the board and tested for short between the ground pad for the LED and all 6 of its legs.  All was good.
I then completed assembly of the thru-hole parts.
At test, 1 of the units will "flash" all white and then the LED with the Temp sensor stays full brigth RED.  Since this is the LAST LED in the series circuit for the LEDs, the short should be easy to find...  Alas it is not...
From top to bottom with the board oriented with text upright, the 3rd pin shorts to ground with a resistance of 1.5 Ohm (My short lead reading is .1 Ohm), or with the 4700 uF cap remvoed about 12.6 Ohm.
This causes the tab on the LED with the temp sensor to read a short as a capacitor charges and then an open.  Its neighbor to the left also reads a short with a slight capactive drop from first contact of the leads to stability of about 500 ms and about 12.6  Ohms.  This reading is polarity insensitive, where as the other will dischage and recharge with lead reversal.  
When you power up and down the board, the RED Led fades as the capacitor discharges.. Finally becoming a faint red glow for many seconds.
I have inspected the VIA's at the pads of the LED and where they attach to the other side, and do not see soldier seepage shorting to ground on that side.
I am concern the this may be a reflow issue where some solder has not flowed, but all looks shiny and spot checks for leg shorts tunrs up nill
The constant current sections are from Left to Right Green, Red, Blue, so the middle one is the red circuit, and I have inspected it also turning up nothing.

My second has a similar problem, except the top row and the left bottom RED ligth up.  The discharge is more rapid, with little to no fade.

THoughts???

I have built 5 units and they all tested good initially (I'm pretty sure) but now I have two units that have the upper right LED (next to the temp sensor) on RED at full intensity all the time.

Any ideas?

-Paul
« Last Edit: December 11, 2010, by PJNMCT »
Leesburg, FL

Offline n1ist

  • Coop Manager
  • Sr. Member
  • *
  • Posts: 760
  • 02148
Re: RED LED Shorted on??
« Reply #2 on: December 11, 2010, »
The LEDs are wired in series.  The one nearest the temp sensor has the anode tied to the positive supply rail, and the ones closest to the constant current chips have their cathodes pulled low by those chips.  If the one nearest the temp sensor is stuck on and the rest are all off, there's a short to ground either at that LED or the next one over.  Also, there's a good chance that the LED will need to be replaced.
/mike

Offline PJNMCT

  • Patron Member
  • Sr. Member
  • ****
  • Posts: 821
Re: RED LED Shorted on??
« Reply #3 on: December 11, 2010, »
You are not allowed to view links. Register or Login
The LEDs are wired in series.  The one nearest the temp sensor has the anode tied to the positive supply rail, and the ones closest to the constant current chips have their cathodes pulled low by those chips.  If the one nearest the temp sensor is stuck on and the rest are all off, there's a short to ground either at that LED or the next one over.  Also, there's a good chance that the LED will need to be replaced.
/mike


Thanks Mike, I thought there might be a short between the led and the surface where the arctic aluminum was placed as I thought I was a little light on applying it originally so I popped the "shorted" led up. It did not change anything so I popped the next one over based on what you were saying above. Now the top row (3) and the bottom left are all red - the other 2 are still ok.

Running out of attempts  <fp.

BTW I touched both leds back down briefly while energized - nothing changed - still 4 of 6 red now!

I'm sure I misunderstood what you were saying above...or not.

Any other ideas? I still have two good leds  ;D ;D (Actually I think they are all still good, it's probably somewhere else).

It would be nice to have some kind of a limited schematic...

Thanks,
-Paul
Leesburg, FL

Offline RJ

  • Administrator
  • Sr. Member
  • *****
  • Posts: 8519
Re: RED LED Shorted on??
« Reply #4 on: December 11, 2010, »
When you are popping them up you are overheating them and ruining them. The pcb is their heatsinks and they can not survive with out them. These leds should never be operated with them unattached from the pcb.

Was the first one the one you took lose before because you thought something was wrong with it?

RJ

Innovation beats imitation - and it's more satisfying

Offline richardb

  • Full Member
  • ***
  • Posts: 83
Re: RED LED Shorted on??
« Reply #5 on: December 12, 2010, »
I couldn't locate any shorts, and have benched these two units until after the first...
BUT....
Now one of the others that worked fine initially seems to be acting up...
With it energized in manual mode (or DMX controlled),
every 4 seconds it will go from "WHITE" to a brief flash "Golden"
Then as time progresses, the 4 seconds shortens until it is about every 1 second...
I am running out of lights to bench, so I guess I will have to look into the problems..
Any ideas...


