Bizarre DIN RAIL SSR Failures in PWM mode to motor.

[Tony Stewart]

A buddy designed/built a product with 21 of above SSR's using PWM control to 21 "motors"** controlled using 1kHz PWM from Arduino. All SSR channels failed with a thermal short-circuit inside when the load current rating was believed to be within the SSR specs using a 24V supply with a 60V rating.

When one SSR was opened up, it revealed that the design was completely different from the simple schematic screened on the outside. Instead of an optoisolator and BJT, it used a quarter sized printed coil with a small coil centered on the other side. The SSR was used in the high-side of load and the failed part appears to be an SMB TVS across the output of a 50A rated Nch FET on a 1W eff. copper area.

Tear-down with step ramp Vdc tests begin tomorrow. The PWM rate is not easily changed but may be too fast for this motor with excessive flyback energy. I plan to add a flyback power diode to 24V rather than rely on the TVS.

The motor** is actually a magnetic brake to a rotating shaft with 320 W mechanical load rating.

The specs of 0.74 - 3.0A and power to 2.6W seem to be inconsistent.

 

[alec_t]

The currents specified in the second row of that table don't seem to tally with the '0.74-3.0A' figures either.
Which of those 6 models of brake is involved here?
The SSR output is only rated 2A. If the brake does indeed take 3.0A, which it might well do at power-up or when rotor-locked, then that would explain the SSR failure.

 

[Tony Stewart]

It is the 320 W model. This current controlled magnetic slip brake appears to be an electromagnetic (solenoid) brake driving a rotating disk pulled into a friction pad where coil current controls the torque independent of RPM. The test was postponed but I did a simulation done at home with RLC measurements.

 

[rjenkinsgb]

It's likely the repetition rate that caused the failure, with the designers expecting enough time for the snubber device to cool between switching cycles.

With PWM drive, the snubber dissipation would be far higher and it sounds like heat just accumulated until it failed.

A flywheel diode across the coil should solve it, as you say.

We manufacture some little power switch modules for controlling clutch-brake units with a typical rate from 0.5Hz to possibly 3Hz at most, and the snubbers (disc VDRs) are by far the largest components in them.

There is no way I'd consider PWM drive to an inductive load with anything other than flywheel diodes.

 

[Tony Stewart]

it's a friction clutch with a power rating of 330W at max temp
where P= HEAT (watts) = RPM x TORQUE/k
I reduced the PWM rate to 30 Hz in simulation and the L/R=T= 4ms is also the flyback time constant. but with only 1.2A peak, The TVS or a Power diode didn't get too hot. So I am surmising the SSR RF pulse envelope detector possibly used did not react well to 1 kHz PWM rates. New Arduino code generated and tests by buddy TBD.

The TVS (zener) being polarized doesn't seem to dissipate much about 125 mW average 1.2Wpk in simulation. The TVS are rated for 300Wpk and 1Wdc, This does NOT explain failure mode yet, so I suspect RF couple is acting weird. when PWM f is too fast.

https://tinyurl.com/26eoyhjy

 

[Tony Stewart]

Problem solved with flyback diode then added shielded sensor input wires with the single ground and fat braid for hinged front door with control dial using shielded wires from a quadrature dial pot He was using single.-ended differential drivers (RS-485) with two different signals in the same strait shielded wire which also caused crosstalk. I recommended a shielded twisted pair or STP wire. He's a Mechanical Eng. with some good electronic skills but limited on theory. Really nice looking layout with Arduino, a plugin SSR for breakout strips for 21 motor drives, an LCD touchscreen for selecting the motor and a rotary handle for the pot to adjust the motor brake tension on the wire winding spool.