H-bridge with IR2110

Thread Starter

phongkem

Joined Nov 19, 2011
23
Hi everyone!
I trying to build a H-bridge with IR 2110 and MOSFET IRF3205 to drive a Motor which have peak current about 200 A (schemantic in the picture)
Hin1(pulse signal),Hin2(pluse signal),Lin1(level signal),Lin2(level signal) is from a Micro Controller. I use 4 battery for this test circuit instate of using bootstrap cap (just for test :)).
Everything is fine before plugging the motor to the circuit. I let Q1 and Q4 on; Q2 and Q3 off. without the motor, output of IR2110 is fine but when insert the motor Q2 is blown out :(.
I demounted Q2 and Q3 and used Oscilloscope to check the Lo1 and Ho2 pins and I found that Lo1 has pulse when the motor runs. The motor stops the pulse disapear.
What should i do now???
 

praondevou

Joined Jul 9, 2011
2,942
What diodes are D5 through D8?

Please post a picture showing the layout.

200A is a lot and the proper layout becomes very important at higher currents. Also, the continuous current limit (package limit) is 75A. What's the cont current you have?

Did you consider the layout guidelines HERE? section 6.
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
I mean peak current here just like the current when the motor start, this motor has its rate current just about 20A. I have read the guidelines and try to obey it. When I used a motor which has a smaller rate current it's just fine. I try to build this H-bridge to fit with motors which has power< 600W. What should I do to make it more stable?
 

praondevou

Joined Jul 9, 2011
2,942
Please post a picture of the layout.

Traces/wires from 12V_2 to 0V should be as short as possible with the peak current supplying component (a battery in your case, normally a capacitor though) as near as possible to the bridge.
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
I don't understand this point. What is the purpose of placing wires and traces as short as possible? It reduces something?
 

SgtWookie

Joined Jul 17, 2007
22,230
1N4148 diodes are rated for 200mA. They are woefully inadequate for this.

You say your motor is a 200A motor. The MOSFETs you are using are rated for 110A, but that is an estimate of the junction; the pins would burn out at ~69A-75A.

You need to read through this page a couple of times:
http://mcmanis.com/chuck/robotics/projects/esc2/FET-power.html

[eta]
You need to keep the distance between the components short to reduce the parasitic inductance.

200A is such a large current that you would need to connect the MOSFETs to a thick copper bus bar rather than a circuit board.
 

praondevou

Joined Jul 9, 2011
2,942
I don't understand this point. What is the purpose of placing wires and traces as short as possible? It reduces something?
Please read section 6 of the application note I posted. The purpose is to minimize stray inductance. If you pass a current (quite a big one in your case) through an inductor (the trace) and then stop the current flow (turn off the MOSFET) a voltage will appear over that piece of wire/trace. The higher the current the bigger the voltage, the longer the wire the higher the inductance and therefore the voltage too.

So if you were able to show us your layout we will possibly be able to give a better advice.
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
Thanks a lot, there are so many thing i have to recalculate. May be my power wire is too long then i will rebuild my circuit and post a layout to have better advices from everyone. But i wonder if in the worst case , wires to the supply can't get shorter so what I should do to reduce the inductance?
 

praondevou

Joined Jul 9, 2011
2,942
Thanks a lot, there are so many thing i have to recalculate. May be my power wire is too long then i will rebuild my circuit and post a layout to have better advices from everyone. But i wonder if in the worst case , wires to the supply can't get shorter so what I should do to reduce the inductance?
1. distance from upper mosfets source to lower mosfets drain as short as possible, trace width appropriate for the expected current
2. DC- bus electrolytic capacitors as near as possible from upper mosfets drain to lower mosfets source
3. if this is still not enough increase the gate resistor values (will be limited due to increased switching losses)
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
By the help of latter posts my Protoboard has worked well with a 20A motor. I also succeed in using boost trap cap(21uF 1electrode & 2ceramics) & diode(1N5819) instead of batteries for high side. Now I took a further test, a 40A rate current - motor was used for this test. All the signal to the Gate pins of MOSFET is just fine before pluging the motor in. But when the motor was pluged in the signal to the Gate of Low side disappeared even when the motor was pluged out(H-Side still fine). I had to reset the control board then the signal was fine again (without the motor). What should I do now?
 

praondevou

Joined Jul 9, 2011
2,942
When this happened did you still have the gate pulses at the input of the 2110s' ? It seems that the uC didn't send LOW pulses anymore. Did you check this?
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
The output of uC to the Lin pin still high level but the Lout was elimnated. This circuit still work well with other motors and that 40A rate current motor is still good. Any ideas else for this problem ??
 

praondevou

Joined Jul 9, 2011
2,942
The output of uC to the Lin pin still high level but the Lout was elimnated. This circuit still work well with other motors and that 40A rate current motor is still good. Any ideas else for this problem ??
Well I know that the output to the upper Mosfet can latch up, but this means it stays ON even if the input goes Low. I don't know if there is a failure mode as you described it.

Please post a picture showing your layout.
L
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
this is my layout. there are 2 modules like this, each of them is a half of bridge. J7 is a pin of motor. j3,c5,c6 are boost trap capacitors, j4 is boost trap diode, j8 is used to supply 12V for motor, j6 -> 12V for VCC. U3 -> 5V for VDD, R1->R3 1k pull-down resistors. j5 for control signals. r4,r5:1k Gate to Source resisor.
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
I have found out the problem and fixed this. This all about decoupling power source. But now I got a further problem. How much is enough for the delay time between turning off H-side MOSFET and turning on L-side MOSFET and are there any advices for building a delay-circuit like that?
 

praondevou

Joined Jul 9, 2011
2,942
I have found out the problem and fixed this. This all about decoupling power source. But now I got a further problem. How much is enough for the delay time between turning off H-side MOSFET and turning on L-side MOSFET and are there any advices for building a delay-circuit like that?
Nice you have it working. It would be nice to share what exactly you did to make it work so others can use it as a reference.

A fixed deadtime can be made like in this thread, post #14.
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
It is quite simple. In old circuit I have Vss pin and COM pin connected. It's the main reason make this circuit to be not stable. So I decoupled these two ground then this circuit work nicely with variety types of 12-24V motor :D.
 

Thread Starter

phongkem

Joined Nov 19, 2011
23
Your post help me much, thanks a lot. Could you please tell me the meaning of each component in the simulation circuit and how to chose the value of resistor and capacitor. Their value depend on frequency right?
 
Top