Simulation of "Mazzilli Driver"
jpsmith123, Tue Jun 28 2016, 03:47PMI want to make a high power resonant driver (1500 watts or so) for one of my HVHF transformers - mainly to be used as part of a capacitor charging power supply (CCPS) - and I'm trying to decide whether to go with a Mazzilli ZVS driver, or a half-bridge "series load resonant" (SLR) type driver, based on an SG3525, or something like that.
I am able to simulate the primary circuit of the SLR driver, using the antiquated spice software "CircuitMaker", but I cannot get the Mazzilli circuit to simulate with this software.
I'm wondering, has anyone successfully simulated a Mazzilli driver with any spice program? If so which software did you use?
What I would really like to know is, what are the usual failure modes of the Mazzilli circuit and what are the possible shortcomings it would have for use in a CCPS?
Re: Simulation of "Mazzilli Driver"
Molenaar, Tue Jun 28 2016, 07:12PM
LTspice is free and lets you simulate a Mazzilli ZVS driver.
If oscillation fails to start somehow, current is only limited by your pcbtrace resistance, your mosfet RDSon and your source, which usually means your mosfets die.
Molenaar, Tue Jun 28 2016, 07:12PM
LTspice is free and lets you simulate a Mazzilli ZVS driver.
If oscillation fails to start somehow, current is only limited by your pcbtrace resistance, your mosfet RDSon and your source, which usually means your mosfets die.
Re: Simulation of "Mazzilli Driver"
Dr. Slack, Tue Jun 28 2016, 07:13PM
Oscillators are usually reluctant to start in a simulator, because unlike in real life, the balancing *is* perfect. You usually have to kick them into oscillation, by using 'initial conditions' on the capacitor, so that the simulation starts unbalanced. I'm not familiar with CircuitMaker, you should have initial conditions, but if not, then lightly couple a pulse generator into one side to kick it, or switch to LTSpice.
Dr. Slack, Tue Jun 28 2016, 07:13PM
Oscillators are usually reluctant to start in a simulator, because unlike in real life, the balancing *is* perfect. You usually have to kick them into oscillation, by using 'initial conditions' on the capacitor, so that the simulation starts unbalanced. I'm not familiar with CircuitMaker, you should have initial conditions, but if not, then lightly couple a pulse generator into one side to kick it, or switch to LTSpice.
Re: Simulation of "Mazzilli Driver"
jpsmith123, Tue Jun 28 2016, 09:57PM
Thanks for the replies.
@ Molenaar: I played around a little bit with a cheap Mazzilli driver from ebay, and the first thing I discovered was that it wouldn't oscillate if the applied DC power ramped up slowly.
@ Dr. Slack: The use of initial conditions was the first thing I tried, but the thing still wouldn't simulate in CircuitMaker. I didn't think of the pulse generator idea however. I'll try that next.
Edit: I just tried LTspice and that didn't work either.
jpsmith123, Tue Jun 28 2016, 09:57PM
Thanks for the replies.
@ Molenaar: I played around a little bit with a cheap Mazzilli driver from ebay, and the first thing I discovered was that it wouldn't oscillate if the applied DC power ramped up slowly.
@ Dr. Slack: The use of initial conditions was the first thing I tried, but the thing still wouldn't simulate in CircuitMaker. I didn't think of the pulse generator idea however. I'll try that next.
Edit: I just tried LTspice and that didn't work either.
Print this page