Class A Solid State Sound

Would someone kindly describe the differences in class A sound of Pass XA.8 series, Accuphase A-70/75 series, and Gryphon class A amplifiers. Does much or any of the differences relate to mosfet (Pass and Accuphase) or bipolar (Gryphon)  output devices?  Thank you!

I like talking tech but i don't think attributing a sound to a transistor is useful anymore. 

Pass tunes his amps for high second order distortion and his fans love it.

Thanks for the information about Pass XA, Erik. I also believe the entire design and voicing are what leads to sonic differences. However, even today, I find mosfet based output stages to generally present a smoother, potentially less objectionable sound...just my opinion.
I have not heard the amps you mention but I do generally seem to prefer amplifiers that use Bi-Polar output devices, such as Clayton's M300 monos.  One possibly meaningful comparison I have made was when I owned both the M300s and Lamm's M1.2 Reference amplifiers for an extended period of time (i.e., months).  I ended up keeping the Claytons and selling the Lamms but I cannot rule out that my preference may have been due to the difference in power between those amplifiers. 
MOSFETs most definitely contribute to the Pass sound. The second order distortion typical in his class A designs are due in part to the fact Pass doesn't bother matching N and P channel parts much, if at all. They also tend to be a bit more linear since they have better thermal characteristics. The circuits they're used in are a big factor too. FET circuits are usually simpler because they take very little current to drive unlike BJTs. But BJTs do things some people like so they use them. It depends on what your goals are. I personally don't mind BJTs for small signal work, but I'm pretty well convinced a FET power amp sounds best. 
Generally, there is a small difference between FET type devices and bipolar devices. Sometimes FET devices can have a colored texture to the sound - more of a warm and fuzzy type sonic signature. This is not always the case, but FET does lean this way. The warmth of Pass Labs (and possibly Accuphase) definitely leans this way. I would not say Pass is very warm, but it’s a nice controlled easy listening sound.

Bipolar is usually a lot cleaner sounding (without that warmth influence). Gryphon is an example of this. Sometimes the high bias Class A bipolar devices can present a very smoothed over and liquidy sound (this is also an extreme coloration, but it’s different than the FET warmth). The new Krell sliding bias Class A amps are like this. New Mark Levinson are also very liquidy sounding. There are many others, but I can’t remember them.
Your description is pretty much what I hear between MOSFET and bipolar or BJT output stages.  BAT solid state amplifiers are another that to me have that soft/warm/fuzzy sound.  The effect is less noticeable in the Lamm hybrid amplifiers I have owned, but still perceptible as a fuller, less distinct and slightly boomy bass presentation, at least to my ears.

High biased push/pull Class-A sound when all is correct, has an ease to it, almost like you need to turn it up.

Complementary output transistors like Bi-Polar’s generally will punch far more current than complementary Mosfets can.

If both are used in push/pull high biased Class-A like these amps you mentioned, you’ll find the Bi-Polar output amp able to drive far more taxing speaker loads, and therefore sound have a lot more punch when needed into those speakers.

Pass and Accuphase use Mosfets and a Gryphon uses Bi-Polar output transistors, on speakers like most all Wilsons ect ect, I would suggest watt for watt the Gryphon’s

Cheers George
Well... Except for the scads of MOSFET equipped Pass amps that will drive 1 and 2 ohms all day long without a burp. Even my dinky little F5 will swing +/- 10 amp into a 1 ohm load with it's 2 little MOSFETs. Then there's the F5T V1, my build with 2 more MOSFETs, measured swinging +/- 23 amp into a .1 ohm load. Of course the F5T V3 will throw upwards of 38 amp at a sub-ohm load without hiccup. 

Measurements seem to suggest MOSFET amps have no difficulty muscling around obscenely low ohm loads. 

BS Kosst, you can’t get complementary mosfets to do current like complimentary bi-polars can into low impedances! End of story!!

That’s why power amps  that are crowned  "amperage masters", are all bi-polar.

And "if" one owns say a pair of Wilson Alexia (0.9ohm EPDR), there you will see the advantage that bi-polars can give.
EG: Gryphon Anitllion Evo mono
175w Class-A into 8ohms
1400w into 1ohm
7500w peak into 1/4ohm
Sorry Kosst, there is no way complementary Mosfets can do these sort of figures.

Cheers George
I have a re-capped Krell KSA-200S and was wondering if it was bi-polar?

I also have a Krell KSA-250 (OLDER MODEL) & was wondering if that had bi-polar transistors (I think it does).

