DRV103 EMI problems

Andreas,

I'm sorry you are having difficulty using the DRV103. I have some questions to help diagnose the situation:

Have looked at waveforms with an oscilloscope? Does the basic frequency look stable on the oscilloscope? You refer to unstable frequency and duty cycle. Can you quantify what you are experiencing? Do you see an unstable waveform as shown in the top of figure 6?

RF interference can be caused by layout issues. Connections to the load should be made with twisted pair. The +Vs side of the twisted pair should be connected close the the DRV103 and its bypass capacitor. It is often better to locate the flyback diode close to the DRV103, not at the solenoid. The selection of the flyback diode can make a difference in RFI. Some slower diode types can create RF noise as they turn off. A fast-recovery type should be used.

Regards, Bruce.

3733.Screenshots.pdf

Bruce,

thanks for your help.
in the PDF Screenshots you can see the waveform in differnt situations.

Figure 1: This is the waveform with no disturbance.

Figure 2-5 shows the waveform, when the conducted disturbance is activated.
in this case the circuit is acted with the conducted disturbance of 5V(80%AM(1kHz)) at the 24V supply-line .
The frequenz where the fault  appear's, is between 100MHz-300MHz.
The disturbance is coupled by a injection Clamp to the 24V supply-line.

We choose the test of conducted disturbance by EN61000-4-6 for the pre compliance test for the
electromagnetic field immunity test by EN61000-4-3.
We must test with 10V/m by EN61000-4-6. this should be equal 8-10V by conducted disturbance with
the injection clamp, because we have seen, that the disturbance is coupled by the lines to the unit.

The only way I found to protect the DRV103 against this disturbance is a extern Filter in the
24V supply line, but this is not practicabe.
Do you know a protective circuit at the DRV103 which solve the Problem too?

I think the problem is not causes by the flyback diode, because we have the same problems, when
we have a normal resistor as load.

Regards Andreas.

Andreas,

Thanks for the additional information. I'm not surprised that 5V of 100MHz to 300MHz on the supply voltage of the DRV103 causes problems with its operation. This level of high frequency voltage would be very disruptive to internal circuitry.

You are assuming that the radiated RF field has a fixed relationship to the conducted level on the 24V power line. I think the solution must be to change this expected relationship. This could come from a various measures:

• Improved bypassing of the 24V power supply. The correct choice of capacitors can make a very big difference. For example, X2Y capacitors can be selected to provide very low impedance in the sensitive frequency range. You may want to experiment with other capacitor values and types, located near the device power connections.
• Routing of supply and load wiring can make a difference. Are you using twisted pair connections to the load? Can these connections be moved closer to a ground plane? Can shielding be added?