New class-T amplifier?

[Undertoad]

http://www.amazon.com/exec/obidos/tg...=UTF8&v=glance

This little 15W amp goes for $30, meant to go between your mp3 player and your home speakers.

http://www.forbes.com/global/2005/1003/026A.html

Audiophiles proclaim it the Next Big Thing, comparing it to multi-$thousand amps.

It wasn't even the company's intent to build something really hi-fi, but they appear to be the first to use this particular type of amplification described in the Forbes article. The critical part used is cheap to produce. I don't exactly understand the electronics behind amplification... I understand what I need to know, but not how chips actually amplify. Audio engineers must be going crazy right about now.

[xoxoxoBruce]

From the second link;

The T-Amp uses Tripath's lowest-end chip, the model 2024, which puts out 15 watts of power and costs $3; its most powerful chip puts out 500 watts, costs $45

New models are coming out, including one with the same chip as the plastic T-Amp but boasting a metal case and heavy-duty speaker connectors for $139, followed next year by a 50-watt version that will cost $349.

Having them built in China helps keep the costs low.

[Silent]

I think they'll find building a clean low power amp on the cheap is not too hard.

Building a clean high power amp, much harder. You need more the just an amplifier IC, your power supply also needs be clean and capable of handling large surge current. That usually requires expensive transformers and capacitors. Also, at high power outputs, cross talk and feedback become issues, so sheilding and a carefull pcb layout also come into effect. Again, not cheap.

So, after all that, you may find that the peformance of the more powerful amps built on the cheap un-do everything gained by the new amplifier technology.

Only one way to find out though.

[wst3]

I don't know, but I sense some snake oil here! I read their white paper, and they make some interesting claims, and present some interesting measurements.

I could pick nits all day, but when people start claiming to measure total harmonic distortion (THD) with an excitation signal greater than 10 kHz I hear alarm bells. If the device or the test set is band limited, as the graphs indicate, then it is meaningless to measure THD with an excitation signal higher than 10 kHz since the first harmonic that you can measure is 20 kHz. If one is making a true THD measurement, the figure necesarilly decreases as excitation frequency increases. If, on the other hand they are making a THD+N or SINAD test (two common alterntives to THD) then it becomes just plain meaningless, since the noise will be much greater than the harmonics generated.

That's one example<G>...

I wish they'd provide a little more detail about just what "Class T" amplification is, I am curious. Like many I've built Class A, Class B, Class A/B, and Class D amplifiers, and even with common components (OK, can't really do that with Class D), they all sound different.

From their own information it is still a switching amplifier, although to be fair PWM may not be the best approach to switching amplifiers, and mixed mode technology has grown tremendously, to the point where the dividing line becomes blurry<G>!

I'd like to hear (and even measure) one of these things. I find it possible that the little amplifier may sound great. I agree with Silent that they may discover some challenges as they ramp up the power handling capability, both for the chip and the final product.

A also find some of the claims in the Forbes article to be a bit odd. But then there are audiophiles who spend $500 on a power cord and then claim to hear the difference.

Interesting stuff though...

[Elspode]

I, too, am pretty much unimpressed with the low-power aspect of these devices. Might make a nice clean preamp for a higher power system, though...?

[tw]

Originally Posted by wst3

A also find some of the claims in the Forbes article to be a bit odd. But then there are audiophiles who spend $500 on a power cord and then claim to hear the difference.

These are the same people who recommend a plug-in UPS to 'clean' power, or spend $100 on Monster Cable wires because their gold connectors make better sound. Or will mark their speaker cables because those wires have polarity - one end connected to the speaker instead of the amp causes worse sound. I also smell the Wicked Witch of the West.

The original amp was Class A. Class B intead amplifies one half of a sine wave. Class C only amplifies sine wave's peak. Class AB is two class B amps where one amplifies top half of a sine wave; other amplifies the bottom half.

Class D does same thing done in CD-Roms, telephone exchanges, modems, and switching power supplies. Class D is a digital amp.

Analog (as in Class A, B, and A/b) means more energy loss. For example, a transistor that is completely off consumes virtually no power. A transistor that is fully on also consumes virtually no power. But a transistor that is only half on consumes half the power; the other half consumed by a load. A digital amp eliminates power loss by operating only fully on or fully off. Switching power supplies are so much more efficient than their old analog equivalents by doing same - fully on or fully off.

Little useful is described in that white paper or article. A digital recording at 32Khz does not have all this distortion. Why then would a digital amplifier at 100 Khz have so much distortion? Well, the devil is in deatils not even implied in those articles. In theory, a Class D amp means perfect amplification - no distortion. And then we apply reality to the concept - those devilish details.

Previously noted was Class AB amps. IOW one class B amp must turn off exactly when the other class B amp (to amplify the other half of a sine wave) turns on. Any overlap means distortion. Still class AB amps have long been the most common, low distortion amps. Class D amps must be designed with the same care. So suddenly only one company finally did design with care?

