DIY with Quantum Boards

pop22

pop22

Well-Known Member
Just scored 4 more Diablo boards! Going to replace my pre-production QB 304s with them. I run my current Diablo light at 400 watts, and even at that level it will stomp most lights out there. Based on HLG's specs, I estimate I'm getting 1000ppfd at the canopy. Not sense in running it any harder, I'm am not running CO2 and don't really plan to, so my plants are maxed on light. Think I'm going to ebay the 304s, and my other spare boards and drivers.

And I'm going to really put the Diablo to work this time. As soon as I get my other 4x4 cleared, I'll move the three plants in 3 gallon pots and the thirty 1 gallon pots will stay under the light.


Diablo light pic1- 11-29-2020.jpg
 
2

2com

Well-Known Member
I wanna post this question here to, just in case some QB648 diablo users/owners might see it, or anyone else can help out:

I see a list of driver options for the QB648 diablo on the few sites that still list them for sale. I don't see any options for CV drivers though, except the "HLG-120H-54" under the advisory "Following driver are Compatible but will require some dimming under continuous operation.:"

Would an hlg-320h-54 run the QB648, and if so, how many?

The 648 fv is 56V, and 5000mA max current.

Thanks.
 
pop22

pop22

Well-Known Member
here is something a little different for you. Found her in the greenhouse raised bed hiding under my Auto Mazar. Transplanted her into a windowbox planter about the middle of October. She's under my converted T5 light, I replaced the T5 bulbs and used the housing and reflector to mount my HLG ( uhh 48? ) two 24" strips running at 60 watts for the pair. I wish they still made them, I really wanted to ad 2 more of these strips! This girl shure likes it!


Strawberry plant pic1 -11-29-2020.jpg
 
pop22

pop22

Well-Known Member
The only way to run the Diablo maxed out would be with a custom heatsink. With HLG's heatsinks your limited to 150 watts per board.

I don't know why they limit the driver recommendations other than to discourage someone thinking they could run a single board on that driver.

I run my light on a pair of HLG240H-C2100 and believe me, 4 Diablo running at 400 watts is more light than you need without CO2! I dimmed mine to 400 watts just to see. It's going to stay there! It's still putting out damned close to as much light as a single ended HPS, and at 400 watts! HLG brought us the promised LED based HPS killer!


2com said:
I wanna post this question here to, just in case some QB648 diablo users/owners might see it, or anyone else can help out:

I see a list of driver options for the QB648 diablo on the few sites that still list them for sale. I don't see any options for CV drivers though, except the "HLG-120H-54" under the advisory "Following driver are Compatible but will require some dimming under continuous operation.:"

Would an hlg-320h-54 run the QB648, and if so, how many?

The 648 fv is 56V, and 5000mA max current.

Thanks.
Click to expand...
 
2

2com

Well-Known Member
pop22 said:
The only way to run the Diablo maxed out would be with a custom heatsink. With HLG's heatsinks your limited to 150 watts per board.

I don't know why they limit the driver recommendations other than to discourage someone thinking they could run a single board on that driver.

I run my light on a pair of HLG240H-C2100 and believe me, 4 Diablo running at 400 watts is more light than you need without CO2! I dimmed mine to 400 watts just to see. It's going to stay there! It's still putting out damned close to as much light as a single ended HPS, and at 400 watts! HLG brought us the promised LED based HPS killer!
Click to expand...
I'm not trying to run it maxed out or anything like that.
I'm just wondering, first, why there are no constant voltage drivers in the recommended drivers lists (except the one suggestion for an HLG-120H-54 for a single board)? Maybe daisy chaining board to board in parallel is dangerous for these boards, or maybe it has something to do with the reverse polarity issue they have/had.

The 650R runs off an inventronics 600+w driver (four boards). The 350R runs off 320w driver. So yea, I guess 150 watts per board is the limit/max, which is good. But I remember these boards had some slight differences in them, and the drivers they can function with. I'm wondering if there's an issue with constant voltage drivers and more than one board maybe? Probably not.

