I'm trying to get a stepper motor running using an L6470 stepper driver. L6470 is connected according to the recommended design. The motor is a NEMA23, rated 2A, inductance of 5.2mH and resistance of 1.9Ohm/phase. The L6470 is connected to an MCU the drives it. All parts are from decent suppliers and no black-market/fake parts.

At first, I used a lab power supply to generate 12V in order to run the motor. After some tests all seems to work fine, so I've changed the lab power supply with a 24V power supply (Main power supply for the whole system) and a DCDC converter just for this motor.

Things seems to work fine for 15 minutes before the DCDC is fried. This happened 3 times so I can rule out bad part. The coil on the DCDC is just warm and testing the coil after failure of the module

The MCU and L6470 work fine and and with the lab power supply the motor rotates as requested with no issues at all. Current never goes up above 1.5A (at least what my 100Mhz scope can measure over a known measurement resistor)

I've attached the schematics of the design. You can find the DCDC schematics here: Murate OKI-T/36W-W40

Any idea what keeps frying the DCDC? Back EMF maybe?

  1. Should I increase the 100uF capacitor?

  2. Should I have any protection against back EMF? Should I worry about Back EMF at all?

  3. Should I add input decoupling capacitors before the DCDC? will this help?

Thanks for your help.

The Design Design

Murata DCDC, TI IC got fried Murata DCDC TI IC got fried


It says on the Murata data sheet (under the maximum ratings) that

The outputs are not intended to accept appreciable reverse current.

This is a warning not to allow anything on that output that can raise the voltage and feed back power to the device.

The L6470 is a H bridge device using mosfets - the mosfets protect themselves by routing energy from the load to the positive supply rail. The parasitic diodes ensure this.

So what you possibly have, when you turn off one of the stepper motor windings is a back emf that raises the supply voltage over that which the murata brick is producing and forces something that can be described as an appreciable reverse current into the device.

The capacitors help a little but not much. The energy in the magnetic field (when it collapses) gets transferred to the capacitors which will limit the rate that which the line voltage rises. They can only limit the rate - they can't stop this rise.

Try putting a schottky diode in series with the Murata supply to the L6470 controller.

  • \$\begingroup\$ Tnx, problem solved \$\endgroup\$ – Gilad Dec 8 '14 at 18:08
  • \$\begingroup\$ What did you do to fix it dude? \$\endgroup\$ – Andy aka Dec 8 '14 at 18:14
  • \$\begingroup\$ I've added a schottky diode between the dcdc and the stepper driver. Seems to work good and no more fried parts. \$\endgroup\$ – Gilad Dec 8 '14 at 18:15
  • \$\begingroup\$ Excellent, glad it works and fingers crossed. \$\endgroup\$ – Andy aka Dec 8 '14 at 18:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.