[Brian Cheney]

If you have specs for a recommended filter please supply them.

Both channels exhibited the exact same behavior, a burst of oscillation at clipping followed by a flattening of the output waveform.  With a 4 Ohm load this behavior occurred at 30V RMS output.  With an 8 Ohm load this occurred at 40V RMS.

I am not involved and have no interest in the sale or marketing of the Carver Pro amps.  I am trying to find out if I can in good conscience recommend them to our customers as a high end product.  

Mr. Curl has the AP filter you recommend but did not feel it would change the power output measurement substantially.

It appears the two output channels are in inverse polarity relative to one another which would lead me to think there might be noise cancellation when the amp is operated in bridged mono.  Is this the case?

[audiojerry]

I, for one, like the fact that Brian took the initiative to attempt independent testing, and that Carver saw fit to respond. Maybe this will result in even better information.

This is another example of why this forum is so great. If we can keep our diagreements civil and maintain respect for one another, the Audiocircle will continue to blossum.

It would be great to get viewpoints from some of our amp designers like Curt, Frank, Hugh, and DVV.

Do any of you guys have any thoughts, concerns, fears about this technology apparently coming of age?

[rosconey]

i agree this has been  very civil and informative  
i wish all topics could be discussed with this type of class

[Brian Cheney]

The discusson on technical measurements may be moot.  Preliminary listening on my big system is very favorable.  More later.

[audiojerry]

The discusson on technical measurements may be moot.  Preliminary listening on my big system is very favorable.  More later.

Whoa, such a statement might be regarded as blasphemy by devout followers of the Church of Measurement! If something doesn't measure well, it cannot possibly sound good, and even if it did, it would only be due to the devil playing tricks with your ears. If you believe your ears instead of the measurements, that is the devil talking. Why do you think tube amps glow red and run hot? Instruments of the Devil!

[Curt]

It would be great to get viewpoints from some of our amp designers like Curt, Frank, Hugh, and DVV.

Do any of you guys have any thoughts, concerns, fears about this technology apparently coming of age?

IMO switching amplifiers are not ready for the audiophile market. They get our attention as an exciting new technology but I think when all is said and done most of us will still be using analog amplifiers in our two channel systems from some time to come. Switchers, today, are just not good enough to compete with the best analog amplifier designs.

A good analog amplifier has a noise floor ~20dB lower than a switcher and has about 10X less THD, those are two important issues. Remember, an audiophile is the person who wants to squeeze every last drop of performance out of his gear, the person who argues over power cords and brands of capacitors... performs tweaks to mods...

When the excitement of the newness wears off and the shoot outs are over I think we will all have had some fun but still be listening to our analog amps until switchers go through a few more generations.

After following some of the new switching amps for a couple years what I like most about them is that they are cheap, energy efficient and don't get hot (no huge heat sinks). The Apogee chips are driven by a direct digital stream into a processor IC then the switching amp, no D/A is required coming from a CDP, that's pretty cool, direct digital audio.

But, here are some of the things I don't like: the high parts count (can break easier); higher noise floor; higher THD; the fact that they are great radio transmitters and can pollute the AC mains; they can cost more to repair; we loose control over the design (it's inside an IC), and they may not be as tweak able as an audiophile likes gear to be.

I see switchers being used in HT systems (5 amps in one box w/ smallish power supply), autos, motor homes, boats, whole house distributed audio systems, computers, sub woofers, and mid-fi systems, all with good success. But not high-end audiophile two channel systems, not yet anyway.

Of course this is just my opinion, an engineer's not a subjectivists, and perhaps I'm wrong but... I'm not ready to “switch” just yet. Maybe in the future when a few of the negatives have been improved.

[cyounkman]

The discusson on technical measurements may be moot.  Preliminary listening on my big system is very favorable.  More later.

Brian,

I'd like to compliment you on the professionalism with which you've handled this whole thing. Needless to say, arranging to do a full bevy of tests is no simple task, and I for one appreciate that fact that you've shared your findings here; results be damned.

Whether it's a testing methodology problem or simply another baffling case of an amp that's terrible on paper and great in the listening room, I appreciate the time you've spent; and, particularly, your forbearance with those who have read too much into your posts...

I'm also very eager to hear the results of your listening tests.  

Is it improper of me to ask what amps you have on hand to compare?

