New classD amp: Gemincore of CL3

[sx881663]

Jaka,
Much less physical mass getting switched at HF, yes a more elegant solution. The +/- 12V power could be a very small SMPS and shielded. I like the way the drive is bootstrapped and limited so a shorted load wouldn’t destroy the outputs due to over voltage. An output inductor would limit the amount of switching current, hopefully. I would imagine the inductor would get quite toasty though.
I don't understand the use of the internal inductor and what it does best. It seems as though the driver would be working into essentially an open circuit at switching frequencies. Of course the 12 volt supply common would need to be referenced to the output, this is not shown.
Roger

[sx881663]

Jaka,
I think the drawing says nh not uh for the inductor value, is this correct? This makes a lot more sense. Just slows the transition slightly, right?
Roger

[Jaka Racman]

Hi,

yes inductor is approx 400nH. It was 10turns on 2.5mm gapped toroid (approx .3mm gap).

Best regards,

Jaka Racman

[classd4sure]

Originally posted by koolkid731
@ Workhorse: Like I said, IMO it's not patented nor patentable, because I know too many prior art source follower circuits, Jacman's is but one example but dated 2/2004. He got TWO source followers in a same schematic. 'nuf said. I already cited the 70's totem poles.

On the subject of patents, many are allowed only due to limitation of knowledge of patent examiners. Most patents are nothing more than a piece of paper on the wall.

I certainly don't worry a bit about any patent. They do however provide inspiration for better inventions.

Ah HAH! You see? What they've done is find a more elegant solution to the known source follower such as Jaka's obese dual follower arrangement..

"I certainly don't worry a bit about any patent."
Bold statement considering..

It'd be interesting to know just how many patents would be violated these days given the most basic class d implementation.

[Workhorse]

Originally posted by koolkid731
@ Workhorse: Like I said, IMO it's not patented nor patentable, because I know too many prior art source follower circuits

KID , I already knew it , thats why i dont care about these damn patents!

K a n w a r

[koolkid731]

More ampliverter: US Pat # 7,015,754

Would be tough to implement such an impractical scheme.

http://patft.uspto.gov/netacgi/nph-P...=50&d=ptxt&S1=(330/251.CIOR.+or+330/251.CIXR.+or+330/251.CIUX,CIDX.)&OS=ccl/330/251&RS=CCL/330/251

[classd4sure]

Hi koolkid,

What is that you find impractical?

Here's a better link to it:

http://www.freepatentsonline.com/7015754.pdf

Given the caliber of it I'd question if they even got a prototype working.

[koolkid731]

Come on, floating gate/base drive, series diodes, and a couple of even more serious issues.

That one won't work.