3
\$\begingroup\$

This has happened twice and I'm hoping someone can tell me what I'm doing wrong so that I don't fry another set of drivers and LEDs.

I have an array of high-intensity LEDs for a crop imaging application looking for plant fluorescence.

We have strips of copper PCBs that were custom made with chains of six green and blue high-intensity LEDs on them (in the application we alternate looking at the plants under green and blue light).

The array

To drive these boards, I am using Mean Well constant current drivers on a custom PCB I made which powers four drivers at a time and routes control signal to two at a time.

This lets me control these ten LED strips from five boards, turning on the blue and green channels as needed.

enter image description here

(Yes, it's ugly... I'm going to eventually have wires cut to length, and have a metal backplate and wire troughs... I just use wood backplates for the first version because I almost always change something...)

For each carrier board, I can control two of the drivers at a time individually. The drivers are Mean-Well NLDD-1400H. Each driver board has an individual connection to a DIN rail distribution block... I'm not daisy-chaining the power.

The drivers are always on if not connected, so I have them wired to the Arduino and added 18k pull-down resistors for each channel.

I am only using two outputs, one for each color and am daisy-chaining the signal to each of the control boards, splitting the one in the middle.

The simplified wiring layout I made in Lucidchart here.

In any case, what has happened twice now, is that I have fried the last set of drivers on the green channel. Sizzling bacon sound, hole appeared in the driver... the smoke got out... very, very dead.

When that happened, it also killed the green side of my LEDs.

The blue channel has never had a problem... and from what I can tell, it is wired exactly the same way.

I don't know when they died the first time, but after replacing them, I applied power and attempted to turn the lights on manually (not through the Arduino, but manually touching the control wire for the greens to ground and 3.3 V to try to get them to turn on).

I don't know if maybe it didn't like brief contact on the control signal because these are also PWM drivers... and a quick intermittent contact did something? But the frying sound happened pretty quickly.

Both times it was the last set of drivers at the very end of the daisy-chained string of driver boards. I don't think that is a coincidence. When they fried the second time, it was a new set of LEDs and a fresh set of drivers on a new carrier board...

Trying an image of my wiring diagram here:

enter image description here

I'm wondering if the daisy-chaining is killing the green drivers? Or maybe me providing a brief control signal... or maybe I have something wired wrong...

\$\endgroup\$
9
  • 2
    \$\begingroup\$ Can you put a wiring diagram in your post as well as a link to the driver datasheet? It is hard to understand what you're doing from just that picture. \$\endgroup\$ Commented Aug 11, 2023 at 1:33
  • \$\begingroup\$ I posted a simplified wiring diagram in that lucidchart link. I tried exporting it as a jpeg just now and found an option for makingi it large enough to be legible... please see the post again for that image... thank you! \$\endgroup\$
    – Tim Keller
    Commented Aug 11, 2023 at 2:30
  • 4
    \$\begingroup\$ What is the total voltage of your 18S lithium battery when fully charged? Does it exceed the maximum input voltage of the drivers when fully charged? I think it would unless the cell voltage is very low. \$\endgroup\$ Commented Aug 11, 2023 at 3:03
  • \$\begingroup\$ Something has to different between the blue and green systems, even though they appear to be the same. Do you have the datasheets for the blue and green LEDs? Also: It appears you also have white strips of LEDs (perhaps white LEDs?) ... how are they connected? \$\endgroup\$
    – jonathanjo
    Commented Aug 11, 2023 at 7:28
  • 1
    \$\begingroup\$ I'm going to ask what user1850479 already did, because it's important: what input voltage are you applying to the driver modules? An 18S LiFePO4 battery would be over 60V fully charged, and those drivers are rated for only 56V in normal operation. \$\endgroup\$ Commented Aug 11, 2023 at 10:01

1 Answer 1

1
\$\begingroup\$

As noted in the comments, I'd check the peak battery voltage.

Depending on charge state, one cell could be as high as 3.4V. 18 of them gives 61.2V. (I am assuming you're not charging at the same time as using, because then the voltage would be much higher?)

The PSU datasheet states Vin as 10V-56V, with 59V as the absolute maximum.

61.2V > 59V & 61.2V >> 56V.

You should also consider twisting the two PSU wires for each driver, as large loop areas with increase the inductance, and may lead to voltage surges during switching transients, which given that you are already marginal/exceeding Vin could add to your problems.

As for why green is failing as blue isn't, the per LED operating voltage is higher for blue than green (bigger bandgap), potentially making it more tolerant of overvoltage and/or requiring less voltage drop in the controller. I know you have current control on the LED strings, but have you checked the maximum voltage of the two strings? If this is significantly less than the LED driver maximum output (Vin-4V) you might have problems under transitory conditions and/or when the driver fails.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ You might also want to verify the negative lines on the green and blue strings aren't interconnected, as that could cause some fun problems depending on the driver internals. \$\endgroup\$
    – colintd
    Commented Aug 11, 2023 at 12:35
  • \$\begingroup\$ The negative lines ARE interconnected... everything here has a common ground leading back to some DIN terminal blocks, even the 12v arduino has a seperate ground wire to the same block so that I have a common ground. Is that a bad thing? I thought it was a necessary thing... \$\endgroup\$
    – Tim Keller
    Commented Aug 11, 2023 at 15:56
  • \$\begingroup\$ Connected on the PSU side is fine. I meant additional interconnect on output side. This could cause a variety of issues depending on the internal psi design. \$\endgroup\$
    – colintd
    Commented Aug 17, 2023 at 7:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.