I am building an array of LEDs that I plan to power up with an LED driver. Here is the one I chose - Mean Well LCM-60U

I have been using it and it works great. However, I have one thing I am trying to optimize on it and that is turning it on and off consistently. That datasheet explains that the turn-on time is about 1 second from cold start and that the turn-on time can vary the more it is turned on and off. I am using PWM control for dimming, but it is not possible to dim all the way to zero current. Turning on and off the AC line works, but I plan to have other controllers working in parallel and I would like them to turn on and off pretty close to the same time (100ms or so of each other).

I have two ideas to go about it, but I am not sure if this will harm the driver in any way. My first idea (that I tried) was to place a relay contact and a few regular diodes in parallel with the LED string. When I click the relay on, the diode absorbs all the current leaving the LED string off. When I want to turn on the LED string again, I open up the relay and the driver goes back to power the LEDs. Since I am cycling between a lower voltage (2V) and a higher voltage (27V), my LED string never gets overvoltaged so I am happy with the result. I just hate that there is wasted energy. My hunch is that no harm is caused to the driver, but I am not sure if this yields more consistent on/off times. Side note: I guess I do not have to use a diode to direct away the current, but a resistor would probably work okay too.

The second idea was to simply have a relay contact in series with the LED string. Thing is, the maximum open circuit voltage can reach 102V. When I close the relay to power the LEDs, I am afraid that the initial 102V will damage my LED string before the current regulates the string and it goes down to 27V.

Any comments on what I have tried so far or other ideas?

  • \$\begingroup\$ Define specs for all added features for sync error, dim levels and input, output power ranges. Step Loads with Relays to CC supplies usually causes transient issues and wasted power below 6% can be shunted with a dummy active load to get better log dimming functions to zero. Define MUST HAVE specs, then NICE to have. Specs. \$\endgroup\$ Sep 5, 2019 at 21:13
  • \$\begingroup\$ Sorry for the ill defined steps. Thank you for the info on the transients behavior. I did not take that into consideration. I originally asked the question because I thought I was not able to dim to off fast enough. Turns out, it was plenty fast. \$\endgroup\$
    – joe
    Sep 6, 2019 at 20:06
  • \$\begingroup\$ Next time list all known specs , assumptions, expectations in point form then state question. \$\endgroup\$ Sep 6, 2019 at 21:33
  • \$\begingroup\$ Again, my apologies. Should I just delete this post? Or do I close this off in some other way? \$\endgroup\$
    – joe
    Sep 7, 2019 at 18:33
  • \$\begingroup\$ if you can edit it with my guidelines, someone will respond. \$\endgroup\$ Sep 8, 2019 at 0:01

1 Answer 1


Old question but I wanted to close this with an answer. I wound up placing a relay in series with the driver output. It was acceptable to wait for a bit in my application. I therefore turned off the LED driver and opened the relay contacts. I then proceeded to measure the voltage on the output of the LED driver and waited until it discharged down to a voltage level I could withstand, and unlocked control to have the relays open whenever I wanted them to.

Thanks to all who contributed.


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.