1

I have similar situation, described here: MeanWell 3-in-1 dimming So, I need to control MeanWell power by PWM signal from Arduino. I tried to simulate scematics, given as answer in that topic, but it seems doesn't work (in simulator).

Schematics:

Schematics1

Results: Results1

As author of that topic mentioned in one of comments, Vg is about -0.47V, which is not enough to close BSS126. One extreme change is to remove 750k resistor.

Schematics:

Schematics2

Results:

Results2

As you can see in this case Vg = -1.4V, which is still less, than required to close BSS126 comletely. I guess because of that maximum Vdim+ is not higher than 5.6V.

I'm quite bad at electronics and already spent a lot of time fighting with this schematics and trying to get 10V if there is no input signal for optocoupler. Can someone advise, what is wrong with this schematics and how to make it work? The goal is to have 0V when there is no input signal and 10V when input signal is 5V.

My MeanWell device is XLG-240-M-AB. When nothing is connected to DIM+ and DIM-, voltage is 14.4V. When I connect resistor 99.7 kOhm, voltage is 9.62V. Which means (if I'm not mistaken) that internal resistor inside MeanWell (please refer to schematics in topic, which I refered to previously) is about 50 kOhm

SamGibson
  • 17,231
  • 5
  • 37
  • 58
Alex
  • 9
  • 4
  • This will not work, due to the fact that the pullup is useless in this situation connected to drain. Connect pullup to some stable rail, because drain goes down, every time You turn on a mosfet. – fifi_22 May 05 '21 at 17:22
  • 1
    What way do you want it to work when your microcontroller is powered down? Full current or zero current? – Transistor May 05 '21 at 17:23
  • @fifi_22, sorry, what do you mean by saying "Connect pullup to some stable rail"? Do you have link to any example schematics? – Alex May 05 '21 at 17:25
  • It’s not an NPN OPTO, but a FET, is pin4 to gate and pin 3 gnd? Open circuit voltage at IF = 10 mA, 8.4 V typical • Short circuit current at IF = 10 mA, 15 μA typical and I assume power FET is a low side driver – Tony Stewart EE75 May 05 '21 at 17:26
  • @Transistor, I would like to keep MeanWell completely off if there is no signal from Arduino. Which means 0V (or close to it) between DIM+ and DIM-. – Alex May 05 '21 at 17:27
  • @TonyStewartEE75, yes, I borrowed schematics from topic I refered to. Author mentioned that "this is an expensive and obscure solution but solid-state". I used it simply because I have no other options :) – Alex May 05 '21 at 17:30
  • Both opto an FET are Depletion mode. You understand? – Tony Stewart EE75 May 05 '21 at 17:33
  • @TonyStewartEE75, thanks for pointing on this! I still not 100% understand, but at least I noticed that this is very unusual optocoupler (if we can call it optocoupler). I will keep reading manuals to get more understanding on how it works. – Alex May 05 '21 at 20:22
  • 1
    Thanks to @TonyStewartEE75 I noticed that my VOM1271 doesn't have enoug input current. I increased it to 10 mA and now both options of scematics, provided by me above, started working. 1st gives 8.3V, 2nd - 8.7. Stil lower than 10V. Any ideas how to increase it to 10V (if it is possible)? – Alex May 05 '21 at 20:28
  • You did’t want full brightness so use 0 to 8.7 as 0 to 87%. Did U confirm pin3 is gnd? – Tony Stewart EE75 May 05 '21 at 20:42
  • @TonyStewartEE75, pint 3 if connected to gate of BSS126, pin 4 - to GND. As I understood original idea, BSS126 needs negative voltage on gate to close it. – Alex May 05 '21 at 21:06
  • Actually I would like to have 100% maximum and control it from microcontroller. – Alex May 05 '21 at 21:17
  • I almost get 10V! What I did (in simulation) is installed 100 kOhm resistor between DIM+ and DIM-. If PWM frequency is extremely low (ex. 2 Hz) it gives range 0-9.7V, exactly what I need! But when I use 500 Hz PWM, result is only 6.9V. I simulate input of MeanWell as shown in answer of Transistor: there is RC-filter. I used C1 = 100 nF and R1 = 75 kOhm. I'm not sure about exact values of my PSU, so I took these values from Transistor's answer and from topic I refered to. I guess I have lower results just because current source 0.1 mA charges C1 with very low speed. – Alex May 05 '21 at 21:17
  • Won’t the lights flicker at 2Hz ? I guess you can put an e-cap across it. For 1% ripple C=Ic dt/dV = 100uA 1s 100/10V= 1mF not good because of leakage R – Tony Stewart EE75 May 05 '21 at 21:29
  • @TonyStewartEE75, it will not flicker, I already tried that. Moreover MeanWell's datasheet recommends using 100Hz - 3Hz for this PWM. As I understood this PWM is not used to feed LEDs. I guess it just used to create voltage using RC-filter, as described on Transistor's schematics. MeanWell's dtaasheet mentions another "PWM fosc : 60~130KHz". I guess it is used to control LED. – Alex May 05 '21 at 21:41
  • I know there is a LPF but if 10% occurred the ripple rms voltage^2/ESR on output cap power dissipation may exceed it’s spec in the power supply output – Tony Stewart EE75 May 06 '21 at 01:12

1 Answers1

2

enter image description here

Figure 1. Input circuitry of some Mean Well LED PSUs. Image source: LEDnique.

This circuit can be driven by a potentiometer, a 0 - 10 V (or 1 - 10 V) source or a PWM signal but the end result is always the same: an analog voltage reaches the IN control pin. See my linked article for more detail.

Connect your opto-isolator's collector to CONTROL+ and the emitter to CONTROL- and you should be able to control the brightness by PWM control of the opto-LED.

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 2. Microcontroller interface.

If the GPIO is 100% off the PSU will be 100% on. You can't easily invert that action. (More on this later.)

schematic

simulate this circuit

Figure 3. The simplest, no-nonsense method is to use a relay which when de-energised with loss of microcontroller power short-circuits the Mean Well control inputs.

Transistor
  • 168,990
  • 12
  • 186
  • 385
  • Thank you for answer, but if I'm not mistaken if GPIO is not connected, it will mean that DIM+ and DIM- are totally disconnected, which will lead to 100% power of PSU. I need opposite behavior - 0% PSU when GPIO is 0 – Alex May 05 '21 at 17:35
  • I stated that. The rice is cooked so I've got to eat that first. – Transistor May 05 '21 at 17:36
  • I just wanted to mention that I already have tiny board with optocoupler and I can successfully make what you suggested (0 at GPIO -> 100% on PSU). The problem is that I need opposite :) – Alex May 05 '21 at 17:46
  • See the update. – Transistor May 05 '21 at 18:25
  • Hm. Do you suggest to use relay instead of optocoupler, right? I've never used relays before. Will they produce noise if PWM frequency is 500-1000 Hz? My fans make noice in this case. – Alex May 05 '21 at 19:39
  • Reed relays can be NC or SPDT and be driven from logic levels. Do that. Then std. opto-PWM when energized open. – Tony Stewart EE75 May 05 '21 at 20:40
  • @Alex, no, look at Figure 3. I'm using the opto-isolator for the PWM control. The relay is just to satisfy your power-down condition. It will click once on power-up and once on power-down. – Transistor May 05 '21 at 21:18
  • OK, got it. @Transistor, TonyStewartEE75 thank you! I will investigate about this solution. Probably it will be simpler than solution with VOM1271. Any ideas about some particular tiny relay, that can be used in this case? – Alex May 05 '21 at 21:37