Officially Class A


2002-03-15 1:15 am
I just bought anX150 for my Apogee Duetta Signatures. A buddy bought a Llano Trinity 200 (200 watts/side) the same time. At an apogee site <> there has been an ongoing hot discussion whether the Llano is truly Class A throughout its rated power using moderate heat sinks that don't get hot. Randy White, the designer insists it is. After a phone conversation with Randy, one member came back with this to report:
Just got off the phone with Randy White. His amps run pure class A to rated power. They run cooler because the new MOSFETs he is using have a much lower impedence than usual. Even so the Amp idles at 300watts/Ch. The Trinity 200 that is. These Amps have an efficiency of 66% at full power. Not bad at all. Of course it stinks at Idle but what the hey, how many of us listen at Idle.

What say you?
Lisandro_P said:

This increases efficiency, but nowhere up to 60%! I agree, it's probably class A at lower output levels and AB at full output.

Well push pull can be just under 50% so i wouldnt right it off totally.

muralman1 said:
What about Randy White's assertion that highly efficient Mosfets are largely responsible for the Class A rating?

Has almost nothing 2 do with it sorry. Pure Constant Bias Class A cannot be over 50%.

I guess althought unlikely, it's also possible Randy's amp uses a tracking power supply.
The lower impedance of the FETs has nothing to do with the amount of heat dissipated in a class A or class AB amplifier. It's only marginally meaningful in switching class D types.

Even if you meant "thermal" impedance. The answer is the same. Hogwash!

If dynamic biasing is used, then it is possible to achieve higher efficiency, but FET impedance is irrelevant.

Jim Hagerman


2002-03-15 1:15 am
I don't care one way or another. The builder insists his amp is Class A. This assertion has met up with a lot of resistance elsewhere. I was hoping for some insight here on how the 200 watt Llano could be made that way. Since The Llano is the only amp around that goes class A all the way without getting hot and with what looks like less cooling apparatus than the 20 watt Class A Pass 150, It seems to me, a builder of class A amps would be curious. this amp is a single triode leading to solid state Mosfets. Someone has said it is push pull. The builder gives credit to advanced Mosfets.

Has anyone looked at it yet?

I haven't seen (or heard) the Llano, but here is one possibility. Note that the driver stage contributes some output current and could therefore technically be considered part of the output stage. As long as the driver stays in class A -- and it will do so even while driving the outputs in class AB -- one could claim that the output stage works class A to full output.

Please, I am not putting words in anyone's mouth -- just wanted to offer one possible explanation. BTW, this approach has been used in the past by other manufacturers of "class A" amps.


  • tricon_1.gif
    6.8 KB · Views: 1,347
Plinius makes the same claim!

Thanks for your input Joe Berry,

Another popular high-end brand that seems to **** ME off is Plinius.

Last year I picked up a brochure of their SA-250 amp and the specs say when bridged, the sucker can do 900 watts into an 8 ohm load! In big red lettering across the brochure it says "CLASS A".

Taking a quick look at the SA-250 manual which is downloadable at

Looking at Page 7 it talks about Bias switching (those familiar with Plinius their SA amps can be switchable between Class A and Class AB modes). Taken from page 7:

"The current required from the mains supply in BIAS A is approximately 1,000 watts, similar to a
small electric heater. The amplifier should not be connected to a wall outlet that is shared with.... "

So right from the sheep's mouth... we're talking about a transformer that draws approximately 1 KVA.

With some simple math we take 900 / 1000 and we get 90% - so what we have is a 90% efficient amplifier? Give me a break Plinius! Even sheeps don't poop out more poop than a male cow.

Even worse, the manual quotes on page 20 explains, "Why Class A is Better".

It's sad to see how hifi manufacturers have (and still today) get away with feeding the public with so much BS. Just as bad as the hifi magazine reviewers, even the definitions of electronic terms BECOME misleading.


Plinius' claim actually may be accurate, at least for normal (unbridged) operation. Assuming that it draws 1,000W from the wall at idle in class A mode, each channel is probably dissipating about 400W, implying a true class A operating range of about 200W, and perhaps the full 250, depending on the bias circuit.

Just eyeballing it, I'd guess that the heat sinking for each channel is at about .1 degree C/W, which would give a 40 degree C rise in temperature for class A operation -- probably still safe for the output devices, but way too hot to touch comfortably.

It's true that bridged operation in itself doesn't increase the range of class A operation beyond what each channel can deliver on its own. If Plinius (or anyone else) suggests that their amp works class A at levels beyond half the amp's idle dissipation, they really ought to offer some defense for the claim.

BTW, I tried earlier to locate the idle dissipation spec for the Llano, but was unable to find it on their web site. Is it there somewhere?

The range of class A operation is mainly determined by the amp's idle current. While bridging allows twice the output voltage swing across the load, each channel in the bridge still transitions from class A to AB operation at the same point in its output current swing.

Take for example a 100W push-pull class A stereo amp in which each channel idles at 2.5A and can develop 40V peak across an 8 ohm load. The amp is class A because the current needed for 40V across 8 ohms is 5A, which is twice the idle current. Given linear operation, each side of a push-pull output stage can deliver twice its idle current before the other side shuts off.

If we now bridge the two channels, we can get up to 80V peak across an 8 ohm load. This requires up to 10A output from each channel, but since each channel still idles at 2.5A, it still transitions from class A to AB at 5A output, at which point one side of the output stage shuts off while the other side delivers the 10A to the load.

Hope this is clear -- let me know!
More on the Llano

Reviewing the info on the Llano web site, I noticed mention of a passive bias circuit. This suggests that Llano is not (any longer) using Hexfets or other trench-type mosfets that need some kind of temperatrure compensated bias circuit, such as the one shown in my schematic above.

Llano may instead be using lateral mosfets, which not only don't need a temperature compensated bias, but also tend to turn off very slowly, giving them an inherent quasi-class A behavior when used in complementary push-pull stages. It may be that the Llano design is leveraging this behavior in making its claim.