advertisement


So Class A - do it if you haven’t!

Arkless Electronics said:
Yep. And on the resistors! The test points will be across said resistor/s. In the most common output stage the point where the two resistors connect together is the actual output of the amplifier before it gets to Zobel and Thiel networks and protection relay etc etc. In the next most common they are between the PSU rails and the output transistors.
Click to expand...
I think I'll bow out gracefully at this point... I've spent the last 20 minutes staring at the schematics and my eyes just glaze over! :eek::D
 
Arkless Electronics said:
There's loads of things wrong in that review and I simply don't believe it can do anything like 60W in class A. In fact for it to do 60W class A and then go on to give 600W is impossible (well each monoblock would need to be the size of a USA fridge anyway!) unless it's actually class G or H. It's also too cheap and the temp rise is too little for class A.

Generalising, an extra zero can be expected in the THD figure for class A and often better still as frequency rises.

For me the biggest attraction in class A is that it increases "intrinsic", "native" measured performance by such a large margin that low NFB can be used (or even no NFB in some cases!) without THD being above say 0.1% and with good damping factor. As I hinted at up thread, a myriad of things are improved by up to an order of magnitude simply by using class A... just one of the many things you never see in non technical writings on it is that in class A/B LOADS of distortion (maybe 30-40%! I'm guesstimating here) is forced back on the power supply whilst in class A it is absent. In some class A topologies it is even possible to make it so the PSU see's just a constant DC draw as if it was just heating a big resistor! There are many more advantages to class A....
Click to expand...
Emotiva hires (or used to hire) a professional independent test lab to verify its' claimed specifications, which I will try to locate for you.

Remember, this is a very large mono amp with dual heatsinks at north of 50lbs per side.

It does get very hot in high bias mode and draws over 200w per side at idle.

And like I said, the claimed measurements improvements of class A in terms of noise and distortion are just not really there in this case.

Here are the Audio Precision PDFs - A and AB, 8 and 4 Ohms:

https://drive.google.com/file/d/17t6RKOOB6Zfmbn3Bqb8qn4NgaYvZg9VW/view?usp=drivesdk

https://drive.google.com/file/d/17dtd46bq5E-a-RtKmtYn0IpogQeW9qEX/view?usp=drivesdk

https://drive.google.com/file/d/17d3RXujYQNmYslymjd83d_hCJnd8xIN1/view?usp=drivesdk

https://drive.google.com/file/d/17c_YM3LtQk4UMox1TB8zNkQWVM_Grm1w/view?usp=drivesdk
 
Last edited:
DimitryZ said:
Emotiva hires (or used to hire) a professional independent test lab to verify its' claimed specifications, which I will try to locate for you.

Remember, this is a very large mono amp with dual heatsinks at north of 50lbs per side.

It does get very hot in high bias mode and draws over 200w per side at idle.

And like I said, the claimed measurements of 10X superiority of class A in terms of distortion are just flat wrong in this case.
Click to expand...

Nothing personal but your level of understanding is not sufficient to make any real claims.... I've refrained from going into all the details of class A specifically because most do not have the knowledge to interpret things correctly and many make the ultimate sin of quoting test results for a specific amplifier as evidence... For a start NFB being the most important thing in amplification means it could be used to move a zero back or forth an order of magnitude anyway...

No discussion trying to get into the meat of the subject will get anywhere at all if people have little or no understanding of things like NFB, transconductance, stability margins, loop gain etc etc I'm afraid.
 
ToTo Man said:
I think I'll bow out gracefully at this point... I've spent the last 20 minutes staring at the schematics and my eyes just glaze over! :eek::D
Click to expand...

Gawd knows why it matters to you but it's R146, R129 one channel and R246, R229 the other and they are 0.1 Ohm...
I have the schematic from when you asked about the phono stage in it once... It's mosfet so bias is rather less critical than with BJT's.
 
Arkless Electronics said:
Nothing personal but your level of understanding is not sufficient to make any real claims.... I've refrained from going into all the details of class A specifically because most do not have the knowledge to interpret things correctly and many make the ultimate sin of quoting test results for a specific amplifier as evidence... For a start NFB being the most important thing in amplification means it could be used to move a zero back or forth an order of magnitude anyway...

No discussion trying to get into the meat of the subject will get anywhere at all if people have little or no understanding of things like NFB, transconductance, stability margins, loop gain etc etc I'm afraid.
Click to expand...
Nothing personal, see above.


From Class A and AB 8Ohm reports above, with the first taken at 400/500w and latter at 600w:

THD+N Ratio, A-weighted: Class A - 0.00344%: AB - 0.00787%

S/N Ratio, A-weighted: Class A - 91.0dB; AB - 89.2dB

THD+N vs. Level, 1KHz: Class A - 0.5%@0.1 mW to 0.005%@0.1W to 0.003%@400W; AB - 0.005%@0.1W to 0.005%@500W, rising to 3%@600W.
 
