High current H-Bridge

[scoe]

Hi!

First i would like to note that i'm not EE but ME so please have patience with me Smile

Ok, i would like to make full h-bridge to drive ~10A motor at 7.2V. I will connect two motors in parallel so it will be around 20A. Problem is that motors can drain around 200A peak.
First, i'm consider to design h-bridge with all n-fet IRFZ46. They already have build in diode which i'm not sure is enough to suppress induced voltage.
I will use uC to create PWM signal and one direction signal. So my h-bridge will have only two inputs. I'll use NOR logic to drive max4420.
For driving Mosfets i have max4420 (i know that they are not quite best choice, but i have it in my drawer so...). I saw somewhere on Net that i need to apply Vin to max4420 slowly, but im not sure why. If that is true, is it enough to use some condenser ?
Also, if i use max4420 do i need to use voltage doubler for upper mosfets to supply enough volts on gate ?

For protecting the uC circuit, do i need to use some kind of optocouplers or is it enough to use those mosfet drivers ? I'm not sure if optocouplers are fast enough.

After some confirmation, i'll draw circuit schematic for further analysis.

Thanks in advance!
Sven

 

[dknguyen]

Hi!

First i would like to note that i'm not EE but ME so please have patience with me Smile

Ok, i would like to make full h-bridge to drive ~10A motor at 7.2V. I will connect two motors in parallel so it will be around 20A. Problem is that motors can drain around 200A peak.

How can they drain 200A when they run at 20A? Or do you mean they have a stall current of 200A?

First, i'm consider to design h-bridge with all n-fet IRFZ46. They already have build in diode which i'm not sure is enough to suppress induced voltage.

Don't count on it. Use an external schottky diode (very fast and low forward voltage) in anti-parallel with all the FETs. It's a parasitic diode that is there because of practical technical issues that arise when making MOSFETs. It's big, and powerful, but slow.

I will use uC to create PWM signal and one direction signal. So my h-bridge will have only two inputs. I'll use NOR logic to drive max4420.
For driving Mosfets i have max4420 (i know that they are not quite best choice, but i have it in my drawer so...). I saw somewhere on Net that i need to apply Vin to max4420 slowly, but im not sure why. If that is true, is it enough to use some condenser ?

Huh?

Also, if i use max4420 do i need to use voltage doubler for upper mosfets to supply enough volts on gate ?

You are right in that there will be a problem. Not a voltage doubler as that is an absolute value and can exceed the max gate voltage of the FET in some cases. What you need is a way to provide a gate voltage that is referenced to the source of the MOSFET. Either an isolated supply or bootstrap capacitor setup will do (there are ICs for this). Pulse transformers also work and provide isolation, but I won't get into that since they have a lot of disadvantages and I am biased.

For protecting the uC circuit, do i need to use some kind of optocouplers or is it enough to use those mosfet drivers ? I'm not sure if optocouplers are fast enough.

You can get fast enough optocouplers since motors are mechanical devices are are comparatively quite slow. You can also use magnetic isolators that use mini transformers inside the ICs like iCouplers from Analog devices or use pulse transformers as mentioned above. You may not need them though...I don't use em on my 24V motor drivers. I woudl though, but there are other circuits involved in my schemeatic that completely defeat the purpose of isolation.

https://www.modularcircuits.com/h-bridge_secrets1.htm

 

[scoe]

Hi,
thanx for reply.

How can they drain 200A when they run at 20A? Or do you mean they have a stall current of 200A?

Yes, i was referring to stall current.

I saw your uM H-Bridge design and it is great, too much option for me. I saw you are using LTC1155 for driving high side bridge. Well i'll try to get one of those cause in that way i dont need to think about gate voltage.

Excuse my ignorance, can you explain what are those Zener diodes are for between source and gate on FETs?

You've noted that you are limiting current on 20A or something, as i can see that is done by uC ? I guess that will not protect circuit from higher current if motors are stalled ? Maybe some kind of safety fuse could be installed somewhere between bat vin and upper FETs.


OT:
Funny thing, your Tumbleweed robot looks like mine with some differences. Well mine bot is build around Twin Detonator, using mini ITX as yours. Also i have steering on both shafts (which gives me a headache to implement control algorithm). Ok, currently my bot is inside the cabinet waiting for me to make him motor driver

 

[dknguyen]

That link was supposed to send you to 3 pages of information on H-bridges in general. Not the product itself.

The zener diode is there to dampen voltage spikes that can occur (exceeding the gate voltage and destroying it) that can form when you try to discharge the gate-source capacitance too quickly through the gate inductance. I think it also involves the source-gate capacitance somehow. They are not always necessary. There are a lot of things like snubbers, clamps, and what not you can add to a motor driver. But you usually only add them if you have problems or are expecting problems.

Always have a fuse.

Why wouldn't limiting the current to 20A with a uC protect the motor from stall currents? The motor just turns off before it ever reaches stall (whenever the current exceeds 20A).

 

[scoe]

That link was supposed to send you to 3 pages of information on H-bridges in general. Not the product itself.

I've read those pages and get general idea about h-bridge. Well some things i didnt knew from before. So the logical step was to see design it self. As i said it is impressive, but i'm not going to use it cause i dont need all the features implemented.

Why wouldn't limiting the current to 20A with a uC protect the motor from stall currents? The motor just turns off before it ever reaches stall (whenever the current exceeds 20A).

Well i guess it will be ok as long as sampling time is small enough. I just though that fuse will be safer, cause uC might get stuck in some procedure or interrupt while motor is on and loading momentum exceed output momentum or in other words motor get stalled. Just as a theoretical situation.

P.S.
i'm not questioning your design, just want to clear things up for me

 

[dknguyen]

Well i guess it will be ok as long as sampling time is small enough. I just though that fuse will be safer, cause uC might get stuck in some procedure or interrupt while motor is on and loading momentum exceed output momentum or in other words motor get stalled. Just as a theoretical situation.

THe fuse always works but it works less well (it reacts slower). THe uC can measure current and shut the motor off faster than the fuse will blow, especially if the current is just a little above rated. Best to have both.

P.S.
i'm not questioning your design, just want to clear things up for me

I don't care. It's not my design, nor my webpage.