Hi Paul, thx for sharing, curious what the motivation is here though (good learning experience if nothing else). Couple thoughts (hope you don't mind):
The original $200 charger thread was also a buck converter. IIRC using a mig reactor at 10s of khz, doing 350v and 35A. It morphed into the multi thousand dollar EMW.
The onboard (inverter) charger thread could use a buck stage, relegating the motor coils/bus caps to 60hz pfc. I'm gonna be busy refining the series cap approach (and other sub-projects) for my own purposes. Getting a charger sorted was the missing piece.
You are going to be playing by yourself if it is all done in Mentor Graphics.
While 8 bit is probably sufficient, there are lots of offerings from atmel (and others) that could probably reduce the external parts count a bit. (differential adc, 20x gain, etc).
There should be some economies of scale, i.e. the "power" section is the only modular part.
Toroids are nice at higher freq. powdered iron will handle higher power levels better than ferrite.
The relay/cap thing is kinda perplexing.
Have you considered the pros and cons of buck-boost (i.e. CUK)? Buck or boost only are always of limited flexibility. A well considered cuk can have good power factor without huge storage caps and be adaptable to a fairly wide range of input/output voltages with one switching element (if I understand it, probably some gotchas I haven't considered yet). Some of the most efficient converters are CUK.
The original $200 charger thread was also a buck converter. IIRC using a mig reactor at 10s of khz, doing 350v and 35A. It morphed into the multi thousand dollar EMW.
The onboard (inverter) charger thread could use a buck stage, relegating the motor coils/bus caps to 60hz pfc. I'm gonna be busy refining the series cap approach (and other sub-projects) for my own purposes. Getting a charger sorted was the missing piece.
You are going to be playing by yourself if it is all done in Mentor Graphics.
While 8 bit is probably sufficient, there are lots of offerings from atmel (and others) that could probably reduce the external parts count a bit. (differential adc, 20x gain, etc).
There should be some economies of scale, i.e. the "power" section is the only modular part.
Toroids are nice at higher freq. powdered iron will handle higher power levels better than ferrite.
The relay/cap thing is kinda perplexing.
Have you considered the pros and cons of buck-boost (i.e. CUK)? Buck or boost only are always of limited flexibility. A well considered cuk can have good power factor without huge storage caps and be adaptable to a fairly wide range of input/output voltages with one switching element (if I understand it, probably some gotchas I haven't considered yet). Some of the most efficient converters are CUK.
