drv8220drlr exploding, when powering a DC motor

[lavaman]

Hi, I am working on a DCC decoder, all things seem to be working fine, except the motor driver. When i wanted to test it on a 100% duty cycle, in the moment i connected it to the motor, it blew up leaving a hole in its case. The board itself continued to function normally no damage was done to other components, only the DRV8220. When i measured the output voltage of the burnt driver it was 0v. Can someone please help me resolve this problem? The driver is the only thing stopping me from completing this project. The motor I used is a regular dc motor for model railway trains. Power supply used was 16v dc. In normal operation, the power supply would be a 16v data line which will get rectified, the signal isolated and processed by the attiny.
Thanks in advance.

[DavidAlfa]

Maybe a short between VM and one of the outputs?
Did it work before?

[lavaman]

I don't think that is what happened. The only thing I did was connect the outputs of the driver to the motor. This circuit is working except for the H bridge IC. Before I fully tested only a proof of concept built with other components on a bread board. This is an PCB designed to fit in model trains.

[NiHaoMike]

Add additional capacitance from VM to ground to suppress inductive spikes.

[Benta]

The DRV8220 is almost impossible to kill.
Silly question, but... are you certain you mounted it in the correct direction?

[lavaman]

The additional capacitance on the Vm pin is a good idea, but I don't think that is the cause of the problem. Yes, I mounted it correctly, before i connected the motor i tested the outputs and tested the PWM with LED. All seemed to be fine, so I proceeded with connecting the motor, which killed the DRV.

[SiliconWizard]

The additional capacitance on the Vm pin is a good idea, but I don't think that is the cause of the problem.

I can't say for sure that it is, of course, but it's not something I would rule out.
Quoting the datasheet:

11 Layout
11.1 Layout Guidelines
Since the DRV8220 device has integrated power MOSFETs capable of driving high current, careful attention
should be paid to the layout design and external component placement. Some design and layout guidelines
are provided below. For more information on layout recommendations, please see the application note Best
Practices for Board Layout of Motor Drivers.
• Low ESR ceramic capacitors should be utilized for the VM-to-GND bypass capacitor. X5R and X7R types are
recommended.
• The VM power supply capacitor should be placed as close to the device as possible to minimize the loop
inductance.
• The VM power supply bulk capacitor can be of ceramic or electrolytic type, but should also be placed as
close as possible to the device to minimize the loop inductance.

We're talking about the internal charge pump, and if it's not stable enough, I wouldn't be surprised if that could kill internal transistors due to having transients during which both transistors of the same half-bridge would short the power supply and blow up. The chip is well protected, but probably not against a dodgy behavior from the charge pump.

[lavaman]

What if I put a PPTC in  front of the motor. wouldn’t that work in combination with installing capacitors between gnd and Vm and output pins of the driver.

[Tnixsw]

For the chip to "blow up" would mean some large inductive spike or high current flow. The DRV IC is supposed to have over current and over temp protection, so I would imagine you could connect up a car starter motor and it may just shut down because of the load. For the spikes, that may be a different matter.

What sort of current does the motor use?

Is there any heat sinking for the device?

How long are the leads to the motor.

Are you sure your software is driving it properly, maybe causing many start and brake cycles leading to spikes. The LEDs will not give much of a problem here, but a high current motor might.

Capacitor selection for the device is outlined in section 11 of the data sheet.

[CaptDon]

Are you sure this whole thing can run from steady D.C.? DCC on the model rails is never D.C., it is A.C. with varying duty cycle. Also, depending on the charge pump design you can only expect safe operation between 5% to 95% 'on' time. When you exceed those numbers the high side H-bridge FET begins to lose gate drive, goes linear and burns up. At 100% on-time I would expect possible failures at motor current levels, you may have gotten away with it just driving your LED. The DCC design has a 50/50 duty cycle on the rails at rest with about 16 to 19 A.C. volts very nearly a square wave. An old D.C. motor train will sit idle with its lights lit. When controlling the old D.C. train the motor is integrating the A.C. duty cycle and when it shifts one way or the other from 50/50 the motor begins to see it as a shift toward D.C.  I am not sure you DCC motor controller will work on a D.C. only input voltage? It may need the A.C. waveform of the DCC control standard to operate or clock internal charge pumps for the H-bridge hi-side FET's??

[DavidAlfa]

From the datasheet:

The inputs can accept DC or pulse-width modulated (PWM) voltage signals with duty cycles from 0% to 100%.

[thm_w]

What if I put a PPTC in  front of the motor. wouldn’t that work in combination with installing capacitors between gnd and Vm and output pins of the driver.

That is only going to help if the motor shorts out somehow.

Start with the recommendation SiliconWizard has given you, the eval board uses 44uF: https://www.ti.com/lit/ug/slou540a/slou540a.pdf?ts=1687825095632
Then maybe consider a ~10nF capacitor across the motor if you are worried about EMI issues (I think thats what you meant by output pins).

In normal operation, the power supply would be a 16v data line which will get rectified

If you are rectifying an AC source, you might need even more input capacitance for smoothing.

[ArdWar]

Start with supplying +12V with a stable power supply, not DCC, and see if it's working properly.