Introducing CobKits.com - specializing in DIY and Citizen COBs
Lost Green
New Member
Hey CobKits,
I've been reading through your threads getting ready to purchase a set-up for a 4x4x6.5 tent. I was all set to by Cree CXB3590s until I stumbled upon a thread or two of yours (so thanks for saving me some $$, I'll hit you back with some sales dollars).
The new HLG-600H-54B looks awesome for my tent size. Does a constant current, constant voltage driver need to be wired in parallel? I thought I read somewhere that COBs wired in a series will draw Volts equal to driver volts/# of cobs, which would imply the answer to my question is yes. Is thermal runaway a concern with parallel wiring for the cobs?
The main thing I am confused about is how Luminus cxm22s can run of 54V though? The Luminus cxm22 data sheet seems to show a maxium forward voltage of 37.5. I am sure I am missing something.
I've been reading through your threads getting ready to purchase a set-up for a 4x4x6.5 tent. I was all set to by Cree CXB3590s until I stumbled upon a thread or two of yours (so thanks for saving me some $$, I'll hit you back with some sales dollars).
The new HLG-600H-54B looks awesome for my tent size. Does a constant current, constant voltage driver need to be wired in parallel? I thought I read somewhere that COBs wired in a series will draw Volts equal to driver volts/# of cobs, which would imply the answer to my question is yes. Is thermal runaway a concern with parallel wiring for the cobs?
The main thing I am confused about is how Luminus cxm22s can run of 54V though? The Luminus cxm22 data sheet seems to show a maxium forward voltage of 37.5. I am sure I am missing something.
klx
Well-Known Member
Cobby, have you ever seen an issue with multiple drivers wired in parallel to a single power lead?
I am having a problem where I can only fire 2 x 250W drivers from a single power socket and if I add a 3rd it will not light up. I would have thought 600W spread over 3 drivers would be no problem but maybe its too much for a single 10A 240V socket?
I am having a problem where I can only fire 2 x 250W drivers from a single power socket and if I add a 3rd it will not light up. I would have thought 600W spread over 3 drivers would be no problem but maybe its too much for a single 10A 240V socket?
CobKits
Well-Known Member
yes
not exactly. say your driver is 186V and 1400 mA.
at 1400 mA your cobs are 34.5V
the cobs use 34.5 X 4=178 V of voltage. the other 8V goes unused. As opposed to 186/4 = 36.5V per cob. Remember cobs can ONLY operate at the corresponding voltage in the datasheet (+/- normal variations)
that datasheet is not gen3. gen2 cxm22 were 36V cobs
klx
Well-Known Member
Yes AC side. The wire is fine because I am using the cable from the drivers connected directly to an extension lead.
I have confirmed the wiring about 10 times. I have redone it, I have run them on their own. Weirdly, when I hooked up each driver with its own power lead and plugged them into a powerboard, only 2 would light up. I can also rule out the house wiring because I hooked it up to a dedicated 15A line and still only 2 would light.
klx
Well-Known Member
Thanks for the suggestion but I figured it out. It was the potentiometer causing it. As soon as I removed it, everything lit up fine.
GrowLightResearch
Well-Known Member
A single 600W driver is often not the best fit. The HGL with the best cost per watt is the mid-power 185 and 240 watt. You have to look at the combination of supplies with the number of watts being used.
CoBs should NEVER be wired in parallel. You should use the high voltage output HGL-xxH-Cxxxx Type A.
Wire them in series.
I would run them light at 1000 mA driving five CXM-22s with a HLG-185H-C1050 Type A.
I would not run these CXM-22s at more than a maximum of 1400mA. Drive four of them with each HLG-240H-C1400 Type A.
The HLG driver automatically adjusts the voltage output to the forward voltage required by the CoBs.
In the case of driving five CXM-22s @ 1000mA, the voltage will be about 5 x 35 = 175V (188V max). the voltage range of the HLG-185H-C1050 is 95V - 190V.
Possibly.
The real problem is the forward voltages of the CoBs are not all identical. When wired in parallel they will all be running at the same forward voltage. They each should be run at their own forward voltage.
Last edited:
If making a linear fixture with these what is a good spacing? They will be run at approx 60 watts ea, the fixture is similar to a tube, one side is a heatsink that the cobs are mounted to. There is a fan at one end to move air over the heatsink. I was thinking about 10" c-c, the whole fixture would be approx 36" long.
Meant to say Luminus CXM22 Gen3 3000K 90 CRI.
Meant to say Luminus CXM22 Gen3 3000K 90 CRI.
