H-bridge & High Voltage/High Side -- Oh my, how to drive??

Sorry for the wordplay, I couldn't resist!

Hello all, I need some help (more like guidance) once again!

I want to design a couple of things that require an H-bridge.. So far so good, I know my way around those as long as it's low voltage (12/24V stuff i've built before, mostly).

However, this time I want to design something that uses a higher DC voltage (more like 90 or 120 or even 180 Volts) and I want some clues on what could be used as driver for the high side mosfets. I want to use P-mosfets for the high side so do keep that in mind please (I dont believe I can get charge pumps for 120Volts, or even more volts, anyway..)

For a project I have in mind for example, what I will have available is a 90VDC source, a 5V/12V stepped down version for logic and other misc components, and of course a (common) ground.

What would you use in such a case to drive the high side P mosfets? Some kind of totem pole / level shifter maybe?

Hi,

Why do you want to use pmosfet? It is more common to use nfets.

90 up to 120V is not that high voltage. Ther are a lot driver circuits for mains applications. 600V is not unusual. Often in SO-8 package.
Instead of charge pumps they use bootstrap circuit to generate the needed voltage to drive high side nfets.

Just do a search in this forum, or google.
One often used driver IC is IR2110. Google for it.

Also look at the semiconductor manufacturers homepage.
Www.irf.com
And many others

Klaus

KlausST said:

...

Click to expand...

Oh. I thought (for some reason - I am not sure why!) that for such high voltages there wouldn't be any good methods to drive N-mosfets at the high side.. It seems I was mistaken!

I think I saw in the datasheet (ir2110) that it can handle 500/600v, thats just impressive!

I have no problem switching the design to accomodate high-side N-fets, after all its all in paper right now, but out of curiosity is there no hope for a P-fet h-bridge?

(I admit I have only used p-fets with some 12V bridges I have built)



An afterthought!

I think I have read in an article somewhere that for charge pump/bootstrap circuits to work properly and provide the high(er) voltage needed the high side transistors must be switched off at some point, ie no 100% duty-cycle..

Is this true? Do such circuits require that the mosfets are not constantly on? Or have I misunderstood something?

Hi,

is there no hope for a P-fet h-bridge?

Click to expand...

For sure you can do this. But it means more effort, more time to develop, more chance for the design to fail (there are a lot of application notes for N-FET drivers and all the problems with it are well known and described)
You have a lot more N-FETs to choose than P-FETs. You usually use identical N-FETS for both high side and low side with identical switch on and switch off time, Mind the dead time to avoid cross conduct of both FETs.

Is this true?

Click to expand...

Definitely yes. This is one problem by using the standard bootstrap circuit. But for sure you may use an isolated DCDC converter also.
The parameters for the bootstrap device values depend on your FETs, the gate drive circuit and the max. ON time.
For a continous switching PWM of 10kHz or more a 100nF should be sufficient. but you can use a higher value. Ceramic.

The OFF time to charge the bootstrap should be at least in the range of more than 100ns. I´v seen circuits with 20ns. If you can spend 1 us then do so.
(mind that for bootstrap charge you need to actively switch ON the low side FET, both FETs OFF may fail)

All my given values are just to get a clue on where to go. All is application specific.

Best is you show us your circuit and we can help you on this.


Klaus

KlausST said:

Best is you show us your circuit and we can help you on this.Klaus

Click to expand...

There is no circuit yet, I am still working with the idea.. That's why I can easily adapt to using high side N-fets hehe

The concept is that I have come across a 90VDC motor that I am thinking of using for a project and I'd really like to build a control circuit for it before doing anything else. I can find ready made ones, but where is the fun in that?

It's just that it's my first time designing a high voltage motor control circuit and I thought I'd ask for some guidance before even starting, for good measure!

As for the circuit, I am thinking about 20-25 kHz frequency or slightly more if the whine is annoying, there is going to be switching for sure (speed control) so maybe I'll just get my hands on an IR2110 and rig it to see what happens..

I've read a couple of informative posts regarding that chip and it seems like a fast & clean solution, wonder why I didn't know of it before!

Hi,

I am still working with the idea

Click to expand...

... then it´s time to bring your idea on paper....

;-)


Klaus

hehe I agree!

Thank you for your help, I will resurrect this post when I have some schemating going on !