Offline richardb

  • Full Member
  • ***
  • Posts: 83
Re: RED LED Shorted on??
« Reply #6 on: December 12, 2010, »
You are not allowed to view links. Register or Login
You are not allowed to view links. Register or Login
The LEDs are wired in series.  The one nearest the temp sensor has the anode tied to the positive supply rail, and the ones closest to the constant current chips have their cathodes pulled low by those chips.  If the one nearest the temp sensor is stuck on and the rest are all off, there's a short to ground either at that LED or the next one over.  Also, there's a good chance that the LED will need to be replaced.
/mike


Thanks Mike, I thought there might be a short between the led and the surface where the arctic aluminum was placed as I thought I was a little light on applying it originally so I popped the "shorted" led up. It did not change anything so I popped the next one over based on what you were saying above. Now the top row (3) and the bottom left are all red - the other 2 are still ok.

Running out of attempts  <fp.

BTW I touched both leds back down briefly while energized - nothing changed - still 4 of 6 red now!

I'm sure I misunderstood what you were saying above...or not.

Any other ideas? I still have two good leds  ;D ;D (Actually I think they are all still good, it's probably somewhere else).

It would be nice to have some kind of a limited schematic...

Thanks,
-Paul

To quote RJ... If the LED is shorting out because of insufficient Epoxy, it will kill the RED channel completely...  And cause other damage...

I am not sure what is causing my RED problems, but I have 2 that are doing it..
One is just the LED at the temp sensor, and the other is all three on the side of the temp sensor and bottom one farthest from temp sensor...

If I figure it out, I'll post... if you figure it out, let me know also....


Offline PJNMCT

  • Patron Member
  • Sr. Member
  • ****
  • Posts: 821
Re: RED LED Shorted on??
« Reply #7 on: December 12, 2010, »
You are not allowed to view links. Register or Login
When you are popping them up you are overheating them and ruining them. The pcb is their heatsinks and they can not survive with out them. These leds should never be operated with them unattached from the pcb.

Was the first one the one you took lose before because you thought something was wrong with it?

RJ



yes it had the upper right red led on full. I had two units in this condition. Now after popping up two leds on one unit, it has four leds on red at full intensity.

Sequence (on the one unit):
all LEDS "glued" down - upper right one has red on full.
popped upper right one - no change.
popped upper middle one also - now upper row (3) and lower left red leds, for a total of four - on full.

-Paul
« Last Edit: December 12, 2010, by PJNMCT »
Leesburg, FL

Offline richardb

  • Full Member
  • ***
  • Posts: 83
Re: RED LED Shorted on??
« Reply #8 on: December 13, 2010, »
After the holidays, unfortunately, I will draw out a schematic and trouble shoot the problem.  Since we are experiencing the same issues, I bet the same fix for both.  There is a mystery via or some other oddity in play here.
If you have a board you have not soldered yet a HIGH RES pic of both sides would aide greatly in diagnosing.  If you or anyone else has one...
The Gerber file would be awesome, but I bet RJ isn't letting that out...

Needless to say, I have 2 of the 5 operating, wanted at least 3... But I will deal with it after the season...

Offline richardb

  • Full Member
  • ***
  • Posts: 83
Re: RED LED Shorted on??
« Reply #9 on: December 13, 2010, »
What I meant was if you have a board and can "scan" the surfaces in, that would be ideal...
But a photo in great detail will suffice...

Offline RJ

  • Administrator
  • Sr. Member
  • *****
  • Posts: 8519
Re: RED LED Shorted on??
« Reply #10 on: December 13, 2010, »
Ok let me say this clearly. If you are "popping" loose and powering these things even for a second you are the fantom issue. The leds will be damaged and then the next one becomes an issue.

The red led's positive lead is connected internally to the heatsink on the tab of the led. The leds are in series just like christmas lights in a string. One led affects the next led. There is no mystery via if that was it I would have a few hundred of them doing this.

The likely areas are since the leds legs are closer together and there is grounds next to each other a extremely small amount of solder is all it takes to short out a circuit.

Or the back being shorted. But heres the catcher you can not just pop the led up and try it. You need to add epoxy and put it back down. Then once dry try it. If users pop them off the board and power them up they are killing them and the next ones get damaged. you can not bypass an led you can not rig it. These are extremely high power leds and doing this stuff will damage leds and driver circuits.

The last led that both of you are talking about will show connection when other parts are attached. That is why you want to check them as you apply them. I am sure you did not that should be enough once the epoxy is dry they are not moving.

If one of you two have not popped any of the leds off the pcb and have a red led on mail it to me. I will find the issue. It could even be a bad led even though it is unlikely.