I like the sound of the old Krells, but some say that some of the newer amp designs are better. I need stability to 1 ohm (Martin Logan ESLs).
I had KMA 100s. Old Krells are all Bipolar output stages. Wonderful amps until one blew.
Accuphase class A amps are stable into one ohm load. 
 I’ve heard diverse opinions on “pure” (non-bridged) class A monoblock amplifiers (Such as Antileon Evo Monoblocks and others) vs bridged class A Monoblocks (like Clayton M300 bridged internally by the designer or Luxman A600 in BTL mode (bridge tied load) 
Many feel the non bridged Monoblocks are better sounding, when output devices from the two channels are summed in Parallel vs Bridging the two channels.  Thoughts or technical explanations?
I direct your attention to pages 17 and 18 where Nelson describes and shows the actual measurement of a bone stock F5, with it's 2 dinky MOSFETs, swinging +/- 10 amp into a 1 ohm load.
In this PDF I'd like to first direct your attention to page 7 where an F5 Turbo V1, merely an F5 with an additional pair of output MOSFETs, is described as swinging +/- 23 amp into a .1 ohm load at 1 KHz. 

Later on page 10 describes a V2 build swinging 38 amp into a .1 ohm load with the same 4 complimentary MOSFETs. It's rational to conclude that the V3 build with 8 MOSFETs would swing +/- 76 amp into the same load.

It looks pretty obvious that MOSFETs have no problem swinging glorious amounts of current through ridiculously low impedance loads. 

Kosst, I thought you were smarter than that, you are making your self look like you have no idea.

Everyone knows given the best implementation for both, the complimentary BJT’s (Bi-Polars) will do current and low impedances far better than complementary Mosfets can. End of story.!!

Cheers George

Oh. So I prove you wrong and all you've got are baseless claims. BJT's only tend to do more current per device in audio because they're large single devices or Darlingtons in a large package. Not sure what that has to do with anything since you can parallel FETs all day and 5 of them are still easier to drive than a single power BJT. But hey... Who cares if you can build FET amps that will basically drive a dead short in a linear fashion? Just because a MOSFET amp can drive 50 amp into a 1 ohm load doesn't prove it can push any current, i guess. Funny.... You're all about Pass and his pre-amps (a subject he openly confesses he doesn't mess around with much), but you don't seem to understand why the guy doesn't use BJT's for anything but cascodes. I'll enlighten you....


The power mosfets used in this project are International Rectifier IRF 130's, a 150-watt, 100 volt, 12-amp device in a TO-3 style case. They are part of a relatively new line of transistors using IR's proprietary hexfet process, which combines a particularly high voltage capability with very fast switching times and low saturation losses. Except for the steep price tags, this type of device has a number of advantages over regular bipolar transistors in audio power amplifier use.

First, their simple low current circuitry eliminates the need for driver transistors. Second, they are immune to the second breakdown phenomenon, which robs bipolar transistors of their power rating at higher voltages. This breakdown results from the positive temperature coefficient which encourages local current hogging within a area on the chip, so that at higher voltages one small part of the transistor tends to do most of the work, resulting in more probable failure. In a power mosfet the temperature coefficient is negative, and energy dissipates more evenly across the surface of the chip, allowing full power application at the highest rated voltage. This characteristic also avoids the bipolar design problem of thermal bias runaway, and eliminates costly compensations to maintain reliable operation over a range of conditions.

Third, because mosfets are majority carrier devices (bipolars are minority carriers) their intrinsic speed is much higher. Rise and fall times are about 150ns while similarly rugged bipolar devices have rise and fall times several factors larger." 
(Full text: )

Hmmm.... sounds like MOSFETs are better audio devices from the technical perspective. In fact, it sounds like they're more linear and reliable across their entire operating range. One might even go so far as to say that they sound and behave more like tubes as well. It certainly doesn't sound like BJT's are the universal current champs. 

You can waffle on all you like with your war and peace post, in our hifi world your wrong.

Cheers George
Post removed 
Post removed 

You can either make it up or you can look it up. You choose to make it up. I go look it up. I'm pretty sure I made my case and you don't know what your talking about. AGAIN. 
You have no idea, what constitutes current ability into low impedance loads with complementary Mosfets vs Bi-polars.


unless V-fet is used, Mosfet always sounded better to me compared to bi-polar.
If memory serves, BTL usually raises the total output impedance. Perhaps that's a contributing factor in liking it or not?

Otherwise, the discussion about current/power delivery in MOSFETs vs bi-polar transistors is silly. Parallel devices, driver stages and feedback are all able to create any amount of current and power with either device type.