Last edited:
Arkless Electronics said:
Also all class A amps go into A/B as the load impedance falls.
Click to expand...
Except for single-ended contraptions, of course. I think that is the position @Old Shatterhand is taking. If any amp delivers even a micro-watt in class AB, cannot be considered class A. That, as you explain below, is utter cack, of course.

Arkless Electronics said:
An amp that makes the first 5W of 50W even in class A is still "pure" class A for the first five Watts and is in fact the most sensible way of doing things as few of us use more than 5W for most of the time.
Click to expand...
 
I love the sound of my Musical Fidelity AMS-35i and have no plan of replacing it even after 10 yrs of ownership.
 
Has anyone ever been able to hear, in the music, the point at which an amp crosses the A to A/B watt threshold?

No A/B amps I've had have sounded as good as the current class A one at low or any volume, but obvs other factors could be at play.
 
Arkless Electronics said:
There is usually a specific quiescent current that gives lowest distortion when aiming for "optimum class A/B" in a bipolar transistor output stage BTW and in the example above of 33mW class A, which comes from a bias of 45mA (10mV across 0.22R), distortion may get worse if you set it to say 80mA in the hope of better results. The true optimum figure needs a THD analyser and oscilloscope to find it but when a service manual says eg "set bias at 10mV across points A and B after leaving unit to warm up for 20 minutes" it will of course be "near enough".
Click to expand...

For anyone interested enough, here's a practical dem of Jez' thorough-going point about 'near enough' on bias - using that classic pfm guinea pig, a Nait 2, and erm, written up rather a long time ago.

https://pinkfishmedia.net/forum/threads/naim-amps-bias-current-setting-and-distortion.36057/
 
Fretbuzz said:
Has anyone ever been able to hear, in the music, the point at which an amp crosses the A to A/B watt threshold?

No A/B amps I've had have sounded as good as the current class A one at low or any volume, but obvs other factors could be at play.
Click to expand...

Virtually no chance of hearing it no. It moves very gradually out of class A.
 
martin clark said:
For anyone interested enough, here's a practical dem of Jez' thorough-going point about 'near enough' on bias - using that classic pfm guinea pig, a Nait 2, and erm, written up rather a long time ago.

https://pinkfishmedia.net/forum/threads/naim-amps-bias-current-setting-and-distortion.36057/
Click to expand...

The only very small reduction in THD when temporarily going class A here suggests there are significant amounts of distortion from other issues in the design... it's low enough for jazz though:) I would expect a much bigger difference had it been done at 20KHz!
 
'other issues in the design' well- yes, TBH the excess gain in the pre-amp section would be my first thought there(and I have others ; ) ). and as I think I wrote - read it as ratiometric, not absolute; it was no attempt at a bench-marked measurement, just a closer look at the form of things end-to-end.

One day I might revisit. Might. I've still got that CD2, that Nait, and the calculator.
 
martin clark said:
Afterthought: certainly don't turn the wick right up in the hope of 'Class A' operation. No, really, don't do it.

Well, since you asked...in the interest of research... and with a little adjustment (ahem) my Nait was temporarily fudged to just squeeze the necessary 90mV of bias to run the test entirely in Class A. Result: 0.0135% THD (whereas just 7.0mV of bias resulted on 0.0165% THD) And at this level the bias was galloping away while I watched. The bias setting represents nearly 400mA of standing current, and the meter drifted well past the 500mA mark in under 15secs of the very abbreviated test run. Left unchecked the amp would have blown in short order. Now look at the 'gain': a difference less than 0.003% (< -90dB difference) but with the certainty of killing the amp. So don't believe anyone that says more bias is better.... at least without checking all the possible consequences in detail.
Click to expand...
What happens when QC is set higher than spec for a class A amp?
 
Very, very little, if the bias is& remains stable!

If the design bias was enough to deliver rated power into rated load - essentially you are just wasting energy in more idle dissipation = heat.
It is conceivable there maybe some fringe benefit for marginal design cases -but you'd have to test for this. I can't think of an obvious route for 'more = better' ; 'more' being grossly increased bias current, rather than just nominal tolerance on specified idle bias.

If the PSU, output stage and heatsinks can handle it - of course you might also achieve a little more delivered power within class A into lower impedances than specified - but since the ear is grossly insensitive to small changes in power output, it'd not be audible. For example - even to get the 8-ohm rated power in Class A into 4-ohmswithin class A you'd need to double the idle bias = not going to happen with any commercial amp.
 
You must log in or register to reply here.


advertisement


Back
Top