Current Source amplifier for low impedance loads

[Dr.H]

I have a very difficult speaker, Apogee Scintilla, impedance dropping to 0.7 ohms or so.
I'd like to DIY a CURRENT SOURCE amp, something like this:
http://apogee-acoustics.de/download/..._Apogee_EN.pdf

Any leads/ideas/schematics?

[AKSA]

Dr. H,

You have no choice but to use voltage source amps - that is conventional designs - with an output stage capable of welding steel. This might require a minimum of 12 pairs of 125W power transistors per channel, and propels you inevitably to a Krell or large ML.

This is the downside of owning Apogees. Fine speakers, but limited by the amps that can drive them, and by and large, such large amps don't do detail all that well.

Hugh

[Calvin]

Hi,

You may want to have a look at Elektors ´The current Amp´ from Nov/Dec 1992 that specialized on driving low impedance loads below 1Ohm.
It later became the ´Ulti Amp´ that was designed for more loads in the typical 4-8 Ohms range.

jauu
Calvin

[Dr.H]

Thanks guys.
Hugh I should just say that using the Krell KMA160 drives these speakers beautifully, but very low levels of detail (like the creak of a pianists chair) is a less audible than with normal 8R speakers with a good amp like the GB150.

[AKSA]

Dr H,

An amp that I believe would be eminently suitable because of the speed of its feedback loop (which is responsible for retrieving fine detail you note) I would honestly think about Lazy Cat's design. MiiB (Michael Boresen) did a very good build out using lateral fets which I believe would be highly suited to this application - but you would need six pairs and about 70V rails.

Cheers,

Hugh

[Dr.H]

Hi Hugh,

Funnily enough I posted the question about using Lazy Cats amp in that discussion but had no reponse.

[AKSA]

Threads are funny; you cannot know who or what will happen, and the oddest things capture people's imagination.

For reasons of speed, the current feedback amp is best, and this configuration is well suited to massive output stages because the multiple switching of all those devices causes loss of resolution in slower feedback loops. You can also get dual die devices for latfets from Exicon, and maybe four of these would be best.

Furthermore, the electrostatic drive of a fet gives excellent immunity from back emf coming off the speaker; high damping factor should scotch this but galvanic isolation does the job very well. This means phase shifts in the speaker don't mess with the feedback loop and the amp maintains composure at all times.

Just today it was suggested to use the CFP high bias driver devised by Mihai Rauta (aka Roender) so capably used in his amp of a couple of years ago.

This too would be ideal for driving six pairs of lateral fets because to avoid switching in the drivers the CFP runs at 100mA, and this is current aplenty to drive a big output stage.

Alex_MM from Romania does beautiful pcbs, and very quickly too, and might be asked politely if he'd have a go. However, compactness is not really needed here; you must space all the output devices adequately to avoid too much thermal power concentration, say every 2 inches on a very large, say 24" long by 8" high heatsink.

Hope this helps,

Hugh

[Dr.H]

That really does, thanks Hugh.

[Lazy Cat]

Originally Posted by Dr.H

Hi Hugh,

Funnily enough I posted the question about using Lazy Cats amp in that discussion but had no reponse.

The answer for you prepared

[Dr.H]

Thanks Lazy Cat!