Using "Smart plugs" to dim constant current drivers not designed for ac side dimming?

Thread Starter

jeffkrol

Joined Dec 8, 2015
56
There is a trend to use smart plugs as a dimmer with things it, at least to my understanding, shouldn't dim.
These smart plugs work on the AC side and will dim leds that are designed to be dimmed on the ac side.
Apparently some found that they also dim constant current drivers .. on the ac side.
Makes no sense to me but I'm less than a rookie here.
My limited understanding that the smart plugs do some work using null crossing point on the ac side.
It also seems to me that this will basically shut the constant current driver off on the ac side.. at what 60-ish Hz?
Doesn't seem like a smart way to go.
https://www.reef2reef.com/threads/want-to-supplement-my-radion-blues-with-white-light.1047635/page-3
See #54 for the plug that "dims" a Constant current driver.
 

Ian0

Joined Aug 7, 2020
10,232
Some constant current drivers are capable of measuring the on-time of the phase-controlled AC input, and varying the output current accordingly.
It should be noted that there is no standard for dimmability. The only load that is definitely dimmable under all circumstances is a filament lamp that operates at the mains supply voltage. Anything else may, or may not be dimmable.
 

Thread Starter

jeffkrol

Joined Dec 8, 2015
56
Some constant current drivers are capable of measuring the on-time of the phase-controlled AC input, and varying the output current accordingly.
It should be noted that there is no standard for dimmability. The only load that is definitely dimmable under all circumstances is a filament lamp that operates at the mains supply voltage. Anything else may, or may not be dimmable.
Thanks.. The drivers in question are not designed for AC dimming..or any dimming whatsoever.
Why an AC side dimmer " works" on it is
???

Afaict the AC side is basically a transformer , diode rectifier, capacitor type thing.
Dc side is just a switching dc w current regulation via voltage bucking.

I suppose the simplistic way to put it is the smart switch just turns the driver off and on thus " dimming" what shouldn't be edimned that way?
And effects on a driver that cycles like 60Hz on/off or some such duty cycle.
 

schmitt trigger

Joined Jul 12, 2010
950
Imposible to tell precisely, without a device tear down and some scope waveforms. Like Big Clive or DiodeGoneWild does.

It may be that it id dimming the circuit by brute force and applying undue stresses on components.
 

Thread Starter

jeffkrol

Joined Dec 8, 2015
56
Imposible to tell precisely, without a device tear down and some scope waveforms. Like Big Clive or DiodeGoneWild does.

It may be that it id dimming the circuit by brute force and applying undue stresses on components.
Yea the stress thing is my concern as well.
Like take a simple ac/dc switching power supply and power cycle it 60 times / sec or something like that.
What are the effects?

I'd love some waveforms. Particularly on the output of a dimmable smart switch..

I believe part of the answer is here..
https://www.makeuseof.com/how-to-make-diy-smart-wifi-light-dimmer-switch-zcd
At least as to how the smart switches work.
 
Last edited:

LowQCab

Joined Nov 6, 2012
4,290
Some devices, especially the really cheap ones, can be smoked by "Chopped-AC".

"Chopping" the AC can create all sorts of
nasty "Spikes" and "Hash" that can damage unprotected Electronic-Circuitry.

It can also "broadcast" RF-Noise into other nearby Electronics.
.
.
.
 

ronsimpson

Joined Oct 7, 2019
3,200
There are many different ways of making the new types of light bulbs. If the bulb says not dimmable, you should believe it.
Many bulbs are built to make a constant light output and will fight you, and maybe not survive for long.
 

BobTPH

Joined Jun 5, 2013
9,254
Thanks.. The drivers in question are not designed for AC dimming..or any dimming whatsoever.
Why an AC side dimmer " works" on it is
???
I suspect the dimmed AC output of the smart plug does not allow the constant current driver to get enough power to provide the current it should and it simply falls out of regulation.
 
Last edited:

Ian0

Joined Aug 7, 2020
10,232
There are many different ways of making the new types of light bulbs. If the bulb says not dimmable, you should believe it.
Many bulbs are built to make a constant light output and will fight you, and maybe not survive for long.
What seems to happen is that at a low voltage the power supply doesn't start, from about 160V to 255V they run as normal, and somewhere in the middle they will strobe as the power supply starts and stops on its UVLO settings. Starting up is the most stressful time for a power supply.
 

Thread Starter

jeffkrol

Joined Dec 8, 2015
56
There are many different ways of making the new types of light bulbs. If the bulb says not dimmable, you should believe it.
Many bulbs are built to make a constant light output and will fight you, and maybe not survive for long.
Thanks but of course that never stops anyone.

See some found that the Quanta LED light bar constant current driver "responds" to a smart switch with dimming.

I've seen people do the same on the output side of the dc driver.. "chopping" the constant current output via a mosfet switch.
Thing is in that case some drivers go to full voltage in a no load situation.
Then when "turned on" by the mosfet need to throttle the voltage down (usually) to get to the current set point.
Assume a driver has a range of say 20-56v. The led string voltage at say the set current of 500mA is 32v.
Driver starts at 56v then throttles to 32v (all approx and for demonstration) VERY briefly the led string sees 56V and will see it at whatever the duty cycle of the mosfet is..stressing the led string.
 

Thread Starter

jeffkrol

Joined Dec 8, 2015
56
Stray inductance and capacitance would likely prevent any damage. A voltage or current never changes instantaneously.
I can give you a real life example ( well real life hearsay) that may not invalidate what you said but to show you one possibility:
Not too long ago building led lights for saltwater aquarium was the rage (still done but not as much)
A common topology was a bunch of Cree royal blue 3w class emitters strung together in series.
A common driver was a Meanwell ELN-60-48D. 24-48v, 1.3A

If one hot swapped the string to a plugged in driver there was a good chance you'd pop some LEDs. I assume from the voltage and concurrent high amp draw.
As you can see the diodes at 1.3A may already be stressed. Just get pushed over the edge.
Anyways most learned not to use current that high.


As stated in here, you don't want to go open circuit, as the drivers loop control can be too slow before LED damage occurs, especially if you have miles of wiring (inductance).
Now all that said I've personally added led strings to a " live" circuit consisting of Meanwell ldd dc/dc buck driver, 56v switching power supply to a string of LEDs with a combined v( f) of like 12v.
Never popped one.
May be driver topology dependent.

Getting off topic a bit
 
Top