Buzz about pure class A solid state amplifiers ?


Fellow Audiogoners:

I would greatly appreciate all your input/insight.

Lately both in and out of Audiogon, I have been aware of a "buzz" and numerous praises about the benefits of pure class A solid state amplifiers, as an alternative to the more common high bias AB designs, and an alternative to the transparency and coherence of tube designs.

I have recently auditioned a current production 60 watt Accuphase class A solid state design, a fine amplifier indeed. However, up against my similarly priced VAC Phi 300 class AB stereo tube amplifier (used in ultralinear), it sounded mildly anemic, less transparent, and less coherent/detailed. This is with due respect to Accuphase products, which I have owned and enjoyed greatly.

So....please help me to understand all the purported benefits of the lack of crossover notch distortion, etc with class A push pull soid state designs, when excellent AB tube amps, with crossover notches, are IMHO more cohesive, transparent and resolving in comparison. Have designers, both tube and solid state, basically succeeded in rejoining positive and negative waveforms seemlessly, without perceptable crossover notch distortion?

Yes I did, admittedly, have an interest in simplifying my system, and going back to solid state after 14 years (with tube preamp); However my tubes are gonna stay :)

To all my solid state friends, please know that I have owned many solid state amps which I have truly enjoyed and have the utmost respect, dating back to the Reference Levinson 32's...and I know there are great SS amps out there, and primarily class AB designs.

Just wanted to share my experience, and learn alittle more about class A solid state, specifically.

With thanks in advance for your thoughts

Brian
audiobrian
Post removed 
Viridian:

Please allow me to clarify...the "buzz" meaning lots of interest, not an audible buzz :)

Thanks
Brian
I've owned some pretty good A/B's, Class A's, and Tubes. I'm a tube enthusiast so take this for what its worth.

First, I think that the best of Class A's and A/B's sound much more like each other than either sound like tubes. Not even close IMHO. That said, I think Class A's 'can' sound smoother because of the absence of switching distortion. CAN sound smoother, but they don't necessarily sound better depending on the sophistication of your other components and the resolution qualities of your speakers, or tonally more correct, including warmth/spaciousness.

About the only SS amps that I have heard/owned that I would use (if I had to) would be old Levinson ML2's or for something a tad more modern and more powerful Threshold SA1'. But thats just me. :-)
At elevated levels it is likely that both sound nearly indistinguishable.

At low levels Class A is likely to have an advantage.

Crossover distortion is not that "level" dependent. Therefore when the music is very soft then the crossover distortion may become noticeable. As you increase the music in relation to the distortion then it becomes less audible - much in the same way "hiss" when playing loud becomes a non issue but can be a nuisance in very quiet passages in classical.

Of course an extremely precise Class A/B design that has negligible crossover distortion may be just as good as a Class A.

Of course components drift with age - so a Class A amp that is five years old may outperform a Class A/B of the same age ( unless you has the Class A/B serviced). On the flip side the higher operating temperature of a Class A may also mean that it ages faster.

There exist power amps that run Class A to significant power levels before switching to Class A/B. This may be the best compromise (power but also years of reliable quality signal at low levels) depending on your viewpoint. Another user could argue that it is simpler just to service the Class A/B amp regularly to keep crossover distortion low.

Note that designers can deliberately choose parts so that they partially compensate eachother in the component aging process...this is probably what distinguishes high quality designs from cheaper low quality ones.

There is no absolute answer - except that build quality (design, component selection) may ultimately be more important than Class.
Post removed 
You did an A/B comparison between class A and class A/B except that A and B were apples and oranges (bananas, if you prefer). Not to mention the interaction with the rest of the system and personal preferences.

The only true test would be on the same amp that could do both, like the Plinius SA series that can be switched.
the "buzz" about class A amps usually results in popular discussions about crossover distortion...it's easier to imagine this voltage based boundary condition than, say thermal distortion mechanisms manifesting non-linearites which may actually do more sonic harm than crossover distortions. The class A approach minimizes the thermal distortions by maintaining a more or less constant device temperature, and therby reducing the distortions caused by heating and cooling (consider a class B amp: junction temperatures and the gain non-linearities rise and fall much faster than the massive heat sinks they are strapped to, which are more an RMS type thing).
Other positives about class A involve less modulation based noise on the power rails (since the output devices are constantly "on" the rails are subject to a more fluid load), and of course the often mentioned cross-over distortion. As Ngjockey states and as I can attest to, the primary differences when running the sa102 in class AB/ vs A are:
When thermally stablized (after about 45 minutes-- and btw, more indication that the benefits of class A are not due entirely to absent crossover distortion) in class A
1)image depth increases noticably
2)vocalists take on a rounder more coherent tone
3)fine treble shadings becoms more apparent
4)overall 'feel' seems more natural and grainless

no big changes in bass power or dynamic "slam"
just more refinement overall...its like you have two amps in one...and its fun to listen for a while in one mode then another and hear this over and over...

hope this helps!


The class A approach minimizes the thermal distortions by maintaining a more or less constant device temperature, and therby reducing the distortions caused by heating and cooling

Hang on - a pure Class A device will cool when driven hard - it will run hottest at idle. So you still have thermal issues - like with any device.
Perhaps your preference for the class AB tube amp, might invite comparison with a pure class A tube amp?
Class A is the least efficent, generate the most heat, and cost the most to repair all those fried transistors. Check the price to repair a pair of Krell MDA300's...about 1500-2500.00 Not cheap, I find it easier to just change tubes. However, can't drive MBL's with tubes very loud so SS it is. Stick with the companies who know how to make Class A, AB, and D sound great. ARC, Pass Labs, Rowland, Krell, Copland, Ayre, and the others. The most important thing is not whether it is Class A or not, it is how the damping factor of the amp is matched to the needs of the speaker. If this is off, all A-D will sound like bad hifi.
The advantage of a Class "A" amplifier is not:

"The class A approach minimizes the thermal distortions by maintaining a more or less constant device temperature, and therby reducing the distortions caused by heating and cooling (consider a class B amp: junction temperatures and the gain non-linearities rise and fall much faster than the massive heat sinks they are strapped to, which are more an RMS type thing).

