Stepper motor deceleration causing large voltage spikes: 15-20V rise

I have one of the DRV8825 boards, which I am using to run a bipolar stepper at 1.4A in full step mode (vref=2). The motor’s only load is a heavy flywheel. My power supply output voltage is 20V, the supply is rated for 5A, and the stepper board is the only load. I have a 220uF capacitor right across the motor power (2mm away), and an additional two 470uF capacitors 2-3 cm away, along with a 0.047uF ceramic between the 220uF and the 470uFs to stop any high-frequency noise.

All is fine when the motor accelerates. However, when the motor cruises at high speeds (6000 steps/sec range), I see transient voltage spikes to about 25V that last in the 0.1s range. Even worse, when I start to decelerate the motor, the voltage on the supply line shoots up to 35-40V for about 1/4 second. During this time, the power supply current draw drops to 0.00 A for a brief period.

I don’t think this is a problem with just the DRV8825’s: I blew an A4988 in this exact situation even though it had the required capacitor, and I suspect that a similar spike blew up that driver.

Hello.

There is a webpage on the Phidgets site that discusses using a transient voltage suppressor (TVS) with a stepper motor that might be helpful for you.

-Nathan