[MPTESE]

Hello,


I desire to power (4) 300W LED COBs simultaneously via a single 120v AC plug. Each COB has their own 300W driver. The driver is AC input and DC output. It is a constant current driver.


Here's the spec sheet of the driver. I assume I would be using the H-AB, which provides a DC voltage range of 29-58v, and 3-in-1 dimming which is another part of this.

https://www.meanwell-web.com/content/files/pdfs/productPdfs/MW/HBGC-300/HBGC-300-H-AB.pdf


Here's the spec sheet of the COB:
https://download.luminus.com/datasheets/Luminus_CVM-GEN4_Datasheet.pdf


Here's a crude drawing of my plan. Very basic.

1. What are your thoughts on compatibility between the LED COB and LED Driver? Seems like a good match to me, I just don't want any chance of frying the LED so would like some other eyes to confirm. Similarly, are there any precautions or additions one should make to prevent overload in any way?
2. Please let me know your thoughts on the wiring. Some might question a 10amp draw like this, and the addition of a plug, but it is very normal for my specific application.
3. I plan on implementing PWM dimming, but that is another question for another time. That is just something I will have to test.

Thank you very much!

 

[FrodoOne]

One learns something every day but it was necessary to "look it up".
A "COB" array means "Chip on Board" array.

This particular one seems to have a "forward voltage" of between 55.1 V and 58.0 V at about 5 A. (5000 mA) with a dissipation of about 300 W.

One of the three constant power drivers mentioned can supply an "Open Circuit Voltage" of 62 V at between 2600 and 8670 mA and would deliver sumwhat over 300 W

If one were to connect these two together, it might be expected that the supply would attempt to drive the COB at up to 62 V which may produce a current in the COB greater than it could handle!

The Supply/Load voltage difference is up to at least 62-55 V = 7 V at 5000 mA.
At the least, one should initially place a dropping resistor capable of dissipating 7 V at 5 A between the Source and the Load.
Since R= E/I this would be at least a 7/5 Ohm resistor (1.4 Ohms) capable of dissipating E * I Watts = 7 * 5 W = 35 Watts.

Having done this you might measure the voltage across the COB and the current flowing and decide if you can reduce the series resistor to avoid wasting some of those 35 Watts while still limiting the dissipation in the load.

While I may be wrong in this analysis, I will be interested to hear if it can be explained where I may be incorrect.

 

[MPTESE]

Hi Frodo,

Thanks for the reply. It's more appreciated than you know.


Doesn't "Open Circuit Voltage" mean when no load is connected (bare leads)? If this is true, then is it safe to assume that when the driver IS connected to a load the max voltage would be 58 (the top of the voltage region?)

 

[FrodoOne]

Hi Frodo,

Thanks for the reply. It's more appreciated than you know.


Doesn't "Open Circuit Voltage" mean when no load is connected (bare leads)? If this is true, then is it safe to assume that when the driver IS connected to a load the max voltage would be 58 (the top of the voltage region?)

That will be the Voltage but there will be nothing to limit the Current.

The voltage across the "Load" will be between 55 and 58 volts, because that is the "range" of the "Forward Voltage" drop across it.
When greater than that voltage is applied, there is (theoretically) nothing to limit the current supplied, up to the capacity of the supply to "feed" any such current - unless the supply is "current limited", which does not seem to be a property of the supply which you referenced.

Just imagine that the supply which you referenced is like a "battery" of infinite capacity which supplies 62 V.
The "Load" will conduct when the voltage applied exceeds the "forward voltage".
The only thing which can stop the current flow exceeding the capacity of your COB device would be a "limiting resistor" interposed to "drop" the voltage across it when the maximum specified current is flowing, the value of which may be calculated as I discussed earlier.


This sort of calculation is often done to calculate the value of the "dropping resistors" required to limit the current (and brightness) of "indicator" LEDs, which have a "Forward Voltage" of between 1.8 V and 3.5V, a maximum current of about 20 mA and yet may be connected across supplies of up to hundreds of volts.

