H-bridge blowing up in my face...

Thread Starter

cmartinez

Joined Jan 17, 2007
7,524
You have highly inductive load yet no any depfer circ to avoid the voltage spikes. Apply the resistor +diode+capacitor and probably will work better.
The back to back zeners connected in parallel to the solenoid act as voltage transient suppressors. That avoids the voltage spikes. Although I've been told that the inherent diodes in the Fets already do the same thing.
 

eetech00

Joined Jun 8, 2013
2,653
Hi

You can try this (circuit is untested). Try using complimentary pair MOSFETs for the H-Bridge.
The opto's remove the VGS(th) constraint when using (control) mosfets driven by an MCU. I've shown an Arduino with PWM at 490Hz. R5/R6 are to help minimize shoot thru and my not be needed. The circuit needs to be tested.

1632502223439.png



1632502260878.png

Edit: relocated current limit resistors R5-R8
 
Last edited:

Thread Starter

cmartinez

Joined Jan 17, 2007
7,524
Hi

You can try this (circuit is untested). Try using complimentary pair MOSFETs for the H-Bridge.
The opto's remove the VGS(th) constraint when using (control) mosfets driven by an MCU. I've shown an Arduino with PWM at 490Hz. R5/R6 are to help minimize shoot thru and my not be needed. The circuit needs to be tested.
Thanks, but your circuit is practically the same (in its working logic) as the one that's been giving me trouble. That is, in your circuit both low side nFets are turned on while idle, and both high side pFets are off. When switching is executed one side of the H bridge is very briefly short circuited, and that's when things go up in smoke...
 

eetech00

Joined Jun 8, 2013
2,653
Thanks, but your circuit is practically the same (in its working logic) as the one that's been giving me trouble. That is, in your circuit both low side nFets are turned on while idle, and both high side pFets are off. When switching is executed one side of the H bridge is very briefly short circuited, and that's when things go up in smoke...
Yes...I believe that is caused by differences in the specs of the P/N MOSFETs in your schematic (different on/off thresholds, etc).
The Mosfets I've shown are complimentary pairs (same package) and are closely matched, plus R5-R8 limit the short circuit
current. Anyway, these mosfets are probably hard to get right now....

Edit: relocated R5-R8 schematic #42
 
Last edited:

Thread Starter

cmartinez

Joined Jan 17, 2007
7,524
A good addition to this circuit would be a current limiter ... but I wouldn't know where to start on that one...
 

Thread Starter

cmartinez

Joined Jan 17, 2007
7,524
The circuit I showed has current limiting.
Yeah ... but if I were to use the same technique I'd get into trouble because those fets can handle currents of only up to 9 amps. Adding 0.1 resistors to the bridge would limit current to 80 amps... far more than what is allowed.
And besides, the solenoid needs 3.8 amps at 16V to work properly.

I need something a little more sophisticated than just low value resistors for limiting current.
 

eetech00

Joined Jun 8, 2013
2,653
Yeah ... but if I were to use the same technique I'd get into trouble because those fets can handle currents of only up to 9 amps. Adding 0.1 resistors to the bridge would limit current to 80 amps... far more than what is allowed.
And besides, the solenoid needs 3.8 amps at 16V to work properly.

I need something a little more sophisticated than just low value resistors for limiting current.
Well...you can size the resistors and mosfets to whatever you need.
I'm wondering how precise the current and pulse duration really has to be..(?)
There must be some minimum operate voltage and current specification for the solenoid.
 

Alec_t

Joined Sep 17, 2013
12,317
I can't claim originality for that circuit configuration.
You might want to add a bit of series resistance in the NMOS gate leads, or else use buffer transistors for the CW and CCW drive signals, if you're concerned about the MCU outputs being stressed.
 
Last edited:

Thread Starter

cmartinez

Joined Jan 17, 2007
7,524
My one concern with that last circuit is that inductive kickback might switch on the upper pFets when the lower ones turn off.
 
Last edited:

Alec_t

Joined Sep 17, 2013
12,317
Hopefully the PMOS/NMOS body diodes would suppress any kickback and prevent that ...... but theory and practice don't always agree.
 

crutschow

Joined Mar 14, 2008
28,203
Below is the LTspice simulation of the Alec_T's circuit in post #51:
Note that the MOSFET body diodes suppress the inductive transient (blue and purple traces), keeping the essentially between 0V and V1.
The adjacent upper MOSFET does have 0V applied to its gate when the inductive transient is occurring, but that is after the bottom MOSFET is already off.

1632663166675.png
 

Thread Starter

cmartinez

Joined Jan 17, 2007
7,524
True, but they are not really needed for this application, since when the lower MOSFETS are off, it doesn't matter what the gate bias is on the upper MOSFETs.
The circuit should work the same if those resistors were to be connected according to the original diagram, right?
 
Top