Hi all
Please have a look:
http://www.extron.com/download/files/whitepaper/class_d_ampl_wp.pdf
Interesting read on how to make class D without any output ripple, however I've got a gut feeling something's wrong in it. Cannot figure out what for now, maybe I am wrong and missing something elementary...
Need your opinions!
Please have a look:
http://www.extron.com/download/files/whitepaper/class_d_ampl_wp.pdf
Interesting read on how to make class D without any output ripple, however I've got a gut feeling something's wrong in it. Cannot figure out what for now, maybe I am wrong and missing something elementary...
Need your opinions!
Yes, that's the patent.
The whole idea is new to me. What's the nature of filtering?
Brickwall filters don't exist and if they do to some extent, they are not for free.
Do you know the crossover frequency between main and auxiliary output and probably the current ripple through the auxiliary output?
BTW, scope shots donn't look very nice at the paper...
The whole idea is new to me. What's the nature of filtering?
Brickwall filters don't exist and if they do to some extent, they are not for free.
Do you know the crossover frequency between main and auxiliary output and probably the current ripple through the auxiliary output?
BTW, scope shots donn't look very nice at the paper...
Well, I believe it is more about being a steep crossover passing audio current into load and everything above to a secondary circuit. No circuit can have such jump in impedance from nearly nothing at 20kHz to very high at 130kHz (switching freq,) and above (harmonics).
So probably there is still some ripple current, just passing into the other coil.
Please correct me if I am wrong.
So probably there is still some ripple current, just passing into the other coil.
Please correct me if I am wrong.
It could help that the amp's output signal looks more like that of a conventional linear amp - which can be a sales argument.
This could however be achieved with other filter topologies as well - which don't necessarily remove the ripple (which assists switching at low amplitudes) from the main filter coil.
Regards
Charles
This could however be achieved with other filter topologies as well - which don't necessarily remove the ripple (which assists switching at low amplitudes) from the main filter coil.
Regards
Charles
Nothing really new here. There is a Bloom and Severns article in the proceedings of the IEEE Power Electronics Specialists Conference (mid-80's, don't have it with me) covering the details.
There is no magic here. A complete analysis will show that you can add two extra poles to the file. In practical, there will always be a couple of parasitic zeros, not all of which are bad in practice; sometimes a little damping is a good thing.
Note that the benefit may be found in that the extra inductor need only carry the ripple and no DC, which can reduce size in some cases over a more conventional series inductor.
John
There is no magic here. A complete analysis will show that you can add two extra poles to the file. In practical, there will always be a couple of parasitic zeros, not all of which are bad in practice; sometimes a little damping is a good thing.
Note that the benefit may be found in that the extra inductor need only carry the ripple and no DC, which can reduce size in some cases over a more conventional series inductor.
John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.