Other positives about class A involve less modulation based noise on the power rails (since the output devices are constantly "on" the rails are subject to a more fluid load)"

as Dpac886 suggests. It is that the amplifying devices are constantly operated in their linear region, above cutoff and below saturation. By using Class "A" topology we can set the bias of the device to it's most linear region. This negates large amounts of negative feedback needed to correct for the extremely nonlinear characteristics of the transistors at the collector cut-off region and the turn-on/turn-off times of the devices in class A/B topology. There is a good explaination at Passdiy.com, Nelson Pass's contribution to the DIY Community.

http://www.passdiy.com/pdf/classa_amp.pdf
I truly appreciate the ongoing discussion. Another point of confusion to me is comparing let's say a Belles MB-01 class A and Karan 450 class A both using approx. 1000 VA transformers/channel, similar capacitor banks, and 20 transistors/channel. However the Belles is said to deliver 75-80 watts pure class A into 8, and the Karan, 450 watts pure class A into 8 ohms.......this huge difference in pure class A output power is not easily understood, even though Mr. Belles and Mr. Karan obviously use different toplogies, transistors, etc.

For what it's worth, in the context of my system only, my personal favorite solid state amplifier, which I do own, and alternate with the VAC, is the class AB Gamut M250 monoblocks. As many know, these monoblocks use one large mosfet pair/channel. I'm not sure if using a single mosfet pair/channel is theoretically better than paralleled devices, as Gamut and, I believe, Vitus believe....another interesting topic, perhaps for a different thread :)

My system comprises a VAC Ren mkII pre and either the Phi 300/Gamut 250 into Kharma Midi Grand Ceramiques :)

Thanks
Well some will call their amps Class A, but obviously it's just a marketing tool for some. As you say, "buzz" word.
All 'Class A amps are not equal, just as all Class A/B amps are not equal, and all tube amps are not equal.

FWIW, I have not heard the Karan KA-S 450, but just looking at photos I find it VERY hard to believe that it could be pure Class A. 450 wpc in pure Class A would require monoblocks with much larger heatsinks.

As a comparison, Pass Labs XA series, which runs in pure Class A, goes to monoblocks with their 60 wpc XA-60.5
Their top of the line XA-200.5 puts out 200 wpc, and each monoblock is larger than the stereo Karan 450. So I would be inclined to believe that the Karan slides into Class A/B at around 50 wpc. This is no ding on Karan, as I have not heard this amp. I'm simply saying that physics makes it impossible for me to believe that this amp delivers 450 wpc of Class A power.

It's just physics, Class A generates a LOT of heat. If it truely is Class A, the power rating will be about 20-25% of a similar size amp which slides into A/B. Many amps will give you 5-10 wpc of Class A before sliding into Class A/B, and some may be bold enough to claim that they are Class A amps, as it is a selling feature, and marketing folks know what buzz words attract customers.

I also agree with Ngjockey, that the best way to compare A to A/B is in the same amp. AFAIK, Plinius and Clayton are the only two manufacturers who offer an A>A/B switch that you can flick on the fly. I've owned both, and experimented. I would say the Class A sounded better, more relaxed and articulate especially at low to moderate listening levels. At higher listening levels the differences decreased, IMHO.

Cheers,
John
Post removed 
Post removed 
perhaps in some designs, but you would have to be running the amp at 100% rms power levels for that to occur, (and the heatsinks would not cool appreciably) most speakers and ears would have been damaged. This argument holds no water. [/url]

Good reply - I do see your point now. I agree that at more modest levels the Class A is going to be thermally much more stable (run hotter but stay that way).

[quote] BTW what amp drives your ATCs?

These are active speakers so each speaker has three separate Class AB amps that run Class A to two-thirds power. The amp circuits are completely discrete (all individual components - no packages) and are actually very simple. They run rather hot when idle of course. The design uses a sliding bias so that the operation stays Class A until 2/3 power. The power output stages of the amp look like any typical simple amp design with two complementary sets of push pull high power MOSFETS - but remember each of the three separate amps are only driving one transducer (no complex crossover) over a very limited frequency range (so the work load is shared and kept as simple as possible - this is perhaps the biggest benefit of the overall design rather than the Class A operation). There is actually more complexity in the line level buffering/pre-filtering for each power amp stage which includes some phase adjustment to maintain accurate phase over the crossover. The three power amps to each driver are totally identical - so power output must be controlled/limited by the load resistance (woofer taking more power than a tweeter).
JMCGrogan2 Vitus SS101 and SM101 have the same feature A/B-A switching as you could read in Ultraaudio review.
Thanks Teddy_bear, I had not heard of Vitus before. However, after looking into it that looks like one SERIOUS amp!!! 176 pounds puts out 50 wpc in Class A!!! $28K price!! Gulp....that makes the Pass Labs stuff look underbuilt.

My back and my wallet scream at the thought! :)

Cheers,
John
Jmcgrogan2 they have a small brother SS010 you could try that :) Power rating for the SS101 is funny though :D