Which brings me to another point: Feedback, and lack thereof, can have a profound effect on the amp sound and characteristics. Luxman amps rely heavily on global feedback in their Class A amps, and they are fabulous. Some other Class A makers use local or no feedback at all. I suspect that speaker matching, and what sound you like lies largely here.
The classic comparison is between Pass And Boulder Warm VS Dry, Tubish vs Transistorish. I am not totally sure about this but there is something about Class A bipolar amps which causes the designers to want to use fans and fancy bias tracking schemes. Do they generate more heat or require higher bias voltages. Pass seems to always avoid this but he does not run his power higher than 300 watts whereas Boulder will go much higher. JC simply switches over to AB which you can hear in the JC 1. Push it into steady AB and it loses its effortlessness. It still gets as hot on low impedance loads as a Pass amp. 

I cant find anything to disagree with in your post up there. It's totally silliness to make any unqualified statement about a particular devices currently delivery ability since so many other factors contribute to it. 

I agree with your assessment of feedback too. My F5 is a class A amp that utilizes "generous" amounts of negative current feedback and I've never heard anybody say that amp sounds bad for it. Feedback is something that needs to be applied with careful consideration and shouldn't be as universally disparaged as it is. 


The reason you see BJT amps with fans and thermal tracking schemes has to do with the characteristics of BJTs. They're highly prone to thermal runaway where as the device heats up, it conducts more current, which in turn raises their temperature, which causes them to conduct even more current. This characteristic does produce distortion which also requires compensation. MOSFETs, when strapped to a big enough heat sink, need no thermal tracking because their conductivity vs temperature will reach an equilibrium at some point so long as the bias is within the operating range of the device. Some MOSFETs, like laterals, have virtually no thermal drift and you can set bias on stone cold devices to find the bias only creep up slightly as the thing hits 75°C. 

It all depends on what the designer is trying to achieve as for what gain devices he chooses. Pass certainly uses plenty of BJTs in his older designs, but he uses MOSFETs these days for the exact reasons you described. Simpler circuits and a more organic, tube-like sound. 
It all comes down to the way the amp sounds into a specific speaker and the only way to know that is to hook them up and have a listen. You can make assumptions all day long and many of them will prove right. But, assumptions are the mother of all f--k ups. A less expensive amp may sound just as good as a more expensive amp into a specific speaker and we are all price sensitive to one degree or another.  
I have heard great examples of both types. It is interesting that Atma-sphere amps have been forgotten in this discussion. They are class A OTLs Tube amps and many say better than any type of class A SS amp.
They all get hot so why not have a little glow with it? 
Ralph designs his amps to drive fairly benign, efficient speakers. If you like speakers like Focal or Wilson or KEF, forget Atma-sphere amps. 
Mijostyn: I enjoyed your Under Seige II reference to assumptions.

As a close alternative to large class A SS amplifiers, the GamuT series of class AB amplifiers with 1 large industrial
size mosfet per phase (2 large NPN mosfets per channel) also provides a full organic sound.

Thanks to all respondents!
Kosst I think Ralph would give you a huge argument on that one. It has to do with the relationship between sensitivity and efficiency which are not the same. Anyway, he says his amps drive low impedance loads just fine. He has several amps on Wilsons and the owners love them.  
The best way to buy an amp is to listen to it in your system which highlights the value of local outlets. We use to take amps to peoples houses all the time and leave them for up to a week. We would install them in the system and make sure all was OK then pick it up a week later. We would do the same with smaller speakers. Larger speakers you had to listen to at the store but we would hook up whatever electronics you wanted to the system.  This was all by appointment. It is a travesty that most of these stores have failed. They had to become theater outlets for wealthy people but many of them were in areas where there were not many wealthy folks or the theater thing did not catch on in time to save them. But, you can't listen to an amp on the internet. Some manufacturers understanding this allow 30 day in home trials gambling that the customer will keep the amp.
Ralph if you are listening I think you should build one of each model for in home trial purposes with a full deposit on the amp which will cut off people who are really not interested. Also make the customer responsible for return shipping. Granted tossing MA-3s around would be.........difficult. But, I think MA-2s would give them a solid idea of what they would get. People  who are looking at the MA-3s are power hungry
They won't settle for a measly 220 watts. That would be me:) That would make the decision to plop $147k on amps easier. I also have another idea. You should have customers with your amps in acceptable and presentable systems bird dog for you. Have people interested in your equipment make appointments with your "bird dogs" to listen to the amps. If they capture the customer give them a fixed fee for their involvement or a percentage of the sale. That might be cheaper than shipping amps all over the country. Just a thought. The Atma-Sphere mini dealer program:)))

