L6470 Stepper Driver - Large cap on VS to ground destroyed??

[JBlackann]

Hello,

I have a circuit which uses the L6470 with a STM32F401.  I based the circuit off of the development board, X-NUCLEO-IHM02A1 which has 2 L6470's on board.  https://www.st.com/en/ecosystems/x-nucleo-ihm02a1.html

The X-NUCLEO-IHM02A1 circuit has a large cap on VS.  The documentation on the X-NUCLEO-IHM02A1 states the power supply can be between 8 and 45 V.  Our circuit is to be used at 24VDC.  The large cap on the X-NUCLEO-IHM02A1 (C7 and C23) is listed as 100uF and 63V.  On my design, I am using 100uF 50VDC.  We built some units and supposedly the units were in service for a period of time but then at some point failed.  I received two boards back that show damage to the L6470 and the large cap that would be on VS to ground.  The L6470's show burn marks on the top of the chip and on board the large electrolytic cap was blown off of the board.  When I checked the cap for resistance it read nearly a short (just a couple ohms).  On another board, the cap is still attached to the board but doing a continuity reading from the input voltage to ground, again reads as a short.  According to our customer, the units were only powered with 24 VDC.  My question is what could have happened to destroy the L6470 driver and the blow the cap off the board or short it.  The stepper motors that were connected to these boards have the correct coil values and work with another board.  I could see if the coils of the stepper motor were shorted the driver could be damaged but the stepper motors appear to be fine.

Any suggestions?

[Peter BENSCH · Accepted solution]

When the driver board is switched off, the motor has probably been turned in the opposite direction to its normal direction of rotation so that it acts as a generator. Its generated voltage is then first running through the source-drain diodes contained in each MOSFET to VSA and VSB or VS, which can lead to very high voltages there if the motor is turned with a lot of force. If you have not inserted a voltage limiter at VSA/VSB, e.g. a TVS, this excessive reverse voltage will then cause your capacitor to short-circuit and burst, which will then roast the output stages of the L6470, as you can see from the mini volcanoes on the destroyed devices.

Solution: Insert a strong protection against reverse voltages that can absorb sufficient energy without burning out. Suitable for this are e.g:

• between VSA/VSB and GND an unipolar TVS of sufficient size and with a well-chosen clamp voltage VCL (not greater than the maximum motor supply voltage of the L6470) on enough cooling surface to burn off the generated energy (unfortunately not included in X-NUCLEO-IHM02A1 for reasons of simplification)
• if necessary, additional bipolar TVS across each stepper motor winding

Please bear in mind that such generated energy can be several times the motor power, so that precautions must be taken to limit it somehow and thus protect the entire driver block. However, TVS are usually designed to protect sensitive equipment against electrostatic discharges according to e.g. IEC 61000-4-2, MIL STD 883 Method 3015, or electrical overstress as IEC 61000-4-4 and 5, so that they are certainly not sufficient in the event of wilful or intentional generator operation and other methods must be developed to burn off the generated energy.

Regards
/Peter