 

[MPTESE]

That will be the Voltage but there will be nothing to limit the Current.

The voltage across the "Load" will be between 55 and 58 volts, because that is the "range" of the "Forward Voltage" drop across it.
When greater than that voltage is applied, there is (theoretically) nothing to limit the current supplied, up to the capacity of the supply to "feed" any such current - unless the supply is "current limited", which does not seem to be a property of the supply which you referenced.

I think I'm tagging along. The driver features a built in potentiometer - which I believe can be manually adjusted via a screwdriver. By adjusting the current in this way, couldn't this be a solution?

 

[RAL238]

Have you thought about how you will dissipate the heat from the LEDs?

 

[MPTESE]

Have you thought about how you will dissipate the heat from the LEDs?

Hey, RAL -

Yes! Please see attached spec sheets for active heatsink per COB.. Thoughts? The three three different "models" are just fan speed settings of the same unit, which you can set. I have a sample in front of me. I was going to mount a small AC-DC converter (fan voltage input is rated from 15-42) onto my build, and feed the AC side from my main plug. Then, split the DC leads to each fan.


I guess I can also attempt to tie into the DC of the led driver, but I'm sure I would need to put something in line to match the voltage rating of the fan..?

EDIT: Taken from data sheet:
"The module consists of a controller circuit, a special low noise, high efficiency, long life fan and an ultra-low thermal resistance heat sink/ The controller circuit converts any input voltage from 11 to 75 volts to a fixed lower voltage as required by the fan. The input voltage is derived from the same voltage used to power the LED array."


So probably the easiest is to just tie into the DC leads of the driver for the fan, as it automatically knocks/sets the voltage.


EDIT AGAIN:
I haven't mounted RAW LED Cobs before, so should be fun. If you have, any recommendation on thermal paste? I'm gonna use cob holders to make mounting a bit easier, and also ones that accept a LEDIL 60degree reflector to concentrate the beam spread.

 

[RAL238]

Ok, looks like you've found a good heatsink designed for the application. There are several different types of thermal compounds (metal-based, ceramic based, etc). I'm not sure what would work best in this case. The LED datasheet has a note at the end about thermal compounds, saying to check with their web site for recommendations. That's where I'd start the search.

 

[seharper]

I think the LED driver has more capabilities than you require. What you really need is constant-current drive at 5000ma output (max), right on the button unless intentionally dimmed. That happens somewhere between 51 and 58 V depending on temperature, age and binning.

So you'll need to carefully set the Meanwell into constant current mode and dial in 5000ma, and maybe put a dab of silicone on that adjustment to keep someone from changing it.

As far as the wiring, as an AC mains guy I am very, very sensitive to wiring methods. Your electrical hookup is correct obviously, but I want to see that it is housed, routed and mounted properly. In particular 4 lights need 4 cords and plugs, unless they are assembled as part of a single chandelier, with wires routed and spliced in internal spaces in the chandelier.

 

[MPTESE]

I think the LED driver has more capabilities than you require. What you really need is constant-current drive at 5000ma output (max), right on the button unless intentionally dimmed. That happens somewhere between 51 and 58 V depending on temperature, age and binning.

So you'll need to carefully set the Meanwell into constant current mode and dial in 5000ma, and maybe put a dab of silicone on that adjustment to keep someone from changing it.

As far as the wiring, as an AC mains guy I am very, very sensitive to wiring methods. Your electrical hookup is correct obviously, but I want to see that it is housed, routed and mounted properly. In particular 4 lights need 4 cords and plugs, unless they are assembled as part of a single chandelier, with wires routed and spliced in internal spaces in the chandelier.

Thank you sir. I should get the driver in a couple days to test. The cob should come in a couple weeks, but maybe I can get my hands on a similar cob within the same voltage range.