RJ

Innovation beats imitation - and it's more satisfying

Offline PJNMCT

  • Patron Member
  • Sr. Member
  • ****
  • Posts: 821
Re: RED LED Shorted on??
« Reply #11 on: December 13, 2010, »
You are not allowed to view links. Register or Login
If one of you two have not popped any of the leds off the pcb and have a red led on mail it to me. I will find the issue. It could even be a bad led even though it is unlikely.

RJ

RJ, I've only been messing with one of the problem boards. Unfortunately, work has me putting in 16 hr nights starting tomorrow for the week - got an attraction startup. If Richardb hasn't followed up, I will drive one down to you next week.

Like him, I am just setting up the others that are working for this season.

Thanks,
-Paul
Leesburg, FL

Offline richardb

  • Full Member
  • ***
  • Posts: 83
Re: RED LED Shorted on??
« Reply #12 on: December 13, 2010, »
I was only addressing the boards with the red LEDs stuck on.  If the LED has overheated due to lack of heat sink, yes, that is a whole other problem.
The mystery via would be the one's located by the "G+" "R+" and "B+" silk screening.
There looked like there was one under there, but couldnt tell... However, I don't thing those are an issue.. But thanks for the pic.  I will review the boards more once things have calmed down.

Offline PJNMCT

  • Patron Member
  • Sr. Member
  • ****
  • Posts: 821
Re: RED LED Shorted on??
« Reply #13 on: December 29, 2010, »
RJ,

I'm sure you know by now that I delivered both to JJ today.

Just some further info - I've been on vacation for a week and I came back to find the GFI tripped. Makes sense due to the rain here while I was away. Anyway, after resetting the GFI, show works fine still but now I have two more Aethers (out of three in-show) with one red LED - the same - upper right. These are not on at full intensity and they seem to work fine with non-zero data. Again, when commanded off, the problem shows up with only these two LEDs (same one on each aether) - one of them is flickering low intensity red, the other has no flicker but on solid at low intensity.

Thanks in advance for the assistance.

I'm thinking it is a construction problem that is common to all my Aethers and the one I have left has yet to exhibit the problem.

-Paul
« Last Edit: December 31, 2010, by PJNMCT »
Leesburg, FL

Offline Gary

  • Sr. Member
  • ****
  • Posts: 554
    • Diamond Crescent
Re: RED LED Shorted on??
« Reply #14 on: December 31, 2010, »
I had similar issues with one of my Aethers. I ordered 6 kits, with the intention of using 5 this year and having one as a spare. Well... it's good I ordered the spare.

I assembled all the boards and once I got them all to work perfectly outside the cases, I started installing them into the cases. The fifth one I put in the case had this issue where the LED in the upper left corner was glowing red. I checked for a short between the rectangular pad on the top of the LED and the board, and there was a short. Dang! Thinking that is must have been a stray piece of aluminum somehow got stuck between the LED and board (which is possible because I didn't put so much adhesive on that it was oozing out... I did just enough to coat it nicely with no obviously exposed metal still visible), I tried compressed air, shaking the thing hoping that a tiny piece of aluminum falls out of some crevice somewhere, poking around, looking closely around the LED legs, but nothing worked to resolve the issue.

I thought it was maybe something on the other side of the board, so I unscrewed the screws holding the board in the case,  and the LED didn't glow any more. After taking it in and out a few times, it's working fine again with the board screwed into the case. But there is a still short between the LED and board. I'm not at the light right now, but at the top of my head, I think the one to the left of it has the same short as well, but it doesn't (and never did) glow. Come to think of it, the same adjacent LED light sometimes had the short issue, and sometimes it didn't. The last time I checked it, there was a short.


Is the glowing red LED (or "short" problem I have now) problem likely due to a bit of Aluminum somewhere?

Perhaps the assembly instructions should be changed to have us place the board into the case, mark the aluminum bracket with a pencil to get the alignment, remove the board, and then do the drilling--thus avoiding all these little pieces of aluminum flying all over the place. I find it odd drilling with such valuable circuit board right in the case, but thought I should follow the instructions in case it made alignment of the drilled holes better. I did the best job I could to remove all specks of aluminum from inside the case, both sides of the board, and I even loosened the screws holding the brackets to the case because of small bits being wedged in between there, so I could clean that out too.

If this light becomes problematic, for future reference, what is the best way to pop off the LED? Do you think I can pop it off while still soldered in, stick more thermal adhesive underneath, or would the LED have to be desoldered and removed altogether first before regluing... that wouldn't be a pretty task.
You are not allowed to view links. Register or Login