If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

# DRV2667: Using the DRV2667 to Drive Actuators with Offset Waveforms

Part Number: DRV2667

Hello!

I am interested in using a TI driver to control a PiezoDrive BA3502 bender actuator. I plan to wire the actuator either using a biased-unipolar/three-wire or an unbiased-bipolar/two-wire method. Here are some images that show the difference between the two approaches (images are modified versions from PiezoDrive):

With the three-wire version, VBST would be a constant 100V, and OUT+ would be 50V amplitude, 10Hz sine wave with a mean voltage of 25V. With the two-wire version, OUT+ would be a 25V amplitude 10Hz sine wave with a mean voltage of 12.5V, and OUT- would be the same signal but 180° out of phase. These two approaches should result in identical 10Hz cantilever bending motions in the actuator.

I'd also be able to offset the midpoint of the bending. For the three-wire approach, I could accomplish this by, for example, increasing the OUT+ signal's mean voltage to 35V while retaining its 50V amplitude. For the two-wire approach, I'd do the same thing but not increase the OUT- signal's mean voltage. Is it possible to command the DRV2667 to do this using I2C? If not, can either be done using the analog input?

Finally, I'd also like to make the actuator have a constant offset. For the three-wire approach, this could be done by changing the OUT+ signal to a constant amplitude and frequency to zero, and varying its mean voltage. For the two-wire approach, I could get a constant offset by doing the same thing to the OUT+ signal but not changing the OUT- signal's mean voltage. Is it possible to command the DRV2667 to do this using I2C? If not, can either be done using the analog input?

• Hello,

Thanks for reaching out. I am looking into your question. I will follow up within 24 hours.

Regards,
Sydney Northcutt

• Hi,

The three wire approach will not work with DRV2667. If you set your boost to 100 V but configure your gain so you have a peak to peak waveform of 50 V, there will be an unavoidable DC bias. You could add a capacitor to filter this out, but that interferes with your desire to have a DC offset later. I recommend going with the two wire approach.

For changing the midpoint, it sounds like you are describing altering the duty cycle of the signals. This would work but there could be some clipping depending on your overall configuration. It will require some testing. I believe the analog input will be the easiest option for control.

The constant offset sounds like you are wanting to drive a DC signal which is also possible with DRV2667.

Regards,
Sydney Northcutt