Need 12V Signaling Supply from 110VDC to Control DC Motor

[reynolds087]

Hi Everyone,
I'm new to this forum, and I signed up hoping to get help with this issue that's been confusing me. I'm building a CNC mill, and I designed a custom power supply for the spindle. It is based around a 70V power transformer that I am rectifying to ~110VDC. The output is filtered and used to drive the motor through a mosfet stage that is controlled by a pwm signal from an ESP32 microcontroller.

The big problem here is that I don't know how to get lower voltage to drive the gate of the mosfets. I have an octocoupler attached to the PWM from the microcontroller, so all I need is approx 12V that has the same ground reference as the motor's power supply.

I've tried a capacitive voltage divider from the AC side of the transformer but then there is a huge voltage potential between the ground from the 12v and 110v so it won't work to drive the gate. Does anyone have an idea how I could derive 12v approx 40mA from the 110 side to switch the mosfet gates and power an indicator led, and all with a common ground reference?

I am attaching a simplified schematic to show the topology of the circuit.

 

[Harald Kapp]

First you're faced with a few issues here.

1. The MOSFET is oriented the wrong way. Source needs to be connected to the 100 V minus side, not to the motor. Drain goes to the motor.
2. Establish the 110 V minus side as common reference GND, i.e. move the "ground sign" to the minus of the 110 V source. This is a minor and mainly visual issue, but makes understanding (and simulating) the circuit easier.
3. The optocoupler is oriented the wrong way. The emitter needs to be connected to the MOSFET's gate, collector to the 12 V source. Observe the polarity of the output transistor in the optocoupler.
4. The MOSFET is actively turned on by the optocoupler but passively turned off by the pull down resistor. This will create a lot of power dissipation in the MOSFET due to slow turn-off and possibly destroy the MOSFET in turn.

Now to the 12 V. There are a few possibilities to create this 12 V from 110 V:

• Build a simple transistor / zener diode based linear regulator
• Build a linear regulator from a high voltage regulator ic (Google).
• Use a switch mode regulator

The first two methods are simple, but inefficient. Depending on the average current drawn to drive the MOSFET a linear regulator will dissipate a lot of power. The last method is very efficient in terms of energy, but needs more effort for the circuit and will be more expensive.

But: there still is the issue with the MOSFET drive. Instead of the homebrew solution have a look at integrated isolated MOSFET drivers. These chips come with everything you need:

• isolated power supply
• isolated signal transmission for the gate drive
• push-pull gate driver output - sometimes even with adjustable current limiter.

 

[reynolds087]

Thank you for looking over the circuit. I connected the mosfet's source to the motor in the schematic, but not in the real circuit. To be honest I can't distinguish the two in the circuit diagram so that was my mistake. As for the collector and emitter on the optocoupler, same thing. Was another oversight on my part when drawing up the schematic.

Your fourth point is interesting. I watched a few tutorials on Youtube that showed mosfet-based motor drivers with a 10k pull down resistor. What do you recommend rather than the method I am using? I am only switching at 5KHz, do you think I will have a problem?

The first regulator you mentioned, the transistor / zener diode based linear regulator seems the best route. Do you have any examples you could attach or link to? The reason being that high voltage linear regulators after shipping are like 10 or 15 dollars at least, plus I was hoping to have something ready within a day or two, and I don't want to use a switch-mode power supply because correct me if I'm wrong, but they are galvanically isolated from the voltage source, through the pulse transformer, and as a result would not have a common ground with the mosfet.

 

[Harald Kapp]

What do you recommend

a dedicated MOSFET driver.

Do you have any examples you could attach or link to?

The basic circuit looks like this. You may add an additinal series power resistor in front of the circuit into the "+" connection to distribute power dissipation between the transistor and the resistor. The value of the resistor depends on the average current drawn by the MOSFET driver circuit.
An additional electrolytic capacitor (100 µF ... 1000 µF) at the output of the regulator may also be a good idea.

 

[reynolds087]

Based on your recommendation, I found this circuit. Would this work for voltage regulation and then I could follow it with a current-limiting resistor?