Again, appreciate why so many audiophiles recommend Monster Cable products. Recommendation not based on science or the numbers. Too often, audiophiles advocate the latest product hyped by Rush Limbaugh gurus. Suddenly someone recommended this T-amp. Suddenly everyone is hyping it as if it were as superior as vaccum tube amps. Nonsense. Maybe there is something uniquely superior to this T-amp. But the history of the source has a credibility problem.

Appreciate that Class D amps are the future. They will get better for the same reasons why switching power supplies output more power in a smaller package. That reason is even touched upon in the cited article / white paper. There is no magic here. Just another example of an amp that has long existed and is getting better over so many years. A technology to be understood by the science - and not by audiophile hype.

[wst3]

Originally Posted by tw

These are the same people who recommend a plug-in UPS to 'clean' power, or spend $100 on Monster Cable wires because their gold connectors make better sound. Or will mark their speaker cables because those wires have polarity - one end connected to the speaker instead of the amp causes worse sound. I also smell the Wicked Witch of the West.

Yup... I think my favorite, after the expensive power cords, is the directional cables. Pretty amazing stuff!!!

Originally Posted by tw

The original amp was Class A. Class B intead amplifies one half of a sine wave. Class C only amplifies sine wave's peak. Class AB is two class B amps where one amplifies top half of a sine wave; other amplifies the bottom half.

One quibble - Class A amplifiers are biased such that they are always on, which is why they are such pigs, and why they sound so good! They amplify the entire waveform.

Class B amplifiers are not biased so that they can only amplify one half (positive or negative) of a sine wave. They can be used as half a "push-pull" stage, or in lots of other applications.

Class AB amplifiers are biased so that they amplify just over half that sine wave, they are more efficient than Class A.

I believe (college was a LONG time ago) the thing you describe as Class AB is actually the push-pull topology. Two class B amplifiers biased so that each one can amplify one half of the sine wave. As you described, the trick is getting the point where one turns off and the other turns on to match exactly. If they don't you get rather nasty sounding cross-over distortion!

<snip>

Originally Posted by tw

Little useful is described in that white paper or article. A digital recording at 32Khz does not have all this distortion. Why then would a digital amplifier at 100 Khz have so much distortion? Well, the devil is in deatils not even implied in those articles. In theory, a Class D amp means perfect amplification - no distortion. And then we apply reality to the concept - those devilish details.

Yeah, those details! What's the difference between theory and practice???
<more snipped>

Bill

[smoothmoniker]

Originally Posted by wst3

Yup... I think my favorite, after the expensive power cords, is the directional cables. Pretty amazing stuff!!!

I was MD for an artist a few whiles ago, and we were doing auditions for guitar players to put on the road. A guy came in who obviously liked the herb, and liked it enough to get loaded up for an 11 AM audition.

The techs roll his rig in, he gets set up, plugs in, and the most god-awful buzz starts coming through his little amp. He, of course, starts bitching about the bad power in the rehearsal studio. We, of course, kindly point out that they are running very clean power, as evidenced by fact that nobody else is rattling the hell out of their speaker cones.

At this point, somebody notices that his guitar chord looks is wrapped from tip to tip in various pieces of electrical tape, that there is some serious fraying and kinking. Somebody kindly suggests that he might want to try another guitar cable, and hands him one.

"F* No! All of the cables made today are shit! My guitar cable is vintage analog; today all the cables are digital. This one is so much warmer."

And there you have it folks. If you wanna really rock, go get yourselves some "vintage analog" cables.

[dar512]

Originally Posted by tw

Or will mark their speaker cables because those wires have polarity - one end connected to the speaker instead of the amp causes worse sound.

Not one end or the other, but certainly having the left and right channels agree on polarity is important. If the wire for one of the channels has inverted polarity then one speaker will be pushing out when the other is sucking in. This causes the volume for center stage sounds to drop off and the performance to sound deadened.

But I go along with laughing at the Monster Cable folks. I buy the cheapest wire I can find that is 16ga or heavier.

[tw]

Originally Posted by wst3

I believe (college was a LONG time ago) the thing you describe as Class AB is actually the push-pull topology. Two class B amplifiers biased so that each one can amplify one half of the sine wave. As you described, the trick is getting the point where one turns off and the other turns on to match exactly. If they don't you get rather nasty sounding cross-over distortion!

I had always thought class AB and push-pull were one in the same. Two Class B amps operating together to perform just like a Class A. How do class AB differ from push-pull?

Sidebar - making one half of the push-pull amp turn off exactly as the other turns on means making the design for wide temperature changes. Semiconductors change appreciably with temperature - which is why transistors are also used as temperature measurement devices. These now so common amplifiers required careful component selection so that the point of crossover did not overlap or separate for all temperatures. Class D amps must do same to avoid same distortion problems even created by room temperature change.

[eiffelenator]

Vintage analog! Hah!
Are there any of these 500 wpc chip amps around for sale?