Boosting the 3.3V PWM output to 10V

[SaleB]

I have successfully used a MOSFET to drive 12V LED strips with 3.3V PWM from ESP8266 in the past. For that, I have used this setup

The MOSFET in the picture was NIKO P3055LD, http://pdf.datasheetcatalog.com/datasheet_pdf/niko-sem/P3055LD.pdf.

Now I have a similar problem but need assurance that it is safe to use it in this way or advice how to improve the solution.

I have a 5V source (ESP8266 module does it's own regulating down to 3.3v). And I have to drive a 1-10V PWM control input of a Meanwell HLG-100H-42 https://avi-solutions.com/meanwell-datasheet/HLG-100H.pdf. For that purpose, I intended to use a XL6009 based step-up converter to get 10V. If relevant, the ESP8266 PWM frequency is 1KHz. Something like this:

Question is can it work in this way? Can I harm the Meanwell? And how can I additionally secure the Meanwell?

In the Meanwell datasheet, there is a mention that the dimming input consumes about 100 microamps. Should I add a 100K resistor in line behind the XL6009?

I know that this MOSFET is an overkill for such a small current, but its the only one I have at hand whose gate is drivable from 3.3v. I cannot find a seller in my country that has any 3.3v logic level MOSFET at hand, so these are from AliExpress and I have plenty of them. Can you suggest a smaller 3.3v logic level FET that is popular so I can order it from Aliexpress just to have at hand for the next similar situation?

Thank you

[james_s]

It's not entirely clear from the datasheet but it looks like the dimming input is a differential input so what you propose will probably work fine. I think I would just use a small
BJT though, something like a 2N3904 will be fine. No resistor is needed but it shouldn't hurt anything if you want to add one.

[edavid]

Since you can dim it with just a resistor, there is already a voltage across the dimming pins.  You should be able to hook up your PWM MOSFET with no additional voltage source.  Perhaps you would need 100K in parallel with it.  I would try asking Meanwell for clarification.

[Ian.M]

That would also be consistent with the nominal 100uA input current in voltage controlled mode, if its actually a Hi-Z input with a 100uA current source pullup.   The one thing that concerns me is that the datasheet states there must be no connection between control circuit negative and LED output negative, but doesn't state what happens if that constraint is violated.   It may therefore be worth using an optocoupler between the 3.3V logic level PWM signal and the Meanwell control input to avoid any risk of damage if any point in its LED string shorts to ground.

[james_s]

It's entirely possible that the LED - lead is not at ground potential so connecting it to the dimming input could exceed the max common mode potential that the input can tolerate. At the very least there is probably a low side current sense resistor.

[SaleB]

@Ian.M:

The output of the LED driver is self-regulating 38-46v output to which are connected a few LED spotlights each containing a series of 1W white LEDs, and it's an entirely different circuit. I have at hand 4n35 optocouplers (https://www.vishay.com/docs/81181/4n35.pdf) that I can use, but I would need help with how to connect them. The diode, I assume, should go between PWM out and ground, but the transistor side I am not certain how to connect and how to choose the right value for the collector resistor Rl.

@edavid:
I am not familiar with the term in parallel with FET. Does it mean between the drain and source, or between gate and ground (source)?

[Zero999]

There is no need for a boost converter to convert 5V to 10V. The Meanwell has a built-in 100μA current source pulling the input up to 10V. All that's needed is a transistor to short the PWM input together. If nothing else in the ESP8266 circuit is connected to the Meanwell driver, then there will be no need for an opto-coupler. And no, if you do use an opto-coupler, the transistor side should go to the PWM input, with the ESP8266 circuit's PWM output going to the photodiode.

To expand further, the Meanwell's input looks like this:


The idea is it will work with a 0 to 10V, a resistor, or PWM input. If a resistor is connected to the input, the 100μA source will pass a current through it up to a maximum voltage of 10V. For example, a 10k resistor will drop 1V across it, with 100μA through it. The low pass filter removes the AC component, so if you apply a 50% duty cycle 1kHz 10V square-wave, it will be smoothed to 5V.

Connecting a transistor between the dimmer inputs will short them together, when it's on and the voltage will increase to 10V, when it's off, thus the logic will be reversed.