 

[reynolds087]

a dedicated MOSFET driver.


The basic circuit looks like this. You may add an additinal series power resistor in front of the circuit into the "+" connection to distribute power dissipation between the transistor and the resistor. The value of the resistor depends on the average current drawn by the MOSFET driver circuit.
An additional electrolytic capacitor (100 µF ... 1000 µF) at the output of the regulator may also be a good idea.

Thanks, disregard my last reply. I wrote it up before I saw yours. I will try that out. Greatly appreciate the help!!

 

[WHONOES]

If I was trying to do as you are, I would use a small transformer to generate a supply for the 12V supply needed for the PWM drive circuit. No danger then of overdriving the gate and no need to drop 70V down to 12V with the possibility of that circuit failing and destroying the FET by applying 70V to the gate.

 

[reynolds087]

If I was trying to do as you are, I would use a small transformer to generate a supply for the 12V supply needed for the PWM drive circuit. No danger then of overdriving the gate and no need to drop 70V down to 12V with the possibility of that circuit failing and destroying the FET by applying 70V to the gate.

Maybe I have misunderstood this aspect of mosfets, but I always thought that the gate and the drain need to share a common ground reference.

With an additional transformer to drop 70VAC to 12VAC, wouldn't the DC output be isolated from the motor's DC supply? And when I attempt to drive the gate of the mosfet with the isolated 12V, how will it interact with the 110VDC on the drain and source?

 

[Harald Kapp]

Maybe I have misunderstood this aspect of mosfets, but I always thought that the gate and the drain need to share a common ground reference.

Maybe you have misunderstood the aspect of a common ground reference? Not really clear what you mean by "gate and the drain need to share a common ground".
A Mosfet can be applied anywhere in a circuit, not necessarily directly connected to ground. Besides, there are N-Mosfets and P-Mosfets. N-Mosfets are the more common type and often are referenced to ground. However, P-Mosfets are not uncommon, too, and are referenced to the positive supply.
I think your misunderstanding is here:

With an additional transformer to drop 70VAC to 12VAC, wouldn't the DC output be isolated from the motor's DC supply?

Easy to fix by connecting the two grounds together:

 

[reynolds087]

Maybe you have misunderstood the aspect of a common ground reference? Not really clear what you mean by "gate and the drain need to share a common ground".
A Mosfet can be applied anywhere in a circuit, not necessarily directly connected to ground. Besides, there are N-Mosfets and P-Mosfets. N-Mosfets are the more common type and often are referenced to ground. However, P-Mosfets are not uncommon, too, and are referenced to the positive supply.
I think your misunderstanding is here:

Easy to fix by connecting the two grounds together:
View attachment 48596

What I meant about the gate and the drain is that the reference should be the same. My understanding is that the gate is basically a capacitor that is charged in order to allow current to flow between the drain and source. So that charge would need to be in respect to ground, assuming it's an N-channel mosfet, right? That is why I have been trying to find a circuit that drops the voltage level without isolating it from the 110V side.

I originally tried a capacitive voltage divider circuit like the kind that are used in some household lighting and other things. I know they are dangerous, but since I am using an isolation transformer from mains, and low current, I thought it would be a nice solution. So in that case there were two separate bridge rectifiers. One for the 12V capacitive voltage divider circuit, and one for the 110V side. When I connected the ground together, it blew the zener diode. Measuring the two grounds, there was around 50VDC potential between them.

I don't really understand why that was happening, and I assumed using a second transformer could cause the same problem.

 

[Harald Kapp]

What I meant about the gate and the drain is that the reference should be the same. My understanding is that the gate is basically a capacitor that is charged in order to allow current to flow between the drain and source. So that charge would need to be in respect to ground, assuming it's an N-channel mosfet, right?

Right, it wasn't all that clear to me from your first explanation, sorry.

I don't really understand why that was happening, and I assumed using a second transformer could cause the same problem.

By connecting the outputs of the two bridge rectifiers you created an almost short circuit. Trace the connections to see what I mean (for comparison see this thread).
When you use separate transformers the secondary AC sides are isolated from each other and the connection as shown in my post #9 above will work like a charm.