CPZ1061 Mosfet getting shorted while powering off the SMPS
Hi,
I am using CPZ1061 instead of CPZ1075 in a design suggested in DER-943, as CPZ1075 wasn't readily available in my part of world. While testing I found that at the first time powering up the SMPS starts OK, gives desire output and works fine. But as soon as I make a power cycle (off then on) my CPZ1061 drain and source are getting shorted. I think while powering off the transformer giving a kick-back which is much more than what the CPZ1061 can handle, and it is getting in avalanche mode. Another point I have noticed is this happens when the a big capacitor(C4) across CPZ1061 is getting discharged very slowly. I have to use P2616 (Material CF139) Core from Cosmo Ferrite and have generated winding stack-up from PI Expert by inputting all the necessary parameters.
I have ordered CPZ1075 now and awaiting the deliveries, but I just want to make sure that it won't be happening with CPZ1075 again. There is no snubber for the CPZ in the design DER-943.
Can any one guide me through?
评论
Hi Aaditya,
The devices significant difference is their maximum voltage rating and power capability; the CPZ1061 is 650V rated, while the CPZ1075 is 750V. The transformer as you've suspected, may be the cause of the failure of the CPZ; leakage energy generated from the transformer may be too high causing a large spike that reaches the maximum voltage rating of the device.
Have you measured the leakage inductance of the transformer you're currently using? Is it within the range of the estimated leakage inductance from the PI Expert?
Also, the snubber in DER-943 is the CPZ itself; it is an active clamp IC.
Hello Aaditya,
Can you clarify what your output load level is when you tested this?
Based on how long the output voltage level was maintained, it seams that you tested this at very low/no-load.
If this is the case, then the charge held by C5 and the output caps is likely because the InnoSwitch is still switching. This would be because even though the input is switched off, the input caps still hold enough charge to maintain the output voltage and keep the device operating for 8-10s.
Thanks for your support. I have ran into a new issue... BP2 Pin rising to 324V (V(drain) level). What might be the point of failure here.
Hello Aaditya,
Using the CPZ1061 in the current configuration can cause the IC to be damaged. You can see a sample schematic on page 4 of the datasheet that shows two additional diodes and a capacitor(D1, D4, and C19) connected to the D and S2 pins of the Clampzero. These components are used to divert the transformer current from the body diode of ClampZero’s high-side switch to minimize the reverse-recovery energy. These components are no longer needed with the CPZ107xM components since these components use GaN rather than the silicon of CPZ106x.
I hope this answers your question.