So, I want to make sure an hlg-320h-54 safely run a pair of qb648s, and would it run four of them?

Thanks.

Edit: I see upon closer look that the qb288 is 54V, and the qb648 is 56V. That probably has something to do with it..?
 
Last edited:
S

Sinfor

Well-Known Member
guitarguy10 said:
I had found this datasheet:

Which says Dimming Capability: Yes ... but this shit is confusing to me in that I'm unfamiliar with how to interpret this stuff. My comprehension of electronics spans V = IR, P = I^2R and some very simple circuits with an arduino microcontroller, so I need help :).
Click to expand...
Your driver can't be dimmed, the datasheet provide you specs and ratings for all types. Your driver can work on CV or CC mode depending on the load, with those QBs and a 2S2P array the driver will work in CC mode and looking to the specs you are going to get around 167w.
Also, if you connect them all in parallel you can have some kind of dimming, like <50% for veg stage.
 
ilovereggae

ilovereggae

Well-Known Member
Sinfor said:
Your driver can't be dimmed, the datasheet provide you specs and ratings for all types. Your driver can work on CV or CC mode depending on the load, with those QBs and a 2S2P array the driver will work in CC mode and looking to the specs you are going to get around 167w.
Also, if you connect them all in parallel you can have some kind of dimming, like <50% for veg stage.
Click to expand...
can u explain the last part? since the qb120s are 24v and driver is 48v how could you wire them all in parallel and how would the dimming work?

I was also suggesting to @guitarguy10 in a PM that if using the 2S2P layout he might want to add fuses or resistors to each parallel branch in case one failed so that the the other one doesn't get too much current. am I wrong to think this? I know most builds on RIU ive never seen anyone use them but I know LED Gardner suggests u add them.
 
ilovereggae

ilovereggae

Well-Known Member
also @Sinfor also one other question. on the QB120 page they show the diagram this way. a series circuit that has 2 boards connected in parallel to each branch. also i admit i'm still learning but doesn't doing it this way cause the first boards to have to carry the voltage for the 2nd, and result in the last boards in the parallel chain being slightly dimmer?

SeriesParallel_grande.png



but the Qb288 page shows 2 parallel branches with 2 boards in series on each branch. I think that this way makes more sense in my mind, but I guess it doesn't matter. Although from my math I thought this would let him get the full 180W. this is also why I thought he should maybe add a fuse on the parallel branches.

Boards-wire3_large.png
 
Last edited:
pop22

pop22

Well-Known Member
It could be, remember that voltage stated is average, and the forward voltage rises as the current goes higher.


2com said:
I'm not trying to run it maxed out or anything like that.
I'm just wondering, first, why there are no constant voltage drivers in the recommended drivers lists (except the one suggestion for an HLG-120H-54 for a single board)? Maybe daisy chaining board to board in parallel is dangerous for these boards, or maybe it has something to do with the reverse polarity issue they have/had.

The 650R runs off an inventronics 600+w driver (four boards). The 350R runs off 320w driver. So yea, I guess 150 watts per board is the limit/max, which is good. But I remember these boards had some slight differences in them, and the drivers they can function with. I'm wondering if there's an issue with constant voltage drivers and more than one board maybe? Probably not.

So, I want to make sure an hlg-320h-54 safely run a pair of qb648s, and would it run four of them?

Thanks.

Edit: I see upon closer look that the qb288 is 54V, and the qb648 is 56V. That probably has something to do with it..?
Click to expand...
 
G

guitarguy10

Well-Known Member
I'm trying to figure out the correct way to wire 4 boards that are 2 series 2 paralel, to put in series you just connect the + to -ve of the next panel, and to put in parallel you just combine the +'s and combine the -ves of the series elements.

Is this correct? I made a pic, ignore the wire colours, it's the directions that are important. If it's not correct what is the correct wiring and why am I wrong?

