Is there any reason not to use ADUM1100 series magnetic couplers instead of fast opto like HCPL-7723?
Interesting idea, however adum1100 contain on-chip pulse transformer, which could be not enough stable if close to the switching stage.
That was my first tought also but...IVX said:
Specs look promising, immune to magnetic field for example 5mm apart from conductor carrying 500khz at 1kA! Unless you put them between airgap of output inductor its hard to mess them up. 😀
Pulse widht distortion is maybe bigger with ADUM-series because? of input schmitt-trigger/hysteresis.?
Well, It seems adum will ok too. BTW, why you don't like HCPL-7723, or usual cascode voltshifter even?
oops, forgot this..IVX said:
I feel that cascode thing is more prone to blow up in high-power designs. And in some cases complete isolation of drive signals is nice to have.
HCPL-7723 looks fine exept bit expensive. Adumxxxx seems cheaper at digikey for example. Optocouplers in general have problems to withstand high dV/dt common-mode transients and archieve high speeds. HCPL-series is ok but expensive.
mzzj said:
oops, forgot this too.. 🙂
I suspected a cascode in my project also, but the THD, measured yesterday, is pretty nice -.0044%@50Hz/220V/100W. 😎
Seems, that cascode not so bad yet, in my case (370V 100KHz) at least.
To be honest, complete thing would be worse, because reference singen will be use Atmega8 with .007% THD own. (at moment EMU1212->amp->EMU1212)
This is also some sort of magnetic coupler but working with a different principle than the Adum. It comes in singles and two kind of doubles. It is a little cheaper than the fast optocoplers with high transient-immunity.
Regards
Charles
Regards
Charles
- Status
- Not open for further replies.