50V, 50A H-Bridge IC needed

[Riccardo1]

Hi,

I'm looking for an integrated H-Bridge IC like those used in motor control applications. I've seen the VNH3SP30-E (http://www.st.com/stonline/products/literature/ds/12688.pdf) but it is only rated for 36V. The current rating it has of 30A is just enough but ideally I would like something higher.

This IC looks just what I need except for the voltage rating (I need at least 50V). I've been searching for hours but can't seem to find what I need. Does anyone know of an equivalent for this? I must be using the wrong search terms because I'm sure something like this will exist.

Cheers

 

[Hero999]

50V@30A.

I think you'll need to build the h-bridge from scratch from MOSFETs.

How fast does it need to switch?

 

[Riccardo1]

Hmm. I was hoping to avoid that as I'd need to make a high and low side driver too. Those IC's also have fault indicators and other useful stuff

It only needs to go at 50Hz or less.

 

[Hero999]

50Hz is fairly low so it should be easy to level shift the h-side P-channel MOSFETs using BJTs and a potential divider or zener to prevent the gate voltage from being exceeded. At higher frequencies this sort of thing gets harder as the resistors slow the switching speed too much.

I'll post a schematic if you don't get it.

 

[Riccardo1]

Yes, a schematic would be great thanks. Do you also know of a control IC that can be used to simplify the driving if it. e.g, prevents shoot through or other fault conditons?

 

[Hero999]

Here's a schematic, no control circuitry is required assuming the lower MOSFETs are logic level.

The diodes prevent both the high and low side MOSFETs from being on simultaniously.

View attachment 40712

 

[Riccardo1]

Wow, you made that diagram just for me! Thanks! I have a few more questions  :

Will I need to change those resistor values if the supply voltage drops to 25V ?

Just to confirm: The high site MOSFETs are ordinatry P-Type, and the low side ones are logic level N-Type ? They must not have reverse diodes built in?

Can I use IGBT's instead of MOSFET's?

 

[Hero999]

R1 and R2, and R3 and R4 form potential dividers which limit the gate voltage to the high side MOSFETs to a safe level, whether their values need to be changed for operation at 25V depends on the drain current at the lower gate voltage.

The MOSFETs are normal enhancement MOSFETs with built-in diodes, if you're using this to power a large motor it's a good idea to add extra diodes because the internal diodes might not be fast enough.

I don't know why you'd want to use IGBTs, at low voltages the on voltage loss is higher than MOSFETs, IGBTs are only any good for higher voltages >400V or so.

 

[Riccardo1]

It will be to drive coils / electromagnets so I was thinking that an IGBT would be better for 2 reasons.
1. They can have a high voltage rating and be more tolerant of voltage spikes.
2. They don't get as hot for a given current.

The supply voltage to the load will be variable (although there will be a constant 12V source available for the control electronics). Different coils would also be used so the output current could vary quite a lot.

For those potential dividers, would it be ok to just add some back to back zenner diodes to prevent over voltage on the gates?

 

[Hero999]

2. They don't get as hot for a given current.

That's not always true.

It depends on the MOSFET and the IGBT, as I said before, for devices with higher voltage ratings IGBTs are best but for low voltage devices MOSFETs are much better.

For those potential dividers, would it be ok to just add some back to back zenner diodes to prevent over voltage on the gates?

If you're going to use zeners then you might as well replace R1 and R3 with zener diodes to limit the gate voltages to a safe level. It's still a good idea to put a resistor in parallel with the zeners, to alow the gate capacitance to discharge, but it's value is much less critical than for a potential divider.

 

[Riccardo1]

Thanks again. I really appreciate your help.
Does this diagram look ok?

I've not found an appropriate pair of mosfets yet. I find one type, but then can't find its equivalent of opposite polarity. I'll update when I find them.

View attachment 40720

 

[Hero999]

No, you clearly don't understand how the circuit works.

What voltage is driving the low site MOSFETs?

I assume it's 5V so why do you need 10V zeners to protect it?

Why do the zeners need to be bidirectional when the gate voltage can only be one polarity?

The high side MOSFETs are connected backwards.

 

[Riccardo1]

:-[
I thought 5V would be too low to switch the mosfet properly. I was thinking of using some comparators to convert the voltage level to around 12V.
I chose 10V zeners because that is high enough to allow the mosfets to switch on but low enough to protect them against any transients.
I made them bidirectional because I'm concerned that transients of either polarity could occur since the load would be inductive and the current is being alternated. Is that wrong?
I just connected the mosfets like in your diagram, but yes, I see now that they (and the diodes I added) should be the other way around. Don't know how I missed that!

 

[Hero999]

If you use logic level MOSFETs for the low side then you won't need to perform any level shifting because they will be able to conduct the full current with 5V at the gate.

 

[Riccardo1]

I've got some STP40NF10L, and IRF5210PBF MOSFETs handy so I think I will use those as they are N and P type respectively and both rated for 100V and 40A.

I also have some comparators, so I will just use those for the level shifting.

Thanks

 

[Hero999]

Those should be fine but you'll need to connect two in parallel for 50A.

Looking at the datasheet, the STP40NF10L should be able to pass 25A with a gate voltage of 5V so you shouldn't need any level switching.

 

[Riccardo1]

Ok, Thanks for your help 

 

[Riccardo1]

Hi there,

I have another question about the diagram posted by Hero999.

On the gates for the high side transistors, there is a pair of resistors (R1, R2). When input A goes high, this should activate Q1 and pull down the gate voltage of Q5.

My question is, How did you choose those resistor values? As far as I can tell, with Q1 activated, the gate of Q5 will be at 32V (relative to ground).

I've tried this in a test circuit with 9V input and a small solenoid as a load, but it wont switch it. Just wondering if the high side gate is not being pulled low enough.

Cheers!

 

[Hero999]

The resistors form a potential divider which keeps the gate voltage to the top MOSFETs below the maximum rating (normally 20) but high enough to turn them fully on.

V = 4.7*50/(10+4.7) = 16V

If the supply voltage is too low the gate voltage will be too low to enable the MOSFETs to carry the required current.