I seriously doubt he would. Look at the specs of those amps. The best of the bunch has a damping factor of 4.5 into an 8 ohm load. That strongly suggests they don't have the ability to control more than modestly reactive loads. Given the volume of words he's written in defense of benign speakers and how they benefit amps, I'd be stunned to see the guy say "NY amps will work great with Sopra No.3's!"
kosst I am not saying that Atma-spheres are the absolute best amp for those speakers but if someone were using a sub woofer and liked the other attributes of Ralph's amps. Why not? Damping factor is not everything. There are those (mostly Mcintosh owners) who say that a damping factor of 50 is more than enough. I certainly do not think that is the case with sub woofers. You can get damping factor way up there by cranking up the negative feedback resulting in a rather harsh sounding amp. Read this
My point is that although you can draw some good assumptions about how an amp will perform on a given loudspeaker you do not know for sure until you pair them up. I draw assumptions all the time but usually when I have overwhelming evidence. I am a SoundLabs fan. The MA-2 is the pet amp for SoundLabs users. These speakers have a rising impedance in the bass. I have heard figures up to 30 ohms. Perfect for these amps. I would also never have a system without a sub woofer array and I always use a very powerful SS amp with high damping to drive it usually a rather inexpensive commercial AB amp. Using an MA-2 for this would be rather silly from a cost and performance perspective. But, you can see why I have a strong interest in these amps.
Atma-Sphere does push high impedance loudspeakers because his amps work better on them as you suggest and most people can only afford his smaller amps. High impedance does have it's advantages in terms of efficiency. Klipsch Heresys on stands with two sub woofers can put on a surprisingly good performance.  Personally when I see Wilson Loudspeakers the first thing that pops into my head is R2D2. If I were to buy dynamic loudspeakers, which will never happen I would go for a large set of Magicos. 
I'm just going off the specs of the amps and all the stuff I've seen the guy post concerning speakers. Ralph doesn't like or endorse speakers that are inefficient and challenging to drive and he doesn't design his amps to drive loads that challenge the amp to actually grip the speaker. Of course, if I'm wrong, I'll gladly take correction from the guy. 

As for what the specs of an amp can tell you about how it'll sound with a given speaker, they really do say a lot if the specs are expansive enough. I like First Watt amps, but I write off most of them based on just the specs of the things. The F2, F3, F4, J2, and all of the SIT amps would drive my speakers horribly. The F5 obviously works VERY well. The F6 and F1 would probably get the job done with style. The F7 makes me curious. That's a tough one to figure out since it's such a strange amp. I honestly have no intention of ever getting into tube power amps. They're lethal to work on and I'm not too fascinated by gain devices that burn out like light bulbs. 
I have not had a tube amp since 1969. I have blown a couple of SS amps, a Fuzz Linear and a Krell KMA 100. Tubes do have a major advantage in that they are easy to change. Transistors not so easy. I only care about what sounds better. If I have to change a tube once in a while that is no big deal. You settle on speakers and build the rest of the system around them. Since the Krells I have stuck with class A amps.
They have a way of sounding more powerful than they actually are. Perhaps they handle transients better. The Atma-Spheres drew my attention because SoundLabs owners love them and they are for sure my next and probably last loudspeaker. The Atma-Sphere and Pass amps are the only amps in the running. Xs300s VS MA-3s. I seem to be always dreaming about the next system usually without regard to the price tag:)
In term of drive, everything else being equal, BJT usually has more output current output so it should be able to drive better vs. FET.  Therefore BJT sound tends to have more dynamic compared to FET.  

With respect to output impedance, I suspect most amplifier output stage uses push-pull config which is a combination of N and P devices either they be BJT or FET.  BJT usually uses common-emitter config, whereas FET would use common-source, so either FET or BJT they should have fairly low output impedance so driving difficult speaker loads probably shouldn't be a problem.  The rest depends on how many devices is being used in parallel and how feedback is used to control the output impedance.

With respect to the sound, the FET is known to have a more tubey kind of sound whereas BJT has a cleaner, more solid state sound.  I think it may have to do with the FET device being a square device.  That is the FET output current is the square of the input voltage proportionally.  So FET tends to have mostly second order distortion just like tubes hence it having similar sound to tubes.

BJT on the other hands is an exponential device that is the BJT output current is exponentially proportion to the input current.  Therefore the BJT will have a lot more higher order distortion.  Higher distortion usually is associated to the sound being a bit lean, more analytical, and hence being more solid states.

Some amplifier designs use FET as input stage and BJT as output stage just like a hybrid amp that uses tubes as input but the output state is solid state.
In term of drive, everything else being equal, BJT usually has more output current output so it should be able to drive better vs. FET. Therefore BJT sound tends to have more dynamic compared to FET.

Cheers George