Teensy 4.1 Damaged From Power Supply Voltage

[c172jeff]

With regard to my robot arm project and stepper motors, I decided to drive the stepper motors with 50 volts instead of 24. When I did this, I also purchased a new buck converter to accept 50 volts DC and deliver 5 volts DC.


Well, this probably destroyed the Teensy.

The DC 5 volts into the Teensy line has startup noise that is shown in the image below. After about 1ms, things seem to be stable. When you see below, is the first of several repeating patterns of noise on the 5 Volt rail. There is a frequency of 92 Mhz inside the shaded yellow. That splotch on the screen is about 4us and has a 50 Volt peak to peak section.

To be clear, the graph below is on the 5volt input line to the teensy and is from a dc to dc buck converter. The Teensey is toast.

Will this likely have damaged the Teensey? What is the best solution to this without designing another board or reverting to a different DC to DC converter from a lower 24 volts? What can the Teensey tolerate with regard to this kind of noise...length, frequency, pk to pk voltage? Thanks

 

[MarkT]

Can you point to the datasheet or pages about the buck converter you have? Some require minimum load and are not happy without this.

 

[c172jeff]

Can you point to the datasheet or pages about the buck converter you have? Some require minimum load and are not happy without this.

This is what I purchased. It takes 50V and is set to output 5.2V. The Teensy and Arduino were running for a short while. Perhaps there was another defect in the system? The first 2 stepper motors worked fine after startup. As well as the Pi.

When I operated the 3rd stepper motor, I got the magic smoke to come out. The third was a larger stepper motor.

Motors and 5 volt buck converter all pull form a 50 volt line. I think the power supply was limited in the amps it was to deliver (not turned up 100%). I'm now wondering if that was the problem. Amps not at 100% and a larger stepper motor caused a voltage fluxation on the 50 volt line which translated to the 5 volt line? All I can say, when I used the 24 volt setup and other buck converter, all worked fine repeatedly.

Amazon.com

 

[MarkT]

I suspect those cheap chinese buck converters don't have great layout and the result is instability, especially with the great reduction range you are needing, 12:1 or so. SMPS are very sensitive to layout.

 

[c172jeff]

I purchased [Chanzon TVS Diodes P6KE6.8CA 600W 6.8V DO-15 (DO-204AC) Axial Bidirectional Channel 600 Watt 6.8 Volt] as a fix. My understanding is that a TVS diode can respond in pico seconds. I am thinking that if I connect these between power and ground, the most volts that the Teensey 4.1 will see on its 5Volt input is 6.8 volts and for at most a micro second. I am not sure, but am curious enough to see if my input voltage is indeed clamped at 6.8 volts and does not explode to the high peak as seen above in the oscilliscope trace. The under voltage might be a problem for stability. I suspect that my problems were sollely due to that startup issue on the trace. It is a cheap chinese buck converter, but its probably fixable with some external capacitors, inductors, or maybe just one TVS diode. Worth an experiment. I wont be installing another Teensey unless I see that voltage capped at 6.8. Hopefully 6.8 is acceptable for the microsecond.