Commercial 1.2-1.5kVA SMPS for AMP2??

Status
Not open for further replies.
Pierre said:
Mmmm, bulky, isn't it?
They don't publish the dimensions, but it seems something like 230x200x55mm. And 3 lbs=1.5Kg aprox.
Besides, it is not what I call cheap: 329$ and you still have to assemble and test it!

I recall that Ghemink did quite a good bunch of tests with it. Perhaps he can tell us about his experience with that SMPS.


Yes, I have them. Had to debug them as the overcurrent protection did not work correctly. Modified the feedback loop as well to able to keep it stable with more capacitive load at the output. Also experimenting with additional LC output filter to get less HF switching noise.

Yes, they are big, about the size you mention, and they are indeed quite expensive as you have to build them yourself as well. Then can delivere a lot of power though.

I also plan to check out Sergio`s SMPS when available as it is a lot smaller.

Best regards

Gertjan
 
Portlandmike said:
Have you or anyone else ever pondered just using a UcD as a buck regulator to regulate the rails?

Its a thought. Don't know if it can do 1.2kW, but assuming its audio its likely capable, like a UcD400 or 700.

Obvously, you'll need one per rail.

The topology seems to be tolerant of capacitance on the output, so use one as a regulator, with a much larger output cap.

Regards,

Mike


Have been thinking about that myself as well, why two per rail? One per rail should be OK?

However, one reason you just can not add more capacitance at the output, say a few 1000uF, is that the the output LC filter would have a strongly reduced cut-off frequency and the whole amp may not work properly and would oscillate (if it oscillates at all) at a lower frequency possible damaging the amp (at least that is what I assume). Adding an extra LC filter may work though. The extra L would isolate the UcD feedback loop for high frequencies from the additional C that could be chosen big. So if the extra LC filter is dimensioned correctly, it should still oscillate at 400kHz and you can filter out that 400kHz switching component (about 1.2V peak-to-peak) with the extra LC filter.

Have fun but don`t blame me for blown UcD modules. Maybe others have better ideas/suggestions.

Gertjan
 
Status
Not open for further replies.