Author Topic: Bypassing faulty LED in a strip, good idea?  (Read 7619 times)

0 Members and 1 Guest are viewing this topic.

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Bypassing faulty LED in a strip, good idea?
« on: May 27, 2021, 08:16:37 pm »
I have got a LED grow light which has got 2 LEDS which has gone open and resulted in one half of the light not turning on.
The whole light consists of 60 x 10W leds. 30 leds are grouped together and powered from one power supply.
First of all I couldn't find any 10W leds locally and it seems that its hard to come by online as well. Only leds I could source locally are 1W to 3W which presumably I couldn't interchange.
There are diodes paralleled with the Leds, I'm not sure what they really are but looks they are 3V Zeners. The diodes which were paralleled with the two dead Leds also were burnt.
I was thinking of bypassing the two empty slots with jumper wires, would this be a good idea? I have seen that done on backlights of LED tvs. I am not too concerned with having two brightness levels. Instead of jumper wires should I use resistors? If so what value should I use. The voltage output of the power supply is 95V (I know these controller would be constant current sources).
Thanks for the help in advance.
P.S: Could you please advise on soldering/desoldering these leds. The whole body seems to be soldered along with the two legs. I am able to desolder the two legs however when trying to remove the body  the whole LED disintegrates leaving the metallic part soldered. I tried with my hot air station and had no luck the plastic cover disintegrates and to take the metallic part I had to tap with a screwdriver. (Refer to photo 2)
 

Offline shakalnokturn

  • Super Contributor
  • ***
  • Posts: 2212
  • Country: fr
Re: Bypassing faulty LED in a strip, good idea?
« Reply #1 on: May 27, 2021, 09:38:09 pm »
I wouldn't go for resistors I can't see the point in dissipating 10W (are they really 10W LEDs?) in a resistor.
If there are two separate constant current sources shorting a couple of defective LEDs in place is the simplest solution.
If there's a single CC source with two LED strings in parallel you'll have to replace the LED's or equalise the number of LEDs in each string.

Funny design... What was the point in sticking mini-Melf zeners in parallel? Did they actually trust those to dissipate over 10W? (I wouldn't trust them with 3W!) Were they meant to overheat then short? Obviously it didn't work as intended.

There are much more appropriate devices for this such as the MP4690.

For desoldering go as high as is reasonably possible (you don't want to melt all the LEDs) with a preheat plate from underneath then remove individual LEDs with high airflow hot air from the top.

 
The following users thanked this post: Yamin

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Re: Bypassing faulty LED in a strip, good idea?
« Reply #2 on: May 28, 2021, 08:21:19 am »
I wouldn't go for resistors I can't see the point in dissipating 10W (are they really 10W LEDs?) in a resistor.
If there are two separate constant current sources shorting a couple of defective LEDs in place is the simplest solution.
If there's a single CC source with two LED strings in parallel you'll have to replace the LED's or equalise the number of LEDs in each string.

Funny design... What was the point in sticking mini-Melf zeners in parallel? Did they actually trust those to dissipate over 10W? (I wouldn't trust them with 3W!) Were they meant to overheat then short? Obviously it didn't work as intended.

There are much more appropriate devices for this such as the MP4690.

For desoldering go as high as is reasonably possible (you don't want to melt all the LEDs) with a preheat plate from underneath then remove individual LEDs with high airflow hot air from the top.


Thanks @shakalnokturn for the reply. They are Melf Zeners right? I'm just assuming because of the blue band and how they are oriented across the LED.
There are two separate CC sources for the two LED strips (30 LEDS each). I took out the controller to read its label and I've attached it here. It states that its 560mA. So if I'm not wrong I'm calculating that its roughly around 1.68W of power for each LED.
V across LED = 3V
P = VI; 3x560x10^-3 = 1.68W
Am I right with my calculation? I don't know why its stated that each LED is 10W and full power for the Light is 600W (60x10W) - This is stated on the box.

Cheers for letting me know about MP4690, I haven't come across this before. 
 

Online tszaboo

  • Super Contributor
  • ***
  • Posts: 8107
  • Country: nl
  • Current job: ATEX product design
Re: Bypassing faulty LED in a strip, good idea?
« Reply #3 on: May 28, 2021, 08:29:55 am »
Yes, you can bypass a faulty LED in a strip that is current driven. Even often times you can do that in voltage driven strips, if the strip is long enough.
Those might be zeners, but I suspect they are LED protectors. Kinda like a SCR latch, it breaks down with high voltage, and then conducts at a lower voltage.
https://www.littelfuse.com/products/led-protectors.aspx
 
The following users thanked this post: Yamin

Offline wraper

  • Supporter
  • ****
  • Posts: 17822
  • Country: lv
Re: Bypassing faulty LED in a strip, good idea?
« Reply #4 on: May 28, 2021, 08:42:36 am »
If all LEDs are in series, you can do that. However most likely it's not worth it, if LEDs started to fail, other LEDs usually will fail soon too. Most likely these LEDs were overdriven by too high current. Also worth to mention that you should care about lumens, not Watts. A lot of Watts on cheap LED bulb often means it consumes a lot of power while emitting too little light for the power consumed, and is overdriving LEDs, thus severely reducing their lifespan.
« Last Edit: May 28, 2021, 08:45:35 am by wraper »
 
The following users thanked this post: Yamin

Offline wraper

  • Supporter
  • ****
  • Posts: 17822
  • Country: lv
Re: Bypassing faulty LED in a strip, good idea?
« Reply #5 on: May 28, 2021, 08:49:25 am »
I don't know why its stated that each LED is 10W and full power for the Light is 600W (60x10W) - This is stated on the box.
Because it's bullshit targeted on unsuspecting buyers.
 
