Seeing another thread here, https://forum.allaboutcircuits.com/...-output-voltage-appears-on-input-side.192993/ made me think about one of my stalled project. Many years ago a member that has since passed suggested to me that an isolated DC-DC convertor can be used instead of the common bootstrap circuit for a high side Mosfet driver.

Seeing the new thread mentioning it brought it back in my mind and I've been thinking on the project again. I need to use Nmos in a high side switch because of the amperage involved and have no need for a low side that could be used to charge the bootstrap cap. So does anyone have any thoughts about whether the DC-DC convertor will work for a boot strap? The cost doesn't bother me because it's not a market driven thing just a personal one.

 

It will work. Connect its ground to the MOSFET source. If you need to control it with a circuit with a different ground, you need an opto-isolator.

 

It was RonV that suggested it to me. At the time I had never even heard of isolated DC - DC convertors. Before asking hear I did some Google searching but didn't really find an answer that made any sense.

So your saying just add the isolated ground to the mosfet source/output will work? That was the part that wasn't clear to me, where the convertor ground would go to complete that circuit. Thanks for the answer.

 

If you ae anything like me, you may have a "treasure trove" of old goodies. Recently, I pulled a DC to DC converter of an old network hub that had a BNC connection. This DC to DC converter is a 5V to 9V 200mA unregulated one, rated at 4KV I think it was. So it is worth having a look on old boards just in case.

 

Yes. I've seen it used in an application note for a 2-transistor flyback or forward converter somewhere.
Also, the LT7001 has a built-in charge pump that operates when the normal bootstrap capacitor stops so that it can do 100% Mark-space
https://www.analog.com/media/en/technical-documentation/data-sheets/ltc7001.pdf

 

Thank you all for your answers. But I have another question about this, do I need a diode between the convertor and the gate driver?
I did when getting in to this find a couple of gate driver chips that had the convertor built in, from Analog Devices, ADuM3223 and ADuM4223, but they are surface mount and I haven't done any surface mount stuff. I see that the LT7001 is also surface mount.

 

I need to use Nmos in a high side switch

Does the MOSFET need to be on for a long time (DC)?

 

do I need a diode between the convertor and the gate driver?

Whereabouts? If you're thinking of the diode to the top of the bootstrap capacitor, then some drivers have it built in and some don't.

 

Does the MOSFET need to be on for a long time (DC)?

Not on for a long time at all. There will actually be two mosfets charging and discharging a capacitor bank. When the bank reaches a certain voltage it stops charging and then the discharge one turns on until it reaches a lower set voltage, and the cycle starts over again. The charge and discharge voltages are measured using a window comparator circuit. The frequency should be ~15kHz or so. This is going to be used in my electrical discharge machine power supply.

The reason for using voltages instead of an oscillator is sometimes in these machines the spark between the electrode and work doesn't ignite. Or the circuit wave form keeps the on time on longer than necessary, making problems in the process.

 

Whereabouts? If you're thinking of the diode to the top of the bootstrap capacitor, then some drivers have it built in and some don't.

Yeah, I was thinking between the converter and the driver. Since the ground of the converter is on the source and not to a circuit ground , I didn't know if it could/would back feed into the converter circuit. Like I said earlier these converters are a totally new thing to me.

 

After finding out it will work Googling farther I found that Murata makes a special series of converers just for this application, the MGJ series in 1, 2, 3, and 6 watt outputs.