TMS320F2806x with Instaspin - New motor causing failures

1. What type is the,motor? ACI or PMSM?
2. What rated power, current, voltage is the motor?
3. What lab project is running?
4. Did the identification of motor run well using lab02a? Is the identified parameters of motor correct?
5. What's the failure? Over cuurent? or motor vibration?

1. What type is the,motor? ACI or PMSM? 

Both are PMSM (BLDCs).

 

2. What rated power, current, voltage is the motor?

Power: 1000W

Current: 1.8A_rms

Voltage: 380A_rms

 

3. What lab project is running?

In our tests where things fail, we are running our software that utilizes instaSpin/MotorWare.

 

4. Did the identification of motor run well using lab02a? Is the identified parameters of motor correct?

Yes the identification appeared to go well.

The Resistance and Inductance parameters obtained were within 10% of the manufacturer values.  The rated flux was one order of magnitude off from the manufacturers value.  Typically, I lean toward the TI software determining the correct flux and utilize that.

 

5. What's the failure? Over current? or motor vibration?

The FETs break and require being replaced.  The motor can spin up and is somewhat stable at lower frequencies but breaks the FETs at higher frequency.  Mind you, the axis this motor is spinning on is rated for up to 3.5A_peak and the current limit (i.e. max current) is set at around 2.5-3A_peak.

(Maybe this is not an instaspin issue.......I'll ask for scopeshots on the FETs in case there is a SOA breach or something)

I asked some follow-up questions, so here is more info in case you guys have ideas:

1. What is your Motor Voltage? (your VDD rail at the drain of the high side FETs)
430VDC

2. Is there anything about this motor’s parameters that is very different compared to the others? (inductance, resistance, flux)
The only one that is very different is the flux. (Approximately an order of magnitude)

3. How which of the 6 FETs break?
The high side V FET

4. Can you share the exact FET part numbers are you using?
FCD620N60ZF

5. Is it possible for you to collect scopeshots of one of the breaking FETs, triggered on the fault somehow? If yes, VDS, Drain current, and the control signal driving the FET would be useful.
We may be able to, our scope has the bandwidth but I’ll have to see how easy it is to set up to try to catch the fault

So I understand there is a good chance the issue here may lie in the FETs or the drivers or something, and I am working that seperately, but Is there any comment you guys can make? At the very least I would like to be able to definitively rule out the C2000 completely so anything you guys can think of to try would help.

When you say phase voltage sensing, do you mean that they have inline current sensors? Or are they doing low side voltage sensing? You mention both in your original post so I want to clarify.

I'm guessing this is custom hardware that the FETs blew on? And custom software? If so, obtaining a baseline on a known software/hardware combo (such as lab11 with HVMTR kit) would be a good place to start. If that works, try next with your custom software. If that works, we can move on to debugging the hardware.

And as you mentioned, scope shots of what's happening on the FETs at failure will be a good resource for us. Does this happen under constant load conditions?

Sean

Brian,

What was the outcome with this? Are you still experiencing issues at the customer?

Sean

Taking this offline