Concerning wiring and application, I should have provided more context. I am attempting a prototype fixture for filmmaking, following the style of brute lights or stage blinders (a quad array). With this application, it’s advantageous to have the driver’s ac leads joined to a singular plug. I got a bunch of wago connectors vs wire nuts this time..

I can’t find a blank enclosure for my purpose, atleast not yet with my minimal funding. So for this build I am starting out with a frame made out of aluminum angle. Almost similar to “diy grow lights” if you’re familiar with that group. However, mine will be much more sturdier and probably thicker metal. It will also have a rear and a front to mount the driver/heatsink on both sides, with cross bracing. This keeps the maximum airflow for the heatsink, but I’m still thinking through hiding and tidying the wiring. The AC drivers will join inside a gang box that will be mounted to the frame prior to the single wire that attaches to the receptacle. I hope to have a drawing soon instead of my words, but it’s hard without all the components in front of me. That being said, should I split the ground from the ac side and bolt it to the frame?

 

[seharper]

Concerning wiring and application, I should have provided more context. I am attempting a prototype fixture for filmmaking, following the style of brute lights or stage blinders (a quad array). With this application, it’s advantageous to have the driver’s ac leads joined to a singular plug. I got a bunch of wago connectors vs wire nuts this time.

Well! Gaffing is a whole different world with its own rules. I am not familiar with them, but they take them very seriously.

One rock band, they had a 100 page rider in their contract, which called out a bunch of things the band wanted, including an M&M bowl with all the brown M&Ms removed, and exactly how to do the wiring safety for their complicated light show. As soon as they got on-site, they checked the M&M bowl. If they did not see an effort to remove the brown ones, that means the venue did not read the rider, and they went over the lighting with a fine tooth comb.

This keeps the maximum airflow for the heatsink, but I’m still thinking through hiding and tidying the wiring. The AC drivers will join inside a gang box that will be mounted to the frame prior to the single wire that attaches to the receptacle. I hope to have a drawing soon instead of my words, but it’s hard without all the components in front of me. That being said, should I split the ground from the ac side and bolt it to the frame?

The only thing that concerns me is the idea of having the driver far from the LED. Cable length won't be an issue; the constant-current driver will override any voltage drop. Just you'd want to have plugs/sockets, and you'd need something totally unique, like a NEMA 2-20 or some sort of obscure Molex connector. Maybe DIN-7 lol, I don't think anyone uses that for anything!

Yeah, if your grounds are any good, you're better off grounding the metal frame moreso than the lamp itself. If you're doing field work, then you have to put up with varying quality of electrical, and you can't trust some random ground in a 70yo house.

 

[MPTESE]

Well! Gaffing is a whole different world with its own rules. I am not familiar with them, but they take them very seriously.

One rock band, they had a 100 page rider in their contract, which called out a bunch of things the band wanted, including an M&M bowl with all the brown M&Ms removed, and exactly how to do the wiring safety for their complicated light show. As soon as they got on-site, they checked the M&M bowl. If they did not see an effort to remove the brown ones, that means the venue did not read the rider, and they went over the lighting with a fine tooth comb.



The only thing that concerns me is the idea of having the driver far from the LED. Cable length won't be an issue; the constant-current driver will override any voltage drop. Just you'd want to have plugs/sockets, and you'd need something totally unique, like a NEMA 2-20 or some sort of obscure Molex connector. Maybe DIN-7 lol, I don't think anyone uses that for anything!

Yeah, if your grounds are any good, you're better off grounding the metal frame moreso than the lamp itself. If you're doing field work, then you have to put up with varying quality of electrical, and you can't trust some random ground in a 70yo house.

Hey Seharper, thanks for the reply.


I'm not sure what you mean by having the driver far from the LED. The leads going from the driver to the LED will be about about 12-14" long? Why would you guess I need something unique? I'd like to use a simple 5-15 plug for the AC side (and not any 125/250v) and split them to each driver input.

 

[seharper]

Oh, that'll be fine. I'm sorry, I don't know why I thought something else.