New Digital Switching Amplifiers


Anyone had the luxury of comparing any of these fine digital switching amplifiers to eachother or to other high end amplifiers such as Pass, Krell etc?

Nuforce
Channel Islands Audio
PS Audio GCA
tpk123
Yup, IMHO don't you believe it when they say digital amps can filter out the HF switching noise. Also the PS Audio was very sensitive to line conditions and I had to remove my surge protector.
How do you define "filter out the HF switching noise"?

Just curious.
Ar_t my question exactly. From the tone of your question I'm guessing you have something else on your mind :-).

If you get Stereophile, check out Aug 2005 page 114 for Channel Islands D-100 specs:
"But you can see from this graph how the ultrasonic output rolls off presumably due to the low-pass filter necessary to minimize the H-F noise produced by the switching output stage.
Even so, a relatively high level of this noise is still present in the D-100's output as can be seen in fig 2."
Cdc, lest anyone get an incomplete picture of the D-100's ultrasonic noise test measurements from your quote of the Aug. 2005 Stereophile review of the D-100, here is a quote from John Atkinson's follow up in the September 2005 issue of Stereophile in which he tested current production D-100 amps as resubmitted by CI Audio:
Whereas the original sample had started out cold with about 0.05% THD+noise (including the ultrasonic spuriae), this increasing to 0.1% after an hour, the new sample began at 0.027% and increased only very slightly, to 0.0285%: a promising beginning.
I understand that my digital amp output has some ultrasonic signal that would not be expected from a linear amp. Last time I checked "ultrasonic" means you can't hear it, but it does keep the bats away. :-)
I'm glad we could bring back this thread which was on its way to oblivion because it is an interesting topic.
Ar_t, maybe I didn't understand your question. I was refering to filtering the HF switching noise the amp itself produces.
Tvad, even the original distortion specs you mention are ridiculously low and probably people wonder how can audiophiles argue over such tiny changes in distortion. The actual printed graphs in Fig 2 however look a lot worse than numbers on a page.
Probably the worst Pioneer receiver is 75% accurate and the best Halcro is 90% accurate so any differences are small on an absolute scale. Nothing is perfectly accurate and you make a choice based on 1) your overall system, 2) preferences, and 3) what you find objectionable.
So if tpk123 is making a buying decision, then the objective is to weed out "bad" products, however small the differences, and buy the right product based on 1), 2), 3).
When I had the PS Audio, it gave me headaches like when I listen to CD's from the Cranberries. Could've been my system, room, CD's, or ears, but it didn't sound as good as "analog" S.S. amps IMS.
Even if they can filter out the HF noise, I'd rather not have a filter used at all. One more component in the chain to mess up the sound.
Also forgot to mention the specs were done with a steady 8 ohm simulated load. Real life loading would probably bring out the switching noise even more.
For an interesting read see the paragraph on "The Quest for that Old-time Religion " section:Hammer Dynamics Super 12. Dick Olsher talks about lowering the tweeter's x-over frequency. During *dynamic* passages "while such a solution results in a smoother frequency response, often the sound quality is far WORSE because of increased distortion. That poor tweeter is made to work much harder than it really wants to. In the end, it is reduced to painful screaming".
I'm saying this as an analogy to Stereophile's steady state 8 ohm load of the C-I amp versus real music.
Don't get me wrong, for the price, etc. I'm sure the C-I amp sounds really nice, it's just not perfect.
There are 2 kinds of noise:

One that goes out on the speaker leads at the switching frequency. We'll say in the 500 kHz range. (I believe Tripath and Nuforce may be closer to 1 Mhz.) The other is all the crud in the 60-100 Mhz region that will mess up TV reception ans some other things. Two different issues, 2 different solutions.

As for distortion changing with warm-up:

You can get the distortion to drop if you increase the "stand-by" current drain. Problem is, you run the risk of destroying the switching transistors if you get it too high. There could be a thermal change in that parameter in the unit under review.
So for a 10kHz signal to the speakers, the amp sends out 510 kHz?
No. There is no correlation to the switching frequency and input signal. If you took x number of any brand of "self-oscillating" amps, put them along side, they would all be different. The ones that I use will vary between 480 kHz to 510 kHz or so at idle.

The old style ones.......the ones that gave "digital" amps a bad name......used a fixed frequency oscillator. It remained constant.
does anyone happen to know, if the outlaw model 200, mono block is a digital switching amp, like what is being discussed here ?
What would it - high freq hash above 20k - sound like over speakers?

Bob Wood
http://www.GreatHomeTheater.com
Thanks Ar_t.
Bob, I don't think there is any sound you can hear over 20k but purely subjectively, it gives me a headache. Just don't ask me to DBT my experience.
Personally I dislike the entire working principle of digital amps
(aka class D,PWM).
First of all the introduction of a convertor into the signal chain
(converts the analog signal into a pulse-width modulated one),
then the fact that the transistors only ever are fully off or fully on
and in the full on only transmit white noise.
That the re-conversion relies on the inertia of the connected drivers.
The high-power fast switching limits choice of transistors to VFets.
Radio interference created by the switching and frequency,harmonic and intermodulation distortions created by the modulator (convertor).

The only reason for their existence is that they are very light and cheap
compared to regular amps which makes them ideal for touring pa application.
They are just about good enough for the job.