Last edited:
CobKits
Well-Known Member
the larger "HLG" drivers do cost very slightly more per watt (~2 cents/W) but are also substantially more efficient so thats a quick payback. also easier wiring for some setups if its a fit
That is false. there are numerous ways to wire cobs safely in parallel. using more than 3 cobs per driver, using drivers with the proper voltage range and limits, etc.
not sure where this suggestion is coming from but there is no reason cxm22 cant be run above 1400 mA with adequate cooling. Their maximum rating is a (very conservative) 2400 mA. That said 1050-1400 mA (50-75W) is a great range and very popular for these chips
Constant current drivers do no form of "voltage adjustment", they control current to the cobs which operate at the proper forward voltage for that current
with all due respect, please stop posting information in my thread on things you dont fully understand. the CXM22s i sell are not 36V cobs
not a "problem" per se. about a max of 1/2V spread in most lots of chips which is less than 10% on a flux basis which is essentially irrelevant in an array. flux per chip is not the same in a series array either as each chip is running at a fixed current but independent voltages
Last edited:
Lost Green
New Member
that datasheet is not gen3. gen2 cxm22 were 36V cobs[/QUOTE]
Thanks - I was definitely looking at the Gen2 datasheet. One last question for you:
If I am wiring 16 cxm22s to the HLG-600H-54B, can I use the Wago 5-wire connectors to do the parallel wiring? First Wago would have the positive in from the driver + 3 cob positives + 1 wire to the next Wago, second - fourth would have the positive in from the prior Wago + 3 cob positives + 1 wire to the next Wago, fifth and last Wago would have the positive in from the prior Wago + 4 cob positives (3+3+3+3+4=16 Cobs using 5 Wago 5-wire connectors). Negative side would just be the reverse of that.
The reason I ask is because I'm not sure if the Cobs on the first Wago connector would get more Volts/Amps than the cobs on the last Wago. Might be a dumb question.
Thanks - I was definitely looking at the Gen2 datasheet. One last question for you:
If I am wiring 16 cxm22s to the HLG-600H-54B, can I use the Wago 5-wire connectors to do the parallel wiring? First Wago would have the positive in from the driver + 3 cob positives + 1 wire to the next Wago, second - fourth would have the positive in from the prior Wago + 3 cob positives + 1 wire to the next Wago, fifth and last Wago would have the positive in from the prior Wago + 4 cob positives (3+3+3+3+4=16 Cobs using 5 Wago 5-wire connectors). Negative side would just be the reverse of that.
The reason I ask is because I'm not sure if the Cobs on the first Wago connector would get more Volts/Amps than the cobs on the last Wago. Might be a dumb question.
Lost Green
New Member
Yessir - I like to get my research done beforehand haha
Last edited:
Lost Green
New Member
You're the man!
Seriously though, you are a wealth of knowledge. Reading through your threads has been a huge help. You have ensured one more loyal customer.
Hope the surgery goes well/did go well.
Seriously though, you are a wealth of knowledge. Reading through your threads has been a huge help. You have ensured one more loyal customer.
Hope the surgery goes well/did go well.
GrowLightResearch
Well-Known Member
The OP posted the electrical characteristics where the forward voltage was 35V
The voltage of a buck LED driver is a function of the LED's Vf. Not requiring any adjustment (aka auto adjust). It's a simple way to explain it to a novice.
WRONG!! None of the chips will be running at their individual forward voltages as by the nature of physics the voltage at each chip will be the same when wired in parallel. They WILL run at their independent Vf when wired in series.
The is no GOOD way to wire CoBs in parallel driven with an HLG. If you do not believe me wire some in parallel and measure the current going through each CoB.
When wired in parallel the forward voltage must be the same because all the anodes and cathode are connected together at the same potential which means they cannot operate at their individual forward voltages. Wouldn't common sense say that is not a good thing to do?
I just ran a test on two CoBs. At 600 mA the measured Vf were 33.4V and 34.2V powered by a HLG-60H. At 1200mA wired in parallel the currents were 768mA and 463mA respectively and the forward voltage 34.9V (higher than either of the individual Vf).
Substantially??? Really? HLG-600 93-96% at 100% load and max voltage. HLG-240 93.5%. But what I was saying it is better to match the load to the driver rather than just pick a driver capacity willy nilly. It is best to maximize the forward voltage and have the string of LEDs just under the driver's max voltage for the best efficiency.
As an electrical Engineer I worked on designing military grade switching power supplies beginning in 1978. It is you that does not fully understand how to properly wire CoBs.
The key there is "with adequate cooling". As you increase the current over the test current the efficiency will drop 30% between 1200mA and 2400mA, the forward voltage will rise (increasing wattage) and due to mediocre thermal resistance, the temperature will rise further diminishing efficiency. Sure you can run them at 2.5 amp, but you will need substantial (i.e. costly) thermal management and efficiency will be significantly diminished. It is best design practice to design a passive heatsink then add a fan. this way when the fan fails you will not damage the CoB. You cannot do economical passive cooling at 2.5 amps.
The case temperature of these CoBs will run about 100°C at 700mA with minimal thermal management. The max operating case temperature of these CoBs is 105°C.