I am designing PWM speed control for 110v DC motor. Since it is 110v, I am going to power it from bridge rectifier connected to mains, with MOSFET on low side for switching.

For driving MOSFET I plan to use dedicated driver (though it doesn't matter for the question I have, so let's assume I drive gate directly from the low-voltage schematics).

For low-voltage side (microcontroller/logic) I want to use typical wall charger, which is switched power supply, as I understand.

The tricky part here is that driver supply voltage (or gate voltage) will have to be applied relative to the "power ground" ("-" of bridge rectifier/MOSFET source). This means "-" of the 12v switching power supply will be connected to the "-" of bridge rectifier. If the 12v power supply was transformer-based power supply, it wouldn't be an issue since 12v part would have been isolated from the mains and connecting two grounds would be fine.

However, somehow I think that connecting these two grounds in case 12v is provided by switched power supply would result in a short, since, as I understand, there will be no galvanic isolation between 12v and 110v DC (bridge rectifier output).

So, the first question, if I am understanding the problem correctly?

Second question is, what options do I have for supplying power for gate driver/MOSFET gate? I can isolate logic part using optical isolator, but I would still need low voltage power supply for driver/gate.

It seems to me that simple option (without using transformers) would be to use resistive/capacitive power supply connected to the bridge rectifier? Although driver should be able to provide relatively high current to the gate (let's say, 5A), as I understand, what matters is how much charge is actually transferred. Assuming Qgs = 150nC and switching frequency of 20kHz it would be Qgs*F per second, which gives Isupply = Qgs * F ~= 150nC * 20kHz = 3mA. Which means, resistive power supply providing 50mA would be plently. 50mA with 100v drop would give 5W of burned energy, which is OK.

So, the third question, am I right in these calculations?

Here is the schematics, but there is not much to draw. I omitted all other components (like snubbers, etc). Box on the bottom is switched power supply. L1 is motor. Imagine I want to open MOSFET by applying 12v to gate-source. My understanding is that I cannot just connect two grounds and connect 12V to Vg -- this would result in short.


simulate this circuit – Schematic created using CircuitLab

P.S. I think, I am wrong about switched power supply. It seems like some of them (most of them?) are actually isolated. Which means I should be able to use it for MOSFET driving circuit.

  • \$\begingroup\$ Please add a schematic. \$\endgroup\$ Jun 4, 2014 at 0:00
  • \$\begingroup\$ You do realize that the rectified voltage will be higher than the RMS voltage, and that a non isolated system will be more dangerous to work on, right? \$\endgroup\$
    – W5VO
    Jun 4, 2014 at 0:55
  • \$\begingroup\$ Yes, I expect DC to be 110*sqrt(2). Yes, I do realize it would be dangerous. Though it is still 2 times less voltage than in the country I came from :) \$\endgroup\$ Jun 4, 2014 at 3:55

1 Answer 1


TI offers some isolation gate drivers that let your do exactly this, without destroying yourself or the other circuits.

Also, beware of the minimum clearances for this type of circuit.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.