• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Class A “class D killer” amplifier with THD less than -120dB

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
1,357
Likes
1,507
Location
Ichihara City, Chiba Prefecture, Japan
There are some advantages to using linear power supplies as opposed to Switch Mode Power Supplies, which by the way work differently from PWM.

Thanks a lot, again, for your kind correction for my naiv confusion of Switching Mode PS and PWM. Yes, I now understand that DP-NC400-4-EXP (Class-D digital power amp) has linear transformer PS (not a Switching Mode Pulse PS) in outer boxes providing power to four Hypex NC400 working in PWM mode;
http://www.dentec-at.com/product258_0.html
 
Last edited:

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
2,183
Likes
2,070
Location
.de
I have built lots of amp kits and other electronics from scratch but landed with three similar Pioneer M-22 class A amps. They put out about 30wpc at clipping and measure like a high quality preamp. I have had two of the three on a distortion analyzer lately and the are working fine, thank goodness, in spite of being new in 1976. They are toasters idling about 250 watts iirc.
PP Class A efficiency is 25% max, so that's 240 W in for a 30 wpc amplifier right there, plus power supply losses... 270-280 W easily. Or if all three are running at the same time, at least 800 W. :eek:

You can potentially double that figure if it's summer and AC has to shovel the heat outside.

I would be awfully reluctant to operate such a bunch of power hogs for any amount of time, it would cost a small fortune.

Anyway, a comparison of the new Topping LA90 amplifier with pma's would be interesting. That's a superb 36 wpc Class AB job with similarly low distortion and even lower noise (a 120 dB dynamic range at 5 W / 4 ohm out and 20 dB of gain is quite something!).
 

garbz

Active Member
Joined
Sep 17, 2021
Messages
102
Likes
125
You can potentially double that figure if it's summer and AC has to shovel the heat outside.
You've touched into one thing that pushed me away from Class-A. It just wasn't pleasant listening to music no matter how good it was when I had 2x 1000W spaceheaters in the room on a 32degC day.

You don't need to want energy efficiency just to save the environment. Sometimes energy efficiency is just to save your own sanity. That and when the coffee machine was on and the fridge compressor started I kept tripping the circuit breaker.
 

gene_stl

Addicted to Fun and Learning
Forum Donor
Joined
Jun 14, 2019
Messages
730
Likes
949
Location
St.Louis , Missouri , U.S.A.
If you don't want to use class A, don't.

The gas company used to have an advertising phrase "the energy you save belongs to all of us, the money you save is your own."
So far, I can afford to run my little toasters and for half the year they make the room warmer , sort of like tube amps which are even more environmentally irresponsible.

Here in St. Louis , in the summer it gets way hotter than 32 degrees C more like 42. (we regularly see over 100F every year) I work on A/C and keep my units running well. I don't leave my audio gear on all the time. And I AM aware that the A/C has to dump the heat outdoors.

I would be delighted to have a stack of LA90s. Anyone that wants to send them to me PM for my address. I suspect that after Amir's recent review they will become unobtanium, at least for a while.
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
6,279
Likes
11,841
Location
Monument, CO
PP Class A efficiency is 25% max, so that's 240 W in for a 30 wpc amplifier right there, plus power supply losses... 270-280 W easily. Or if all three are running at the same time, at least 800 W. :eek:

You can potentially double that figure if it's summer and AC has to shovel the heat outside.

I would be awfully reluctant to operate such a bunch of power hogs for any amount of time, it would cost a small fortune.

Anyway, a comparison of the new Topping LA90 amplifier with pma's would be interesting. That's a superb 36 wpc Class AB job with similarly low distortion and even lower noise (a 120 dB dynamic range at 5 W / 4 ohm out and 20 dB of gain is quite something!).
Minor correction: theoretical max efficiency for push-pull class A is 50% IIRC. It is down to 27% or something like that for single-ended designs. I have the derivations in my notes, somewhere, probably in the black hole masquerading as our basement storage room...
 

MakeMineVinyl

Major Contributor
Technical Expert
Joined
Jun 5, 2020
Messages
3,345
Likes
5,257
Location
Santa Fe, NM
I would posit that the majority of class A amplifier users (especially ones like me who use SETs for the HF in an active system), have very sensitive speakers and could easily (like me) get by with an amplifier which is < 5 watt (mine are 2.5 watts), so the total AC power used is not very significant. In the case of my SETs, there is next to no heat coming from the tubes because the power levels involved are so minuscule. If anything north of around 5 watts is needed and / or the speakers are relatively inefficient, a more conventional A/B or D amp would be indicated.
 

gene_stl

Addicted to Fun and Learning
Forum Donor
Joined
Jun 14, 2019
Messages
730
Likes
949
Location
St.Louis , Missouri , U.S.A.
You can get similar or better performance with a Neurochrome Mod 86 amp without the crazy heat and cost of a Class A. https://neurochrome.com/products/modulus-86
I have linked to Neurochrome numerous times here and have a kit waiting to be assembled. What sent me to his website was the statement at Sigfried Linkwitz site "I don't know how to measure such low distortion".
 