[Jack]

IMO switching amplifiers are not ready for the audiophile market. They get our attention as an exciting new technology but I think when all is said and done most of us will still be using analog amplifiers in our two channel systems from some time to come. Switchers, today, are just not good enough to compete with the best analog amplifier designs.

A good analog amplifier has a noise floor ~20dB lower than a switcher and has about 10X less THD, those are two important issues. Remember, an audiophile is t ...

Try it, you'll like it.
Specifics will be on our website. I will post the link when ready.
Sincerely,
Your friends at Carver Pro

[warnerwh]

Curt: the extra information regarding the digital amplifier technology is appreciated. As engineering goes it seems to me that it usually takes more than a few years to work out all of the problems.  An RF transmitter is not needed in my system.  The sound though may trump the drawbacks.  I'm surprised that if the noise floor is 20db higher on these amps that it would be tolerable. Is this correct?

[Brian Cheney]

On hand for comparison with the Carver are a Plinius 250IV, two pair of Ampzilla 2000 monoblocks, a 10 year old Parasound 2200, and the QSC PL236, a pro amp with a switching power supply but a class H (linear) output stage.

I agree with much of what Mr. Wishman says.  The Carver is not designed around audiophile lines.  Since the adapters I tried detracted from the sound I ended up jamming half a spade lug into each of the insulated "binding posts" (BP's in appearance only, they accept only bananas or bare wire, and only from underneath the amp which means the unit must overhang the back of any equipment rack).

Most analog designers I know are put off by the extremely high noise levels of a switching amp, even if they are supersonic.  If your class A amp oscillated at full power at 500kHz and every waveform it produced was overlayed with hypersonic garbage, you would be offended.  Yes a filter can be added to the measurement chain but not to the listening chain.  The Audio Precision distortion meter must "null" the fundamental in order to display residual distortion.  This "null" was hard to come by in our tests, since the switching frequency modulates the signal, causing a constant sideways jump (in frequency) in both directions.  This is most disconcerting.  You have to examine the output on a spectrum analyzer like John Curl's HP 3563 in order to separate true harmonic distortion from the all-pervasive noise.  Fourteen harmonics of 1 kHz at high levels are also very disconcerting to see, as is the high IM.

[Curt]

Curt: the extra information regarding the digital amplifier technology is appreciated. As engineering goes it seems to me that it usually takes more than a few years to work out all of the problems.  An RF transmitter is not needed in my system.  The sound though may trump the drawbacks.  I'm surprised that if the noise floor is 20db higher on these amps that it would be tolerable. Is this correct?

A switchers noise floor is way below any listening rooms noise level and the SNR of the switchers (as stated by mfg) is pretty good so, yes it is quite tolerable and not a real damaging issue, just something I don't care for.

Two issues I don't like most are the broadband noise and the higher THD.

[Curt]

Try it, you'll like it.
Specifics will be on our website. I will post the link when ready.
Sincerely,
Your friends at Carver Pro

Thanks Jack, I'll watch for your link.

[Jay S]

Here are the manufacturer's specs for the Acoustic Reality eAR Two ICEpower amp:

Max. power output (RMS) 2 x 500W / 4Ω, 2 x 250W / 8Ω
Power stage efficiency > 93%, 200W / 8Ω  (only ICEpower)
Peak output current >80A (peak power into 0.6 ohm = 5500 Watt)
Output impedance 2mΩ @ 1kHz
Dynamic range (A-weighted) >117dB
THD+N < 0.009% 100W @ 8 ohms
Linearity 10Hz - 20kHz: +- 0,2dB
Bandwidth 5Hz - 70kHz
Input sensitivity (full output) 2V RMS
Dimensions (WxHxD) 380x130x280 mm

Of interest is the claim that distortion is so much lower than what's been recorded for other digital amps.  

The measurements from this review seem to support the low claimed distortion.
http://www.acoustic-reality.com/articles/eng_highfidelityreview.pdf

As for noise emitted by digital amps, I can tell you that my system's sound quality improved very significantly when I plugged my digital amp into my balanced power supply.  Now, whether that is due to a reduction of noise going into or out of the amp....