I am wiring them like this because the driver that I have is an ELG-240-48, which is non-dimmable (so CV and CC (I believe at least, because its actually 'blank', not A or A/B) output 48V, 3.75A (180W) on 120V main (it also has an output for european 220V of 48V, 5A (240W) but I live in North American 120V).

The boards are QB120 v2 which are each 24V max 2800mA (70W) each so putting 2 in series would yield 48V* and putting those two sets in parallel would maintain the 48V since parallel splits current and not volts (right?). How would the currents work out in this circuit, like the 2 in series would they consume 2.8A** x2 = 5.6A but since there are two of these split in parallel the current is divided by 2 for a current draw of 5.6A / 2 = 2.8A (and like 2.8/3.75= 74.66% efficiency)?

Please don't mock if I'm completely wrong, I am struggling to find *specific* information on how to plan and calculate for a DIY led board configuration, the only videos I am finding show HOW to wire them, but not why or what values were used to determine which driver and/or wiring config (series, parallel, seriesparallel etc.) to use. I am leaning on my decades old knowledge of ohms law here lol

*rated output of my driver is 48V and 3.75A
**max rated current for each board
 

Attachments

Last edited:
S

Sinfor

Well-Known Member
ilovereggae said:
can u explain the last part? since the qb120s are 24v and driver is 48v how could you wire them all in parallel and how would the dimming work?
Click to expand...
QB120 specs
Screenshot_20201127-211658_01.png
V-I derating curve with that specs will look like this
IMG_20201129_205856_01.jpg
To put the ELG to work in CV mode, the current drawn by the load should be lesser than the max. rating current at 48V or the load @3.75A must need a voltage higher than 48V, two boards in serie @1.875A=44.48V

4 boards in parallel needs around 21.73V
21.73Vx0.94Ax4=81.6W
8160/166.8=48.9% dimming

ilovereggae said:
I was also suggesting to @guitarguy10 in a PM that if using the 2S2P layout he might want to add fuses or resistors to each parallel branch in case one failed so that the the other one doesn't get too much current. am I wrong to think this? I know most builds on RIU ive never seen anyone use them but I know LED Gardner suggests u add them.
Click to expand...
You are right if one series goes off, actually the left one is going to take all the current. If you extend the curve I made, @3.75A the voltage should be less than 48V.
Instead of fuses or resistors I would use the remaining voltage adding to each string three royal blue in parallel. Xpe-2 @625mA drops around 3.1V plus 44.48V gets you close to 48V, if one fails the driver is going to enter in CV mode.
Fuses are more complicated, they can trip at currents higher than the rated ones or can take a while before blowing or even not blowing at all.
For a limiting resistor: 3.5V/1.875=1.86ohm, the closest value is 1.8ohm. 15W
 
ilovereggae

ilovereggae

Well-Known Member
Sinfor said:
QB120 specs
View attachment 4755508
V-I derating curve with that specs will look like this
View attachment 4755512
To put the ELG to work in CV mode, the current drawn by the load should be lesser than the max. rating current at 48V or the load @3.75A must need a voltage higher than 48V, two boards in serie @1.875A=44.48V

4 boards in parallel needs around 21.73V
21.73Vx0.94Ax4=81.6W
8160/166.8=48.9% dimming


You are right if one series goes off, actually the left one is going to take all the current. If you extend the curve I made, @3.75A the voltage should be less than 48V.
Instead of fuses or resistors I would use the remaining voltage adding to each string three royal blue in parallel. Xpe-2 @625mA drops around 3.1V plus 44.48V gets you close to 48V, if one fails the driver is going to enter in CV mode.
Fuses are more complicated, they can trip at currents higher than the rated ones or can take a while before blowing or even not blowing at all.
For a limiting resistor: 3.5V/1.875=1.86ohm, the closest value is 1.8ohm. 15W
Click to expand...
@Sinfor thank you. i forgot the account for the boards drawing less voltage at lower current. also thank u for the explanation of how the CC/CV drivers work, I didn't understand how to calculate when it shifted between modes.

