Full bridge circuit is built using IR2110 and DC motor is moved but spikes at o/p

Thread Starter

azorfa

Joined May 26, 2010
3
Sir,
With the circuit implemented to drive the DC motor using the two IR2110 IC's constructed in a full bridge form,
we are able to drive the DC Motor with the deadtime generated at the LIN and HIN inputs. But the output signals at LO and
Ho, with the CRO probes at the input of the Gate resisitors (22Ohms) the pulses we obtained are as shown in the attached files. Even though the dead time is generated, before the lower MOSFET turns off , the Upper MOSFET is being turned on, resulting in a very large spike at the current sense resistor. A diode resistor in parallel are also connected at the Gate of MOSFETs. I think , the MOSFETs we are using (IRF540N) are able to withstand this large current, so as per now we are not facing the short circuit problem. But at times, may be due to this reason our entire Full bridge section is getting burst and fail.
Please suggest me how to solve these large current spikes, and how to obtain a noisy free PWM waveforms at LO and HO.
PWM Frequency : 24KHz
MOSFET : IRF540N

Please acknowledge me at the earlier.
Thanking You Sir,
 

Attachments

Thread Starter

azorfa

Joined May 26, 2010
3
I have attached the circuit we are using to drive the motor, please suggest what has to be done to reduce the spikes. One more thing i have observed is when not applied the 70V to the mosfets, the spikes wont appear, but on giving the voltage, i can see the spikes. Please go through the waveforms i attached in the previous mail.

Please help me if anyone has faced this problem.
 

Attachments

SgtWookie

Joined Jul 17, 2007
22,230
You do not have reverse-EMF diodes across the MOSFETs.

Relying on the MOSFET body diodes for the reverse-EMF will cause the MOSFETs to dissipate a great deal more power than you would normally expect, particularly in the ~40% to ~80% PWM range. Use fast-recovery diodes rated for the average current that you will put through the motor at 50% PWM.

You are attempting to use a single current sense resistor. You cannot do that. You must use separate current sense resistors for the low-side MOSFETs. Otherwise, you will not be able to measure the current through the motor accurately.
 

Dead Short

Joined Apr 30, 2010
4
Hi Sgt W

Do you mean one should use two separate resistor (one off each low side MOSFET to GND) and then a separate opamp to measure the current and feed to the micro?

Some say that the Hi and Lo side can be driven simultaneously and others say rather switch Lo side to GND and then drive the Hi side with the PWM - whats the recommended practice?

Have been asked by the boss to ensure that I have over current sensing and shut down across each MOSFET - are there any recommended techniques for this? I could just use a more fancy IR device but we struggle for some components in South Africa so am tryin to stick to simpler devices such as the IR2110.

Regards

DS
 

Thread Starter

azorfa

Joined May 26, 2010
3
Thank u for ur reply Sir,
I checked with the fast recovery diodes (HFA25TB60) across the mosfets, but of no change. I have changed the bootstrap capacitance to 0.22uF and also with 1uF, and i reduced the Gate resistance to 10Ω and also to 47Ω but i am finding no change. I am not able to understand what is the mistake i am doing. As u said i have seperated the current sense resistors also. . Please suggest me what should be done.
:(
 

SgtWookie

Joined Jul 17, 2007
22,230
Thank u for ur reply Sir,
I checked with the fast recovery diodes (HFA25TB60) across the mosfets, but of no change. I have changed the bootstrap capacitance to 0.22uF and also with 1uF, and i reduced the Gate resistance to 10Ω and also to 47Ω but i am finding no change. I am not able to understand what is the mistake i am doing. As u said i have seperated the current sense resistors also. . Please suggest me what should be done.
:(
Try using SMALL capacitors across the MOSFETs, from drain to source. Start with 220pF. Don't go higher than 1nF (1,000pF or 0.001uF), or you will wind up with high power dissipation in your MOSFETs.

Keep in mind that the only way to initially charge the bootstrap capacitors is to turn OFF the high side of the bridge, and turn ON the low side for a period of time.

After the bootstrap cap's initial charge, you will still need to re-charge them. If you leave the high side MOSFETs on for very long, the bootstrap caps will discharge, and you will wind up with high power dissipation in the high side MOSFETs.

You may wish to slow the MOSFET gate discharge time. This will result in higher MOSFET power dissipation, but will lower the reverse EMF spike seen when the current is turned off to the load.

As long as the reverse EMF spike does not exceed the supply voltage or go below ground by more than a volt or so, you really don't have much to worry about.

Note that the caps for Vdd should be significantly larger than the bootstrap caps, as the bootstrap caps are charged from them. I suggest using a 0.1uF (100nF) metal poly film or ceramic cap in parallel with a larger capacitor.
 
Top