Home Propeller Head Plaza

Technical and scientific discussion of amps, cables and other topics.

And now for a completely different point of view...

Hello Kurt,

I'm not normally a big fan of single-ended. For one thing, the power supply rejection ratio sucks unless you start getting fancy with current sources and the like. But then your simplistic approach goes out the window. However, in this situation (with the battery supplies), you don't need all that much power supply rejection ratio. So here's what I'd suggest:

FETs can be good, but there are a bunch of 'gotchas' to watch out for. I would avoid any of the modern vertical FETs. The input capacitance is extremely non-linear, especially with large signal swings. This only leaves a couple of choices -- lateral MOSFETs or JFETs.

The only lateral MOSFETs are the Hitachi devices (now only available in the plastic packages) and their clones made by Semelab (under a couple of different brand names). Both are fine for the N-channel parts, but the Semelab P-channels are much better than the corresponding Hitachi parts.

If you want to get wild and crazy, there's absolutely no reason you couldn't parallel a boatload of small-signal Toshiba JFETs together. Each one is typically rated at around 400 mW. So just parallel 100 of them (for instance) and you have a 40 watt device. This isn't as stupid as it sounds, only a 10 x 10 array. You wouldn't even need a heatsink. If you use the 2SK170s as followers, the total input capacitance would be 600 pF. This isn't terrible, but you would have to make sure it would work with your circuit.

I would also avoid servos. This isn't too hard. Almost everything Ayre makes is a true DC amplifier. For a well designed follower, it shouldn't be too hard to keep the offset below 5 - 10 mV. This is no problem at all for a loudspeaker.

Have fun!
Charles Hansen


This post is made possible by the generous support of people like you and our sponsors:
  Atma-Sphere Music Systems, Inc.  


Follow Ups Full Thread
Follow Ups


You can not post to an archived thread.