By the way switching frequency and amplitude is fixed in all these amps
as their functioning depends on it.
golix@macunlimited.net...The switching frequency is not fixed in Tripath-based amps, and some others. Your understanding of digital amps is dated in other respects too. As to their advantages, light weight is not important for home audio, but low cost certainly is.
I agree. I believe he is assuming that this genre of amps has not improved any in the last 20 years.

It has.......I wouldn't waste my time with them if they hadn't.

As for "conversion".........some types do. The PCM/PWM ones do. Mostly, they just modulate the power supply, which is all that any amp does to begin with. It is just a different form of modulation.

I do not understand how they "only transmit white noise". Waiting for an explanation on that one. RFI, yes they do produce that. Some of us go to great lengths to eliminate it.

As for "cheap". Yes, some are. About the only real cost eliminated is the heat sink. You can get by with a smaller power transformer, as you don't have to heat the atmosphere at idle. Other than that, they are not that much cheaper to build.

Well, the goods ones aren't........
To follow up on the conversion issue. What requires more conversion, digital to analog or PCM to PWM? Like it or not, digital is probably the most popular source.
Unsound...Not all digital amps accept digital input...yet.
How do the Channel Island amps differ?
Eldartford, yes I'm aware of that. In fact it seems as though very few do. My question was earnst. Which process involves the greatest amount of conversion? Perhaps a better question, might be which conversion is the least detrimental?
Just wanted to check in on the SOUND of these amps, as I currently have the Channel Islands 100 wpc monoblocks in my wife's system. No, it's not a permanent thing, just letting a friend demo his stuff at my home.

First, the obvious. Yes, they do run cool. Despite leaving them on 24/7, I presume they are impacting my electric bill in quite a benign way. Build quality looks better than good, they are quite solid. For those who care, while certainly not glamorous, they are not unattractive. I'll say my wife has not even noticed them; that can appeal to many of us.

In terms of the stuff most of us here actually care about, I find them to be fine sounding amplifiers. They seem to be capable of delivering the kind of power a demanding loudspeaker, which is currently a part of the setup, require. These speakers look to be a tough impedance load, one my Jadis DA30 was certainly not up to the task of handling - the result of which was really dead sound. With the Channel Islands monos, things really sprung to life.

Like most solid state, in my opinion - which is the polar opposite of the conventional wisdom, they present a rich, robust, and relaxed sound. They have no issue in sounding stout in the bass - where I do agree with most in terms of solid state. The amps are dead quiet, but I should add that seems to be a gift of the new room I'm in, as everything I've tried in there has also been dead quiet - do I dare tempt fate by installing dedicated lines??? Clarity and channel separation also seem fine. I don't hear them really doing anything wrong, and that in itself makes them worthy of an audition by anyone interested.

The only area of complaint I could level at them is the lack of the sweetness of my Jadis, but I could aim that at most amplifiers, solid state specifically. I can't disagree that a good tube amp brings out the emotion in music.

In the end, at $1600 for a pair, they certainly win my recommendation in spades. Were I in the market for this type of amplifier, I'd buy them, though it's probably apparent that I like them more with my head than my heart.
Unsound...The answer to your rhetorical question is that any digital-to-digital conversion can be done EXACTLY within the resolution of the digital data. Even Fractional bits, resulting from the conversion, can be represented by oversampling.
Positive Feedback just released their review of the CI DM 200s. This may answer part of the orginal thread. IMO, the review is accurate.
Thanks for the head's up on the review, Cellorover!

I will agree with most of the review. However, the area I think the review got completely wrong is that sonics of these amps to me is anything but tubelike. Again, for me, what I hear from a tube amp goes against the typical descriptions of such. These amps are as solid state sounding as they come. And, again, they certainly have garnered my recommedation - breaking out Kraftwerk last night showed the music up in excellent fashion.
((((Personally I dislike the entire working principle of digital amps
(aka class D,PWM).
First of all the introduction of a convertor into the signal chain
(converts the analog signal into a pulse-width modulated one),
then the fact that the transistors only ever are fully off or fully on
and in the full on only transmit white noise.
That the re-conversion relies on the inertia of the connected drivers.
The high-power fast switching limits choice of transistors to VFets.
Radio interference created by the switching and frequency,harmonic and intermodulation distortions created by the modulator (convertor).)))
I wounder if this is the reason I dont listen nearly as mutch while I have the ARC M-150 in my systum as win I have the VTL MB-175 mono's in. But I swaer the ARC dose everthing better.
Eldartford, you give me too much credit. While that is what I suspected, I really didn't know for sure.
Thats 150-M.
I own a Bel Canto EVO2-the newest gen2 model-
had it about a year now and all i can say it is
it is very very good.To get the most out of it You need a really good Source and Pre-Amp.I have a Audio Research cd3mkll Cd player and Audio Research PreAmp-Ls16mkll
and with good MIT cable both to speakers and as interconnect in a balanced configuration and Vandersteen 3a Sinature Speaker-the sound is magnificent-some of est I have heard anywhere.
How can someone say that digital amps(actually, some of the Japanese ones I heard in Best Buys gave me a headache also-but I'm thinking Nuforce here{yes, I know they really aren't digital amps}give them a headache, but not notice that cd's give one a headache?
I know with me some do:
"When I had the PS Audio, it gave me headaches like when I listen to CD's from the Cranberries"