Water Electrolysis Power Supply - 250KHz, 1:3 Duty Cycle, +26V 50A

[pnielsen]

I would like to build a power supply for experimenting with water electrolysis. The information I have calls for a square wave pulse, timed 0.001ms "on" and 0.003ms "off" (250KHz). It is supposed to supply +26V at 50 amps into the cell.

I am considering the power stage of the attached circuit. But rather than build a signal generator from scratch, I intend to connect a function generator at the junction of IC1/2 pin 1 and R11.

Am I on the right track or can someone suggest a better solution?

If the 250KHz is problematic, it may be possible to reduce it to as far as 25KHz.

 

[Pommie]

1ms on and 3ms off is 250Hz - not 250kHz. Which one is it?

Mike.

 

[pnielsen]

Sorry. Should have picked that up. It is 250kHz. I have edited my original post accordingly. Thanks

 

[ChrisP58]

He's asking for 0.001 mS on and 0.003 mS off, which is a period of 4 uS. So that will be 250 KHz.

 

[Pommie]

He's asking for 0.001 mS on and 0.003 mS off, which is a period of 4 uS. So that will be 250 KHz.

Originally he didn't.

Seems you're making a 1kW 250kHz transmitter.

Mike.

 

[ChrisP58]

Originally he didn't.

Seems you're making a 1kW 250kHz transmitter.

Mike.

Sorry Mike.

I now see the timing of the edits.

It always frustrates me when posters do that. It makes it hard to follow the continuity of a thread.

It's far better to make another post with the corrected information.

 

[Inquisitive]

Originally he didn't.

Seems you're making a 1kW 250kHz transmitter.

Mike.

Yes, Mike is correct he was not seeing things. The original post was edited after that detail was brought out in the thread.

Inq

 

[ronsimpson]

Sorry. Should have picked that up. It is 250kHz. I have edited my original post accordingly. Thanks

I am a little confused as usual. The text says hz or khz who knows and the picture from #1 clearly says hz.
The poor way the MOSFET's gate driver is built it must be hz. (slow not fast)
---edited----
Please show a link to where you got this circuit from.

 

[pnielsen]

My apologies for error in my OP. The diagram I attached specified it is capable of up to 3KHz, and I mentioned 25KHz would be acceptable if 250KHz is not. It is a PWM speed controller. I was considering to use or adapt the power stage after IC1/2.

If there is a problem with the gate driver, where can I find a better cuircuit? Note that this is intended to power an eletrolytic cell, not transmit RF.

Can we continue on this basis or should I repost this thread?

 

[ChrisP58]

If you have a function generator that you can set for the frequency and duty cycle that you want, then just use it to drive a mosfet gate driver.
Here is one that is still available in a DIP package. **broken link removed**

The output of the mosfet driver drives the mosfets, but I recommend a resistor (10 to 100 ohms) in series with the gate of each mosfet.

Stay with this thread. No reason to start a new one.

 

[pnielsen]

Thank you for your reply. The circuit below looks like a solid design. It includes the series resistors you recommeded.

I assume the IR2110 will accept a variable duty cycle, logic level signal from my function generator.

Any comments?

What could I expect as the estimated maximum switching frequency?

https://www.infineon.com/dgdl/ir2110.pdf?fileId=5546d462533600a4015355c80333167e

 

[ChrisP58]

The IR2110 is a highside and a lowside drver, instead of only a lowside driver like the one I proposed, and like your original circuit.

See that extra Q1 mosfet? It pulls the output up to the positive supply rail. Do you need that?

 

[pnielsen]

Thank you for your suggestion. Yes, I probably don't need the high side feature, unless it has better load driving properties I have yet to learn about. All I want is current flow from the cell's positive electrode to the negative.

Do you see any need to improve upon the circuit below? The original circuit I posted used two N-channel MOSFETs in parallel and seemed to have extra protection. I need to supply about 50A at 26V to the cell, which is not a typical load, and ensure the driver will not be adversely affected.

https://microcontrollerslab.com/use-mosfet-driver-1r2110/