@guitarguy10 this should help explain things a bit better. tbh this is a bit more complicated of a first time DIY build than I expected.
 
G

guitarguy10

Well-Known Member
ilovereggae said:
@Sinfor thank you. i forgot the account for the boards drawing less voltage at lower current. also thank u for the explanation of how the CC/CV drivers work, I didn't understand how to calculate when it shifted between modes.

@guitarguy10 this should help explain things a bit better. tbh this is a bit more complicated of a first time DIY build than I expected.
Click to expand...
I'm actually more confused now lol. If the max rated for each board is 2.8A then why would 2 boards in series draw 1.875A (instead of 2.8A x 2 = 5.6A)? and *how* would they do this, I can't control the current on this driver?

Also:
"Instead of fuses or resistors I would use the remaining voltage adding to each string three royal blue in parallel. Xpe-2 @625mA drops around 3.1V plus 44.48V gets you close to 48V, if one fails the driver is going to enter in CV mode. "

What? Xpe-2? Is that a COB? is it to act as a sink for current if a board fails instead of a fuse? How would I wire this in parallel? Parellel with what specifically?

Isn't the driver that I have alerady in a constant voltage (48V) mode?
 
ilovereggae

ilovereggae

Well-Known Member
guitarguy10 said:
I'm actually more confused now lol. If the max rated for each board is 2.8A then why would 2 boards in series draw 1.875A (instead of 2.8A x 2 = 5.6A)? and *how* would they do this, I can't control the current on this driver?

Also:
"Instead of fuses or resistors I would use the remaining voltage adding to each string three royal blue in parallel. Xpe-2 @625mA drops around 3.1V plus 44.48V gets you close to 48V, if one fails the driver is going to enter in CV mode. "

What? Xpe-2? Is that a COB? is it to act as a sink for current if a board fails instead of a fuse? How would I wire this in parallel? Parellel with what specifically?

Isn't the driver that I have alerady in a constant voltage (48V) mode?
Click to expand...
so to answer the first question, current is pushed, not pulled. (at least with LEDs that how I think of it. maybe in other circuits where the load dictates how much current it tries to pull this wont work, but with LEDs they will only use what they are supplied). in your case at. 110V you only get 3.75A out of your driver, so that is going to get split when you do the parallel branches to 1.875A. So that is what you will be sending each board.

With that in mind, looking at the HLG spec sheet like Sinfor pointed out, the boards actually dont take a full 24V at lower currents. Yours will only draw 21.73V each board at 1.875A.

Now for the more confusing part. The driver you have actually can run in either CC or CV mode. The datasheet shows a curve of where it will switch over. In your case since you aren't drawing enough voltage it will stay in CC mode instead of CV.

That is the best I can explain it, if someone else can simplify please do so!

Like I said earlier, this is actually one of the more complex DIYs we've seen since we are trying to fit a square peg into a round hole to make this ELG work. It can do the job, but a properly matched driver would make this be a much simpler build.
 
ilovereggae

ilovereggae

Well-Known Member
Hey so I have a question regarding @guitarguy10 and his qb120 dilemma.

What do you all think about running 2 in series on a Delta LNE-48V150WDAA? This would send 3200ma to the circuit which is a little overpowered but if he dimmed them to 2800ma he could run at 65/70W each no problem. this would be a really easy build and low enough voltage to be "safe".

The other option would be to use the LNE-24V120WAAA. These would require being setup in parallel tho. Not sure if he should consider adding resistors or fuses at that point to protect from one board dying. I dont know if I'm overthinking this bc HLG doesn't mention it, but LED Gardner has me rethinking that one.

I got a bunch of both of these drivers on blowout from Arrow for $11 each a few months ago. I found out about the sale from RIU so figure I should pay it forward and help this guy out so his first experience w QBs isn't so frustrating.

If you all agree this is a good setup, I will drop 2 in the mail before the end of the week.
 
You must log in or register to reply here.
Top