The following users thanked this post: Yamin

Online coromonadalix

  • Super Contributor
  • ***
  • Posts: 6860
  • Country: ca
Re: Bypassing faulty LED in a strip, good idea?
« Reply #6 on: May 28, 2021, 09:36:03 am »
theses are at least 1 watt or 3 watt leds,  not 10 watts for sure

you have a thermal pad under the led, sometimes it can or could be soldered on the pcb to help dissipate the heat,  or i've seen some thermal glue in one case ....
« Last Edit: May 28, 2021, 09:39:08 am by coromonadalix »
 
The following users thanked this post: Yamin

Online tszaboo

  • Super Contributor
  • ***
  • Posts: 8107
  • Country: nl
  • Current job: ATEX product design
Re: Bypassing faulty LED in a strip, good idea?
« Reply #7 on: May 28, 2021, 09:54:54 am »
I don't know why its stated that each LED is 10W and full power for the Light is 600W (60x10W) - This is stated on the box.
Because it's bullshit targeted on unsuspecting buyers.
Or its one of those "equivalent incandescent lamp" markings. 600W would get hot very quickly and it would light up a historical monument nicely
 
The following users thanked this post: Yamin

Offline shakalnokturn

  • Super Contributor
  • ***
  • Posts: 2212
  • Country: fr
Re: Bypassing faulty LED in a strip, good idea?
« Reply #8 on: May 28, 2021, 01:50:13 pm »
P = VI; 3x560x10^-3 = 1.68W
Am I right with my calculation? I don't know why its stated that each LED is 10W and full power for the Light is 600W (60x10W) - This is stated on the box.

The label rates the output between 80-110V at 0.56A so that's about 60W (not 600) output which is about 1W/LED.
Vf is likely closer to 3.3V but the rated current I guess is total output so maybe some 0.28A per string.
 
The following users thanked this post: Yamin

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20132
  • Country: gb
  • 0999
Re: Bypassing faulty LED in a strip, good idea?
« Reply #9 on: May 28, 2021, 02:58:11 pm »
Shorting the LEDs might increase the curret flow, or cause other components to run much hotter. Why not simply replace the LEDs with new ones? They're not that expensive and you'll be doing a propper repair, restoring it to its original condition, rather than bodging it.
 
The following users thanked this post: Yamin

Offline YaminTopic starter

  • Frequent Contributor
  • **
  • Posts: 309
  • Country: mv
Re: Bypassing faulty LED in a strip, good idea?
« Reply #10 on: May 28, 2021, 07:21:57 pm »
Hi guys, thanks so much for the replies.
I bypassed the two faulty LEDs and took some measurements. The output voltage of the controller goes down to around 84V (it is 95V on the "good string"). The current draw remains to be constant at around 550mA on both strings.
Vf measured across the LED measures around 3.1V. So is my calculation from before correct? and the power dissipated is 1.68W (per LED)? Could I use 3W, 3V LEDs available locally as a replacement?
I am slightly confused with the Watt ratings of a LED. When its says '3W LED' does it mean that its the highest power it could dissipate or does it mean that it'll draw that much current from the supply for it to dissipate 3W?(hope I made sense). What would happen if a LED strip is made up of different Watt rating LEDs but all have got the same Vf?
When I replace the LEDs I would have to change the mystery zener diode or the safety device which is "paralleled" right? Could I use the led protectors suggested here for the two replacement or use 3.3V Zeners off the shelf? Could I leave it for the time being?
I'll include a picture of the label of the light which shows the 60x10W,I also figure it some marketing BS.
Once again thanks so much for the help really appreciate it.
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 10083
  • Country: gb
Re: Bypassing faulty LED in a strip, good idea?
« Reply #11 on: May 28, 2021, 08:34:52 pm »
Higher power rated LEDs will have approximately the same forward voltage drop as the existing ones. If you feed them with the same current, they will run at the same 1.68W output power dissipation, just more efficiently and hopefully a bit cooler (better thermal coupling).
« Last Edit: May 28, 2021, 08:40:05 pm by Gyro »
Best Regards, Chris
 
The following users thanked this post: Yamin

Offline shakalnokturn

  • Super Contributor
  • ***
  • Posts: 2212
  • Country: fr
Re: Bypassing faulty LED in a strip, good idea?
« Reply #12 on: May 28, 2021, 10:05:15 pm »
Amazon Galaxyhydro / Roleadro ganja light.
https://www.amazon.com/Galaxyhydro300-600-Grow-Light-300W/dp/B01HI3AFYM

They keep saying it... 10W per LED!
In a similar sale they say "Our 600W only consumes 134W..." (Must include a patented free ganja energy generator.)


[Edit]: "Zenner to protect each led, if one want out, the others leds won't be affected."
So that was the plan! Not ESD protection.
I'm sure they do more testing of the plant production than light products quality check, that way "If two wank out they're in the dark."
« Last Edit: May 28, 2021, 10:14:33 pm by shakalnokturn »
 
The following users thanked this post: Yamin

Offline wraper

  • Supporter
  • ****
  • Posts: 17822
  • Country: lv
Re: Bypassing faulty LED in a strip, good idea?
« Reply #13 on: May 29, 2021, 08:34:30 am »
Amazon Galaxyhydro / Roleadro ganja light.
https://www.amazon.com/Galaxyhydro300-600-Grow-Light-300W/dp/B01HI3AFYM
It's not Amazon but just some Chinese seller selling on Amazon.
Quote
Business Name:Li Xugen
Business Address:
3rd Floor, Block 2 Yijiayang Industrial Park
Huaming Road,Dalang Street Longhua District
Shenzhen
Guangdong
518109
CN
« Last Edit: May 29, 2021, 08:37:32 am by wraper »
 
The following users thanked this post: Yamin


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf