IM trying to replicate a hybrid switch in the half bridge rectifier to drive a pmsm machine. i have separate circuits of a turn on and turn off delay working fine in isolation with a simple pulse generator input. but they i try and implement the turn-on and turn-off in the half bridge rectifier, it doesnt seem to make any impact.
Am i doing anything wrong with the setup? is there a better way to do this?
It would be helpful to post a model and a description of how exactly you want the signals to behave. Attached is an implementation of a turn-on, turn-off, and a combined turn-on & turn-off delay:
onoffdelay.plecs (30.9 KB)
However, some users might be imagining a “non-deterministic turn-off delay”. See this forum post for additional discussion.
Hi Bryan,
Thanks for your reply.
Please see the screenshot of my circuit below:
What i am trying to achieve is turn on the MOSFET earlier than the IGBT and then turn off the IGBT earlier than the MOSFET.
I have those Turn-on/off delays setup do just that and they work on their own, in a similar circuit to the one you have attached,
However, in this configuration they dont seem to be making and impact, and when i try and add on turn-off delay to the MOSFET, it gives me the following error:
Many thanks,
Wahaj
What do you mean by “don’t seem to be making an impact”? If you look at the gate signals you should see the expected driving behavior. However, electrically the impact may be quite small unless you have very high electrical losses in your switch. Perhaps the differences will only manifest if you add a thermal loss model, which I suspect is the motivation for this arrangement.
The error message points to the fact that you are shorting the DC source because both upper and lower switches in a leg are closed at the same time. It’s important to note that the turn-off delay is a “latch” extending pulse. So, you need to tune the timings such that all switches in the upper branch are off before closing any devices in the lower branch (and vice versa).
You can add a non-zero switch resistance to eliminate the error message. This can be useful in debugging the modulation signals. Obviously your electrical simulation will have issues (excessive currents that would destroy your switch) until the underlying problem is resolved.
when i observe the switching on and off behaviour for both switches with only the turn-off applied to the IGBT, they both still turn-off at the same time (regardless of the magnitude of time delay i add).
I have added a time delay to the lower set of switches. when i dont add a turn off delay to the mosfets, everything works fine. but as soon as i add a turn-off delay, it says a short circuit is created (even through the turn off delay is much smaller time magnitude than the turn on delay of the lower set of mosfet switches.
One cannot debug further without a model and specific timing details, but in order to generate a short circuited voltage source both gate drive signals need to be closed.
Also note you should be using the “Voltage Behind Reactance” implementation of the PMSM which allows direct interfacing of arbitrary external networks and will behave correctly during the deadtime interval.