LEDgardener

I was reading this article http://ledgardener.com/diy-led-strip-bu ... bridgelux/ and in the notes section it says

On the constant voltage/parallel systems, all strips must be connected at once to prevent the chance of strips being overpowered. If the design calls for 10 strips, do not only hook up 5.

Still a noob and the most confusing part is matching up these drivers. This quote kind of goes against what I know.

For example in the 3x3 grow it recommends 18x Samsung LT-Q562A 2′ Strips wired in parallel with a Mean Well HLG-320H-24B Driver. First of all it looks like this driver is a CC so that could be a typo. But couldn't this system work with just a single strip if need be? If some of the strips went bad it wouldn't need to be shut down right? Each strip would still get 22.3V and only draw as much current as it needs.

Was just a bit confused by this note and wanted to see if it was true.

Thanks.

That driver is constant voltage + constant current. It runs in cv mode normally, Once you max out the current with the pot, it becomes constant current.

If you hook up one strip you suggest to the driver you suggest, that strip will pull it's max current and voltage running at max, and degrading the life of the strip. You only want to run at about 75% of any strips max, unless you have serious cooling. Running at max has a much higher chance for thermal runaway.

If you want to run less strips, and leave room for expansion on a constant voltage driver,you have to get the a version of that driver, not the b. The a version has 2 internal potentiometers. One that limits volts and one that limits current. By being able to control voltage you limit what the max voltage is for each strip that is wired parallel to the driver.

Maybe some real world numbers will help you understand whats going on.

I'm running 10 Samsung 2' LT-Q562 strips with a MW HLG-240H-24B driver (Smaller driver and less strips then the article you are quoting). I have the adjustable potentiometer on it with a digital readout of voltage, current, and wattage going to the lights. I have switches so I can turn on and off the lights individually if I wanted, I just turn down the wattage accordingly.

Here are ACTUAL readings from my light with all 10 LED strips on (40 LED/strip);

100.8w = 22.21V @ 4.54A = 0.454A/strip = 10.08w/strip = 0.252w/LED
200.4w = 23.17V @ 8.65A = 0.865A/strip = 20.04w/strip = 0.501w/LED
240.7w = 23.53V @ 10.23A = 1.023A/strip = 24.07w/strip = 0.602w/LED

I believe (I may be wrong) that the Samsung rated max for the LT-Q562 is 1.2A (it use to be 0.900A but they increased it)

I typically run my lights at 200W (0.865W/strip) = 72% of Samsungs rated max for the strip.

For reference the HLG 288v2 (301b diodes) (https://horticulturelightinggroup.com/c ... v2-led-kit) runs a max of 150w/288LED's = 0.521w/LED. That's a little more then I typically run at (0.501w/LED). If you were to run the 18 2' strips of LT-Q562A with the HLG-320-24B driver your max would be 0.444W/LED.

If you max out the HLG-320H-24B driver on 18 strips it's 17.777W/strip, that's lower then I typically run. The HLG-320H-24B is a CV driver, if you dial it down it's a CC/CV driver that runs a curve. You can run it with a single strip but you'd be powering that single strip at absolute max that the strip could take, this would generate a lot more heat (fire risk), reduce the lights efficiency, and shorten the lights life. You can turn the potentiometer way down and run only a few lights (one light is still probably unrealistic and unsafe for such a large driver).

The article on the build looks to be a good one, if you follow it should should be fine (I didn't write the article and use the information at your own risk)

So long story short. If I have a Meanwell HLG-240H-36B(6.7amp) hooked up to 4 boards that pull 1.7amps at 36 volts and 1 of the boards goes out the other 3 boards would be pulling way over their max because the 1.7 amps from the missing board would be distributed to the remaining 3 boards?

How does one have any peace of mind with this type of setup knowing that at any time the system could be over powered. It seems a potentiometer can't even prevent this. The only preventative is to not have a board fail which is out of control.

Someone smarter then me can explain how it works in a series wiring setup, to my understanding its even worse.

I'm pretty sure that the reason people are going with more expensive passive cooled systems is because early active cooled leds would fry if the fan gave up. The passively cooled systems people build here are already way more stable then what existed in the past. In the last year reading about strips I've barely encountered someone who blew one. The lifespan is generally predicted based on operating temperatures. A 10c shift in running temperatures changes the 50000h rating of lifespan quite a lot.

Also I notice that the LedGardner always recommends builds that are well cooled, and could mostly support the temporary raise in temperature that could come from a strip giving out. In general there are lots of ways this could be accounted for in the original design:

More/ smaller strips/ boards per driver

Difference between your running temperature and the max temperature of the strips, if its small, you have less tolerance for error.

Building a system that follows strip temperatures and cuts lights when they rise.

I have one setup where I have a 75w driver per 2x2foot strips. If a strip fails in that setup, the other one might really fry too. That said they run at around 40c and have a max temperature of 80c, so chances are very low, and two replacement strips would cost 16e.
I have two other fixtures where a 75w driver feeds 6x1foot strips and 10x1foot strips. Sometimes the ghetto wiring fails and a strip goes out, especially in the build with 10 strips. The system has way too much tolerance to care. Strip temperatures stay under 40c.

miketoasted wrote: ↑

Tue Sep 25, 2018 7:04 pm

So long story short. If I have a Meanwell HLG-240H-36B(6.7amp) hooked up to 4 boards that pull 1.7amps at 36 volts and 1 of the boards goes out the other 3 boards would be pulling way over their max because the 1.7 amps from the missing board would be distributed to the remaining 3 boards?

How does one have any peace of mind with this type of setup knowing that at any time the system could be over powered. It seems a potentiometer can't even prevent this. The only preventative is to not have a board fail which is out of control.

If you have a CV driver that outputs 5a max, and you have 5 strips wired to that driver that are max of 1a each. If you want to drive those strips at 750mA each (75%,) you have two options.

You could go with the b version of the driver which only limits current. If You limit the current to 3.75a, each strip is getting 750mA. If one burns out, 4 strips Will be pulling 3.75 amps now. That's 980mA each. If two burn out, that driver will only run each strip to its 1a max, because each strip will only pull the current it needs. But running strips over 75% is never recommended.

If you go with the A version of the driver, you can limit the voltage that each strip will run at. By limiting the voltage you control how many volts each strip runs at in a parallel system. Current drawn by each strip is relative to the voltage each strip is ran at. So each strip will run at 750mA weather you have 1 or up to 5 strips connected. Once you have more than 5 strips connected it will start dividing the current more and more until it reaches the minimum voltage range of the driver or strips as possible.

Oh! Each strip will only ever pull its max. So it won't totally fry if strips burn out it will just run at max. Peace of mind enough for me. I knew this sounded too intense for safety.

Thanks.

That is how it works in theory. I'm not saying they won't pull,over max. Running strips at max has a higher probability of thermal runaway. Stuff can malfunction, especially if the stripsget hotter than they are rated for.

Not enough torture tests done...