[_scotty_]

Here are the manufacturer's specs for the Acoustic Reality eAR Two ICEpower amp:

Max. power output (RMS) 2 x 500W / 4&#937;, 2 x 250W / 8&#937;
Power stage efficiency > 93%, 200W / 8&#937;  (only ICEpower)
Peak output current >80A (peak power into 0.6 ohm = 5500 Watt)
Output impedance 2m&#937; @ 1kHz
Dynamic range (A-weighted) >117dB
THD+N < 0.009% 100W @ 8 ohms
Linearity 10Hz - 20kHz: +- 0,2dB
Bandwidth 5Hz - 70kHz
Input sensitivity (full output) 2V RMS ...[/quot      
 
These measurements reflect the use of AES 17 filter which rejects noise above 20kHz at something like 60db/oct. If other digital amps were measured using this method similar figures might be generated.

[Curt]

It's nice just to filter out the bad and improve your THD+N spec. Too bad most users don't listen with one of those filters on their amp.

It's true some of the high freq garbage can effect the lower frequencies and the AC power, it should be removed for normal listening too, not just for the spec mesurments.

You would think an amp should be measured just the way its to be used.

[warnerwh]

I recently read an article by a professor of psychoacoustics. He's one of the "if you can't measure it you can't hear it" people.  His testing indicates that some individuals can indeed detect frequencies well above the 20khz point.  Wish I could find the article but if memory serves me correct he stated they'd tested as high as 35-40 khz.

[kana813]

Warnerwh- if you enjoing reading what college professors think about
HiEnd audio, check this one out:


http://signal.ece.utexas.edu/seminars/dsp_seminars/01fall/AudioMyths.p
df

[_scotty_]

It's nice just to filter out the bad and improve your THD+N spec. Too bad most users don't listen with one of those filters on their amp.

It's true some of the high freq garbage can effect the lower frequencies and the AC power, it should be removed for normal listening too, not just for the spec mesurments.

You would think an amp should be measured just the way its to be used.

Curt, This is one of those times, that if it sounds good, it is good. Not investigating and listening to this new technology because of objections to what amounts to theoretical drawbacks due to noise issues is probably
going to have you inquiring in the not so distant future about the the number of the truck that hit you.

[DVV]

I, for one, like the fact that Brian took the initiative to attempt independent testing, and that Carver saw fit to respond. Maybe this will result in even better information.

This is another example of why this forum is so great. If we can keep our diagreements civil and maintain respect for one another, the Audiocircle will continue to blossum.

It would be great to get viewpoints from some of our amp designers like Curt, Frank, Hugh, and DVV.

Do any of you guys have any thoughts, concerns, fears about this technology apparently coming of age?

Agreed on the maturity of argument, Jerry, right on! A forum will by default have different views, and that's as it should be, provided we can keep it civil.

Regarding digital amplification, in my mind there's no doubt whatsoever that this is the technology of the future. It makes sense, since our sources are rapidly turning digital, to keep the signal in the same form for as long as we can, preferably all the way. I never liked converters of any kind in between, anywhere, not if I could help it.

As of this writing, my experience with them is limited, so I can't say anything conclusive from my point of view.  What I have heard is not what I'd call breathtaking, however, we must be realistic and note that this is a very young technology, still in its infancy. Look at CD - we are now 20 years down the road, and still have so much to explain and learn about it. See how that has evolved - far from perfect, but also far from its initial days, when effective 14-bit resolution was cause for an applause.

On the other hand, more intuitively then anyhow else, I feel this is a technology showing tremendous potentials and promise. At this time, it merely approximates high quality analog amplifications, but it will continue to evolve and at some point, it will catch up, and then overtake traditional analog technology. How soon, or when, I can't even guess, too many variables, but I'd say 10 years from now we will be auditioning some very serious samples.

Until then, I'll continue with traditional analog designs, but not for a moment will I allow the digital scene to get out of my eyesight.

Cheers,
DVV

[Curt]

I recently read an article by a professor of psychoacoustics. He's one of the "if you can't measure it you can't hear it" people.  His testing indicates that some individuals can indeed detect frequencies well above the 20khz point.  Wish I could find the article but if memory serves me correct he stated they'd tested as high as 35-40 khz.

But, we can measure it, JC did measure it and there was some stuff, a bunch, it the audible range. I believe that's what Brian said in his post.

I not sure I agree with if you can't measure it you can't hear it either What about the stuff we don't have measurement techniques for, like capacitor dielectric. It's hard to measure the difference between caps but I can hear differences, can't you?