Help with MOSFET and driver selection

Hi All,
I'm in the process of building DC power supply that will run off 240vac mains.
The main reason for this build is so that I can take advantage of PWM on the Arduino and the fact you can change the frequency of the PWM easily within a sketch.

I can manage getting 240vac to a smooth DC form just fine, there is plenty of info out there to take care of that.
What I'm struggling with is the choice of mosfet and driver. There is like a billion to pick from.

My basic parameters are:
+/-240vdc. Not sure what the out come will be so I figure a mosfet rated for 650v should cover all bases.
1-5 amps. This is a testing rig so may draw up to 10 or 15 amps on occasion.
Switching ability. I'd like to set the PWM frequency anywhere between 50Hz and say 100,000Hz or more if possible.
Type of electrical load, I'm unsure of. What I'm running will be an electrolyzer. I don't know if that's classed as inductive or not.

I'm hoping for a little guidance on gate driver and mosfet selection. Here is what I have in the shopping trolley at the moment. Do you think they will work??

Please let me know if you need more info but I'm not sure what else to supply at this point.

The Mosfet selection requires more information about the circuit, planned frequency and especially the transformer.

You should include the method of feedback from the output to the Mosfet since there can be no electrical connections between the output and the mains.

Do you have an idea for EMC control?

It appears the parts may work, but without a schematic and data about the load this is only a guess. Note one of your parts the TC4420VPA shows a 38 week lead time, the other shows 39 weeks. If you use these parts order lots of spares so you have them if something blows. at 100,000 Hz you need to be careful with the circuit design. If you decide to add flyback diodes pick them carefully. This gives you time to draw a schematic, not a frizzy picture and post it, we will look at it and offer advice and answer your questions.

Do you want the Arduino to select and monitor the output voltage with external regulator.
Regulate the voltage?

A circuit diagram of your concept would be great.

Tom.. :smiley: :+1: :coffee: :australia:
PS. I have yet to see an Arduino controlled regulated power supply.
Lots of projects where the Arduino selects the voltage and possibly current limit, but external component actually doing the dynamic regulation.

Where did you see that? It says they both ship immediately from my end.

I will no doubt need help with the circuit design but I'll put something together and post it.

I don't plan on using a transformer. I'll tune my electrolyzer to suit the end DC output voltage. Voltage fluctuations aren't that important at the load end.

EMC, hadn't even crossed my mind. I do have an EMI filter off an old treadmill somewhere that goes on the 240vac side. I could put some ferrite rings on there maybe? (I can see you shaking your head right now)

Thanks for the info, I will go do some research and get a schematic back to you.

I don't see this ending well.

Lol. Yeah I'm messing with high voltage. I know the dangers. I'm also messing with extremely explosive hydrogen and oxygen gas from within the electrolyzer. I better update my will.

And NO isolation from your mains power wiring.


What are you electrolzing?

Tom... :smiley: :+1: :coffee: :australia:

Illegal and dangerous :frowning:

As @Paul_KD7HB said "I don't see this ending well."

Do you understand it is likely your power supply will be putting mains 240 VAC on the ground output of you power supply? And if someone touches it and another device it is likely they will die?
You should know I'm not a nit picker with regards to other's posts. What I am saying is not some arcane rule normally only of interest by some super safety folks, its a real danger similar to crossing a busy street without looking for oncoming vehicles.

I'll be electrolyzing water. And hopefully no one else.

All jokes aside JohnRob, I do take this seriously. I need to make this safe. That is why I'm on here asking for guidance.

I would like to get yours and others approval before I go plugging anything into the wall. :slightly_smiling_face:

Here's the issue....

Lets assume for a moment your circuit has a common (aka ground). This common starts at the mains and becomes your negative output.

You plug your circuit into the mains. You have a 50/50 chance of which connection (pwr in or common) the "hot" wire of the mains connecting to your circuit common.

Now if someone touches your circuit common and has some path to ground (ie a water pipe or some properly grounded electrical device, they will receive 230VAC right through them. Lets hope you don't like that person as you will likely not have to talk to them again.

Put another way, what you are considering doing is literally the same as pluging in a single wire to a mains receptacle and holding on to it while you perform your experiments.

Don't let my attempt a levity suggest this is not an important issue it is literally life or death :skull:

Ok correct me if I'm wrong but mains ground from the 230vac should not be connected in any way to your dc side of the circuit.
It will only be the phase wire and the neutral wire that connects to the full bridge rectifier. The rectifier will then produce the +vdc and earth or -vdc for the dc side.
The AC main ground wire (green) should be grounded to the casing if the casing was made of a conductive material.

Ok JohnRob I see what you mean now. I was just reading on another website about a CAPACITOR FED POWER SUPPLY. I didn't even know you could do that, I thought you had to have a rectifier.

What makes you think you will need 230Vac to do the job.

A 12Vdc supply and a PWM controller will do the job.
Electrolysis relies on current, electron flow to make the process work, not 100s of volts.

You project will need DC voltage.

Can you post a concept drawing of your project.
Safety with Oxygen and Hydrogen is dealt with proper plumbing and storage.

Thanks.. Tom... :smiley: :+1: :coffee: :australia:

Yes your right TomGeorge, It actually only needs 2vdc per cell. The more cells you have the more voltage is required. This is all one big experiment so everything is design in progress.
I do have an ulterior motive driving this build for the high voltage dc. I have some 180v dc motor projects in the pipeline.

With the limitations of physics today there are only two ways to electrically isolate the mains from your output.

  1. magnetic (aka transformer)
  2. optical (currently no way to transfer more than a very low current)

Any type of "CAPACITOR FED POWER SUPPLY" will NOT provide isolation.

I'm trying to not be insulting but I believe you are in over you head here. Building switching power supplies is an EE specialty. I suspect you could spend less and have a better result if you purchased a power supply. If you are thinking of cost, you can probably get a part time job, spend less time at that job and make enough money to purchase such a supply.


I clicked your link, at the top it shows Manufacturer Standard Lead Time 39 weeks. I did find on the top right 463 in stock when I scrolled all the way up, so I stand corrected. I do not purchase from them so I am not familiar with there layout.

Say I use a transformer. 230vac on the primary side and for the sake of this discussion I want 230vac off the secondary side and I reqired the full 10amps. Both windings will be identical. I will have isolated my circuit from the mains power but the risk of electrocution is still the same is it not??
I comprehend what your saying but am I missing something here?

Yes you are still dealing with hazardous potentials, however at least you know where they are. With isolation you won't have any hazardous potentials between you output and earth ground.