-
Autor
Some time back I bought a dozen of the Plus 0-10 v dimmers to control Meanwell PWM LED drivers. The Meanwell dimming input (according to the datasheet) accepts a variable resistance (so it sources voltage) and sinking PWM from the dimmer but also accepts what it calls "additive DC voltage" and "additive 0-10 volt PWM" on the same input (not at the same time, of course). In other words it also sinks sourced PWM and DC voltage as additional control modes.
I want to be energy efficient and not have the LED drivers energized by 120 volts when they are not in use (even though standby power draw for each dimmer is less than 1/2 watt for each driver. ) Unfortunately the Gen 3 1/0-10 v dimmer is not available in the US and in any case I already have the Plus sinking version. I don't need power measurement or want the extra heat from a shunt resistor inside the Shelly case (even 1 milliOhm). The dimmers are in a central panel so it's no problem to add extra components. Assuming it works the components for the solution below will cost under $5 per dimmer/driver, including the LDO 10 volt regulator that cost under a dollar each. (It's rated at 150 ma so I don't need one for each dimmer/driver. Probably I'll put a small resistor on its output to ensure it's not overloaded if mutltiple dimmer/drivers are activated at once).
I have Panasonic 2 Amp Solid State AQG-22105 Rectifiers that can be turned on by 4-6 volts DC. They have an input imedance of 330 ohms (rather low I thought) and seem to want 10 ma at 5 volts. I also have a source of 10 volts to convert the Plus 0-10v dimmer to a source of PWM voltage. I can't use the PWM drivers as a source of 10v because they will be off normally when the Shelly dimmer wants to turn them on.
This is a little complicated to arrange given what thgoebel in the Plus 0-10v teardown thread advises regarding placing a 3.3kOhm resistor in series with the 10 volt source. Using the usual RC network with this resistor in series would seem to delay the turning on of the relay through the usual sort of RC network that is also needed to keep the relay turned on during as little as a 6% PWM duty cycle. (The Meanwell drivers dim to zero below 6% PWM). I explored using a PNP transistor fed with the inverted PWM signal from across the 3.3k resistor as suggested by thgoebel but I was not confident I could get the timings right and avoid all the drivers being turned on briefly on system power on -- also there would be a constant current flowing in the off condition in most of the configurations I examined, negating the purpose.
So to isolate the RC timing from the 3.3k resistor I decided to use a CMOS integrated circuit buffer -- 1/6 of a hex CD74HC4050E that happily accepts more than 10 volts on its input. It outputs 5 volts at up to 25 ma for charging a 2 microfarad capacitor through a 220 ohm resistor. Then this voltage is fed through a 2.2kOhm resistor to the base of an NPN transistor which is placed between the - terminal of the SSR and ground. The transistor only needs a milliamp at its base to cause it to reliably conduct the 10 ma needed for the relay. I've checked that there will be enough voltage and current at the transistor base to keep it on during the minimum 6% duty cycle and the relay should turn on within a full cycle at 60 Hz (one half cycle for the transistor to conduct even at the worst case 6% PWM input, and another to be sure the zero-cross relay turns on). This seems like a tolerable delay.
Here's a diagram. I'll report back after I've tested it.
.