levimax

Major Contributor
Joined
Dec 28, 2018
Messages
1,374
Likes
1,985
Location
San Diego
I have linked to Neurochrome numerous times here and have a kit waiting to be assembled. What sent me to his website was the statement at Sigfried Linkwitz site "I don't know how to measure such low distortion".
I have a tri-amp system and I have used 4 Nuerochrome amps for the mids and tweeters for several years now. The kits go together well and have performed as advertised for me. I am not a big believer that you need distortion so low but at least you don't have to worry about it. What I did notice is that the amps are dead quiet, even when directly connected to the tweeters. Good luck and have fun.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
1,357
Likes
1,507
Location
Ichihara City, Chiba Prefecture, Japan
I would posit that the majority of class A amplifier users (especially ones like me who use SETs for the HF in an active system), have very sensitive speakers and could easily (like me) get by with an amplifier which is < 5 watt (mine are 2.5 watts), so the total AC power used is not very significant. In the case of my SETs, there is next to no heat coming from the tubes because the power levels involved are so minuscule. If anything north of around 5 watts is needed and / or the speakers are relatively inefficient, a more conventional A/B or D amp would be indicated.

As you may well aware, I am always on the same path!
 
OP
pma

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
3,012
Likes
6,050
Location
Prague
THD vs. power at 1kHz with 4ohm load. We can see that about 18W the amp moves from class A to class AB, still keeping very low distortion. During this transition, we can see in the spectrum that the distortion that was purely H2 and H3 within class A suddenly shows also higher order distortion components, though very low in level.

classA_1k_thdpower4R.png


and THD vs. frequency

1651136262515.png
 
Last edited:

garbz

Active Member
Joined
Sep 17, 2021
Messages
102
Likes
125
Minor correction: theoretical max efficiency for push-pull class A is 50% IIRC.
There's a major caveat to that. The only reason a push-pull Class-A can approach 50% is due to them needing to be transformer coupled, otherwise in a push pull configuration one transistor will switch off (which is the whole point of the design) and its no longer considered Class-A. But you can make single ended Class-A designs with 50% efficiency, you just need to transformer couple them. ... which we don't because it performs poorly.

All of this is purely academic though. Actual efficiency in operation is tiny compared to theoretical efficiency at clipping point.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
6,279
Likes
11,841
Location
Monument, CO
There's a major caveat to that. The only reason a push-pull Class-A can approach 50% is due to them needing to be transformer coupled, otherwise in a push pull configuration one transistor will switch off (which is the whole point of the design) and its no longer considered Class-A. But you can make single ended Class-A designs with 50% efficiency, you just need to transformer couple them. ... which we don't because it performs poorly.

All of this is purely academic though. Actual efficiency in operation is tiny compared to theoretical efficiency at clipping point.
That is not the way any class A push-pull circuit I have seen works. Standing bias current must support class A, but as long as you control the offset properly the amp does not need to be transformer-coupled. And you can design a class-A single-ended amp that does not need to be transformer-coupled, but I have no idea what all the little single-ended class-A audio amps are doing ( not my field).
 
OP
pma

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
3,012
Likes
6,050
Location
Prague
In a push-pull class A, and provided that we stay in class A (both halves remain conductive with I>0), we can get maximum efficiency of about 35-40% in a real circuit. 50% is just a theory, no loses, no saturation voltages, ideal components.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
6,279
Likes
11,841
Location
Monument, CO
In a push-pull class A, and provided that we stay in class A (both halves remain conductive with I>0), we can get maximum efficiency of about 35-40% in a real circuit. 50% is just a theory, no loses, no saturation voltages, ideal components.
Correct, agreed, just like 27% is the theoretical max for single-ended. And you have to add input stages, bias and control circuits, etc.
 
OP
pma

pma

Major Contributor
Forum Donor
Joined
Feb 23, 2019
Messages
3,012
Likes
6,050
Location
Prague
Some thoughts on efficiency of this amplifier sample. Total supply voltage (+rail, -rail) is 47.2V. Current consumption is 1.45A, it makes 68.44W power consumption at idle.
LT1122 (allowed rail voltage up to +/-20V, real supply +/-19V here) gives very good results in this circuit:
classA_1k_thdpower_LT1122_4R.png


We get 16.8W in pure class A into 4ohm (this would make bad efficiency, 8ohm would make better), but also 30W at THD=0.004% in class AB, and it would make 44% efficiency, which looks much better. 1% distortion is at about 35W and then we get that 50% efficiency, though not in a pure class A and tolerating 1% distortion (quite high in my view).

This class A amplifier handles difficult load very well, below is a distortion spectrum at 10kHz into 4ohm resistor and 4ohm//1uF resistor-capacitor parallel combo. 9W/4ohm power. This is quite extraordinary.

classA_10kHz_9W_4R_dB.png


classA_10kHz_9W_4R+1uF_dB.png
 

garbz

Active Member
Joined
Sep 17, 2021
Messages
102
Likes
125
In a push-pull class A, and provided that we stay in class A (both halves remain conductive with I>0), we can get maximum efficiency of about 35-40% in a real circuit. 50% is just a theory
Can you give a simplified diagram of what that would look like? Because all my imagination is failing me. If both halves of the push pull pair need to be biased into full c conduction for all 360deg then the max theoretical is back down in the 20s unless presented with a varying inductive load, if they are biased such that one of them falls out of conduction then it is no longer class A is it, since the definition is 360 Deg conduction of each transistor not just the stage itself.
 
Top Bottom