Another technobabble alert
Just a side note here. During the 12 VDC testing a few days ago, I noted a 1 MHz pulse train at the start of each switch-on pulse - it ran for about 16uS which is significant at the 64uS PWM rate I was running (now 9mS).
Here is a picture of the rubbish at the switch on point measured at the gate drive output. This is with my 9mS PWM rate (110Hz).
I tracked it down to excessively long clip leads going around my test bench causing the switch pulses to get back into the control analogue input. I largely suppressed the effect with a 10nF capacitor across the TL494 voltage op-amp output to -ve input; but it may have been the cause of the IGBT failure. I added the cap. before I realised that it was the leads causing it - but left it in anyway. The input control wires are shielded cable so maybe it wasn't the cause of failure - anyway - worth mentioning.
End technobabble (largely)
Once the test was done I shut it all down, quickly unplugged the 600 VDC connections, lifted the lid (not screwed down yet) and tested the IGBT and various other components for temperature - the finger test. The IGBT heatsink was cool and the only hot components were the two current surge limiters (they are supposed to run hot) - and of course the ceramic element which had already cooled so it was only warm to the touch.
That done I took the box inside and stuck everything that might move firmly in place with the red RTV silicone I've come to know and love.
The heater box is ready to re-install.
No comments:
Post a Comment