View single post by iceman
 Posted: Mon Dec 5th, 2005 06:33 pm
PM Quote Reply Full Topic
iceman



Joined: Wed Apr 6th, 2005
Location: Uzbekistan
Posts: 76
Status: 
Offline
Hi Sawdust123,

I can see from reading your post that you are very wise in the ways of transformer size reduction, and understand that higher frequencies are a friend of the power designer. Regrettably they are a curse to the rest of us. In fact, I was just suppressing some unpleasant thoughts about conducted emissions on a current project when I came across your excellent tutorial. May I be the first to compliment you on your succinct summary of the essence of inverter design, not to mention the very appropriate warning about not using a square peg in a round hole (pardon my metaphor). I would also add that anything that has low frequency magnetics (e.g. 60Hz transformers) do not like square waves, although you can get away with it on small power cubes if you keep them cool, and if the inverter doesn't choke (current limit) on the unexpected heavy load.

Regarding your question, my best guess is the FETs failed. Typically this is caused by excess heating in the devices, which happens if the snubber circuit fails to do it's job. The snubber fails when either the diode or the cap give up.

This thing that really bugs me about this is the design problem is probably very easy to fix. Something simple like "wrong PIV on the snubber diode" or equally trivial. However, if you don't look at your warranty center return data, you will never know that you have a problem, much less how to fix it.

I was hoping that the provocation directed towards Panasonic (Matsushita) in the invective above would stimulate a sentient response. Perhaps if I was a bit more explicit.

PANASONIC MICROWAVE DESIGN CENTER: I have a smoking example of what is wrong with your product in my garage. Send me your FedEx account number and authorization to ship, and it is yours.

Last edited on Mon Dec 5th, 2005 06:35 pm by iceman



____________________
iceman