VFD Output Snubber

[KCJoe]

Hello,

I have a piece of rolling equipment, it consists of one 7.5HP Motor and one 10HP motor. Each motor powers a separate roller, the load on the rollers spikes high for a brief time as we pass parts through it.

Each motor is controlled separately via a VFD, these are rated at 440V and the supply is 380V AC 3-Phase.

The supply to the machine is isolated via a transformer, steps from 208 line to 380V.

The issue is that I get a high voltage alarm on the VFD's which causes them to shut down. I assume it's when the load quickly drops from peak it causes a high voltage spike.


I'm pretty certain I need a snubber on each output of the VFD, what I don't know is what type of snubber I should use since it's a frequency drive.
How do I calculate the values for an RC snubber for this drive, or should I use something solid state.


Help will be greatly appreciated,

Thanks,
Joe

[SeanB]

If you are getting high voltage alarms you probably need more than anything else to connect the DC bus lines together of the inverters, if they are all fed from the same isolator and circuit breaker, and or install braking resistors on each unit, to clamp high voltage developed from the units absorbing power from the motor during slowing down of the load.

As the VFD absorbs energy this can only be transferred to the DC bus, and if there is nothing to absorb it ( like another inverter unit drawing power) or a braking resistor and chopper controller to modulate the bus voltage within the high voltage and operating limits, the bus voltage will spike, and the VFD will shut down to protect the power output stage from being destroyed.

No need for snubbers on the outputs, the VFD will operate fine without, and the only time you want this is if you have long power leads to the load and include a low pass filter LC circuit to reduce radiated emissions. On the input side read the VFD instructions on how to parallel them, and as you are feeding them from a local 3 phase transformer that is probably only rated for 15-20kVA you will likely need line reactors as well to control input diode current and improve efficiency.

[Benta]

My suggestion would be 3-phase load reactors between VFD and motor. Some information here:
https://www.ecmweb.com/design/line-reactors-and-vfds

[duak]

KCJoe, I agree with SeanB that from your description that the DC link voltage is rising enough to throw an alarm.  This happens because the VFD is trying to decelerate the motor or is trying to hold back a heavy load that is trying to accelerate the motor.  Does the VFD regenerate or have a braking resistor?  Connecting the DC links together is an excellent idea, especially if one motor is pushing and the other is receiving.  However, the AC line into the interconnected VFDs would have to be simultaneously broken to prevent a hazardous condition where one VFD provides DC power to all.

If it truly is voltage spikes on the load connections to the motor, a load reactor could help.  How long are the cables to the motor?  Is there a frame grounding connection between the VFD and the motor?  Is it in the same cable and is it connected solidly at both ends?

What are the make and models of the VFDs?

Cheers,

[Kalin]

I haven't heard about a voltage spike when the load is removed from the motor and it is still running at reference frequency. Typically you would expect a voltage spike if, for example the motor is running and it attempts to slow or stop a high inertia load. Is the line feed synchronized in some way? Is it possible that the stock being fed into the next roller is moving to fast and the motor is actually trying to decelerate the stock?

[Jeroen3]

Are the VFD's U/F or process controlled? What if they have overshoot in the control loop that it attempts to slow down? In that case it's just parameters to disable braking or improve stability.