Cyrill Hammer (Souloution)
"if
you want to have your product performing at the cutting edge it is not
possible with today’s known switching technologies. In order to come
close to the performance of the best linear design we would need
high-current semiconductors that provide switching frequencies of
several MHz or even GHz."
This is the one, that I
believe hits the nail on the head, and why to some that hear it Class-D
sound in the upper ranges hasn’t come of age yet.
Only Technics
with their unobtainable $30k SE-R1 comes close with a 1.5mHz switching
frequency instead of what all the other are using today 400-600kHz. But
as Cyrill Hammer states it should be even higher.
The quote from Mr. Hammer is false. Here's why:
As @merrilaudio states, the problem is dead time. Dead Time is the delay time that has to be inserted in the circuit so that both output devices are prevented from being partially on at the same time. If this isn't done, a phenomena called 'shoot through current' occurs. This causes the output devices to heat up and fail. The problem is that dead time causes distortion. So with any given output device, there is always a certain minimum distortion and associated maximum switching speed.
Technics sort of got around the problem by brute force- going to an output device that was so much faster that they could reduce deadtime and also switch faster (BTW their 150 watt version of this is about $18,000....). However eliminating dead time is what works- you can actually have lower switching frequencies with lower distortion if you don't have to have deadtime circuits!
As some of you may know, we've been working on a class D for a bit over a year and a half. Although we are known for tube gear, in particular OTLs and balanced tube preamps with direct coupled outputs, it turned out that we seemed to have something to bring to the class D table. May a year ago we had working concept, today we have prototypes that demonstrate how a class D sounds if deadtime is eliminated. Our amp is not based on anyone's modules and we have a patent filed with another on the way. So far the prototypes have switching speeds between 250KHz and 500KHz. The amp is very smooth and extended, lacking grain or harshness. It easily compares to many tube amplifiers in that regard. It is zero feedback and exhibits soft clipping.
We don't think its ready for prime time yet; in that regard I've yet to hear a class D (including the Technics) that is, but we are very excited about our class D as it sounds better than any we've heard. I think merrilaudio knows what we are talking about; eliminating deadtime is the single most important hurdle to cross in class D amp design (especially if you can run MHz+ switching speeds), and I'm going to go on record that any amp that employs deadtime to work is an amp that will become obsolete.
