3
\$\begingroup\$

I've been looking for answers for a few weeks, so now I'll actually voice my question!

My plan is to power 109 LEDs in series off of one constant current driver. The goal is to create supplemental light bars for a greenhouse. 3 bars powered from 1 driver. The plan is to maximize efficiency, reduce electrical waste, focus on the specific wavelengths for photosynthesis, and take advantage of the Emerson effect.

The driver model is Meanwell HLG-185H-C700. In order of forward voltages from low to high, there will be 11 at 1.7v, 46 at 2.2v, 25 at 2.8v, 11 at 3.5v, and 16 at 3.6v. The driver outputs 286vdc @ .7A. All the drivers are rated at .7A.

Can I simply connect all these LEDs in series?

If that isn't a workable solution, what are my best solutions?

I've considered the following, but any complications to the circuit increases costs and decreases efficiency:

  1. Wiring each group of Vf to its own CC driver with the appropriate Vdc output

  2. Installing a resistor of the appropriate value for each diode so that all Vf are equal (this seems counter-productive to me, will lessen the number of diodes, and increase heat and waste.)

I'm fairly ignorant electronically, so any help is welcome!

\$\endgroup\$
  • \$\begingroup\$ Make sure that the insulation on the LEDs is rated for > 286 V. i.e., If the LEDs are in close proximity to, for example, a metal frame they could arc to the frame. \$\endgroup\$ – Transistor Feb 3 '16 at 19:34
  • \$\begingroup\$ I presume you have no green LEDs in your setup as photosynthesis does not use it. (That's why leaves reflect green light.) \$\endgroup\$ – Transistor Feb 3 '16 at 19:35
  • \$\begingroup\$ in larger currents like 0.7 A junction potentials of the LEDs will be a bit larger from Vf(typical) and also you should always put a space for internal reference and leaks (say about 2V) so you should consider that the LEDs voltage potential in series will be larger than 286V and also LED string should not exceed 286 - 2V = 284V. So I think you should avoid from tying all LEDs in series. Or get rid of a few of them. \$\endgroup\$ – Alper91 Feb 3 '16 at 19:41
  • \$\begingroup\$ Data sheet links for the LEDs is the only sure way to confirm this. \$\endgroup\$ – Andy aka Feb 3 '16 at 22:41
  • \$\begingroup\$ You said "All the drivers are rated at .7A" Did you really mean that, or that all of the LEDs are rated at .7A? If 700mA is the actual absolute max rating on the LEDs, then you probably don't want to run them quite that high. If you're willing to drop the current to 500mA, you could use the -C500 version of the supply, which has 200V-400V compliance range. It's still a 200W supply, so you can adjust the number of LEDs to get the same total power. Either way, as the others have said, leave yourself some margin both in the power supply specs and the LED specs. \$\endgroup\$ – Entropivore Feb 3 '16 at 22:52
5
\$\begingroup\$

The total forward voltages of all the LEDs is specced at 286V. this is the same as the output of the supply which means that the supply has essentially no wiggle room to account for manufacturing differences. This could result in damage to the supply and is not recommended. Either remove 12-20V of LEDs, or get a larger supply.

\$\endgroup\$
3
\$\begingroup\$

As for the part of the question that is about "mixed forward voltages", or mixing forward voltages - yes, no problem so long as the CURRENTs are the same (or only run as high as the lowest-rated current device in the series string) and driver can supply adequate voltage for all of them (which, per Ignacio, seems to be a problem with the specific string of LEDs specified. I gave an upvote for bothering to take the time to add them all up.)

LEDs are primarily current sensitive and current specified devices - the forward voltages are a range, and temperature-varying as well - even for a single part number of a single color there will be variance. So long as the current is correct, the voltage can fall where it may (unless, as transistor's comment mentions, it's so high on the high end of the string that it flashes over to some other part or ground.)

I shall hope you have the cooling under control - that tends to be the hardest design task for high-output LEDs, and has a HUGE impact on their useful lifespan, or lack of one.

\$\endgroup\$
  • \$\begingroup\$ I was obviously chasing the wrong problem, but I solved it good! Thanks, I didn't grasp that differing voltages wouldn't alter the current as it passed through the series. Yes, cooling is handled. the three bars I spoke of are actually heatsinks 36"x2.75". With this new information, I'll bump each bar down to 33 LEDs on each. If it's generating too much heat, I might experiment with a lower current driver. \$\endgroup\$ – Barney Rubble Feb 5 '16 at 7:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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