Question about PWM dimming

[sfsuphysics]

So my main tank is dimmed via the Storm controller (really not a fan of this controller), but my 40G and 20G frag tanks are simple on/off schedules. I would like to tie them all into a controller (since PWM controllers for LEDs dont' seem to be particularly cheap).

However one question that has been bugging me, is that the LDD drivers connect to the controller and one of the legs connects to the neutral as a "reference point" (I'm guessing?). How would that work out if I just pulled some Cat5 from my others tanks to the controller, they would not be sharing the same neutral, different power supplies as well as different circuits, would dimming just not work in that case?

Is there a way I can make this work? I know in the past with multiple power supplies I could tie the neutrals of each Meanwell PSU together and that seemed to work, but I'm wondering if that's advisable when they're on different circuits.

 

[mcgyvr]

You tie the grounds together....
"Neutral" is the AC side and not involved here...you want a common DC side ground (not chassis ground) so one signal is not floating at a different potential vs the other..

 

[sfsuphysics]

Well damn, all this time I tied the controller to V- not ground. Guess I'm lucky I didn't screw it up.

Ok tying all the grounds together of different power supplies, presumably this is not going to be carrying any real power (at least as long as things are going good) so I can simply piggyback off the Cat5 for the single to the LDD drivers?

 

[mcgyvr]

Umm.
V- may be the grounds you need to tie together...
On meanwells Ive used im the past v- is the main supply gnd/return and should be grounded together

 

[sfsuphysics]

Ok guess V- is what I was supposed to tie to, my vernacular was just screwy with my calling V- = Neutral. Tried to connect the GND to AC ground and that did nothing for it.

It's a hot day here in SF (for SF standards) I'm not thinking straight.

 

[mcgyvr]

Yes...all v- must be common

 

[perkint]

Since mcgyvr has pretty much answered the technical bit, just to clarify the simple bit - yeah, you can run multiple independent LEDs from one dimming signal (ie one controller) by sharing the signal round (so long as you tie the -ve/GND/Vin together) - someone did the maths a while back (might even have been mcgyvr?) and I don't remember the exact number but it was something silly like 100 LDDs should be OK from a single ATMega pin (which I think is what the storm is based on).

Tim

 

[sfsuphysics]

100 LDD drivers per pin? Yikes. Well I will have (short term) 4 per pin, so I think I'm well within those levels

So since I got people here, on a somewhat related note, if I wanted to do less white on a tank, could I just put in like a LDD 350 and the dimming will be the same at roughly 1/3 the output power?

 

[mcgyvr]

Lowering the current will decrease the brightness of an led...

 

[sfsuphysics]

I know that, I was just wondering if the PWM worked the same way, i.e. if I had say 50% max on the white channel via the drivers it tells the LDD driver to only put out 50% power, regardless if it's different max powers.

I'm sorry, I know this sounds like a stupid question.

 

[mcgyvr]

During 50% PWM dimming the current output of the driver is not reduced.. Its only on 50% of the time..

Its turned on/off quickly and averages out to be ~equivalent to 50% power..

 

[perkint]

In very tough terms, if 100% PWM on 1000mA is 1000 lumens
50% PWM on 1000mA is 500 lumens
50% PWM on 500mA is 250 lumens

Sort of. But hopefully that helps clear it up?

Tim

 

[oreo57]

Not to get too technical or obfuscate the simplicity but there is a minor GAIN of output w/ lower currents.

Like 100% at 1000mA = 1000 lumens
100% at 500mA is 550 lumens..

Generally watt efficiency (L/watt) increases as current (and concurrently heating) decreases.

Same w/ cooling
100L/ w at 60C or 110L/w at 45C

Minor thing but................................and not sure "exactly" how PWM affects that..does 50% cool and gain?
I suspect so.

 

[perkint]

Yeah, I was speaking roughly. As you say, 2 LEDs at 500mA gives more light, less power and less heat than 1 LED at 1000mA. But, I was ignoring that to try and give a simple answer...

Tim

 

[perkint]

Minor thing but................................and not sure "exactly" how PWM affects that..does 50% cool and gain?
I suspect so.

Yes, but 100% at 500mA is better than 50% at 1000mA. I can't remember how much, but I'm sure one of the guys (Ibraheem) measured it on UR. I'll see if I can dig any figures out.

Tim

 

[lingwendil]

Simply put, yes, you can run a cat 5 cable to the other tanks, and use it to supply the PWM dimming signal to the other fixtures. I run a cable about 30 feet around the perimeter of a large room to supply the fixture on my nightstand freshwater tank with four channels of control (from perkint's controller design, actually ). Since I only have four channels, I wired headphone connectors at the end of two pairs each of the wire in the cat5 sleeve. blue/blue white and orange/orange white is channel 1 and 2, green/green white and brown/brown white are channel 3 and 4. I tied the white of each pair to V-/ground. The fixture uses a pair of regular 1/8" stereo headphone jacks to receive the signal. Nice, straightforward, and tidy.

You could do up to seven channels if you tie only one of the whites to ground for them, but this way worked for my four channel use, and allowed for a nice clean cable with two connectors neatly heatshrinked at either end.