It is an expression I never heard before, but it is indeed perfectly understandable when you are Dutch.
> dit verskil is soos kak en kaas
"This difference is like cheesecake and cheese" ?
(Oddly, Google Translate thinks it is Afrikaans, not Dutch. Yes, I know one comes from the other.)
"This difference is like cheesecake and cheese" ?
(Oddly, Google Translate thinks it is Afrikaans, not Dutch. Yes, I know one comes from the other.)
Agree.
But complex class AB when done right, can outperform simple class A. I design class AB that can sound better than simple class A (a popular DIY amplifier).
Class AB can have far less high-order distortion, too (even very small THD).
Regarding the use of energy, I expect that most people listen at relatively low levels most of the time (see also Pano's test http://www.diyaudio.com/forums/mult...much-voltage-power-do-your-speakers-need.html ). When you listen at a low volume, you don't need exceedingly large quiescent currents for an amplifier to work in class A.
Hence, it could be useful to have an amplifier with adjustable bias current that indicates with an LED or so when it works in class A and when it transitions into class AB. You can then adjust the bias such that it works in class A at the volume at which you are actually listening.
I had an idea on how to make such an amp five years ago, but never actually built it. I will post the schematics and explanations in this thread:
http://www.diyaudio.com/forums/soli...variable-bias-indicates-whether-works-ab.html
An explanation of the basic principle is already there. The full circuit will not be very simple and it will probably require debugging, but maybe it will be of use to someone anyway.
Hence, it could be useful to have an amplifier with adjustable bias current that indicates with an LED or so when it works in class A and when it transitions into class AB. You can then adjust the bias such that it works in class A at the volume at which you are actually listening.
I had an idea on how to make such an amp five years ago, but never actually built it. I will post the schematics and explanations in this thread:
http://www.diyaudio.com/forums/soli...variable-bias-indicates-whether-works-ab.html
An explanation of the basic principle is already there. The full circuit will not be very simple and it will probably require debugging, but maybe it will be of use to someone anyway.
In my place, for a class A build I see two problems,
A. Suitable heatsink is not easily available.
B. 45+° heat index weather condition.
...guess class A is never going to be popular here. 🙂
A. Suitable heatsink is not easily available.
B. 45+° heat index weather condition.
...guess class A is never going to be popular here. 🙂
I think the Inverting JFET Circlophone has a sliding bias to keep in semi Class A. There is no manual bias adjustment on that amp.
I just looked it up, but it looks to me like a non-switching class AB amplifier with class-AB control loop, just like the original Circlophone and just like my design from the mid-1990's.
Actually that's assuming that C1 in http://www.diyaudio.com/forums/solid-state/189599-my-little-cheap-circlophone-34.html is only meant to be a frequency compensation capacitor. If it is chosen so large that the ripple on the bias currents is small within one signal cycle, then you are correct.
Personally I would prefer a manual control over a slow control loop that always lags the dynamics of the signal, but that's just a matter of personal preference.
Personally I would prefer a manual control over a slow control loop that always lags the dynamics of the signal, but that's just a matter of personal preference.
Marcel, I would think Ian Hegglun's CubeA would also fit that category. Class A yet only 1/4 of the 'normal' bias.
Jan
I think Ian Hegglun's square and cube law amplifiers are in a class of their own: they are somewhere in between non-switching class AB and normal push-pull class A.
In normal push-pull class A, the output transistors conduct current continuously, and for each milliamp that one transistor is driven above its quiescent current, the other is driven a milliamp below its quiescent current.
In non-switching class AB, the output transistors conduct current continuously, but when the current through one output transistor becomes large, the current through the other output transistor approaches some fixed minimum current.
Ian Hegglun's square and cube law amplifiers do something in between.
In normal push-pull class A, the output transistors conduct current continuously, and for each milliamp that one transistor is driven above its quiescent current, the other is driven a milliamp below its quiescent current.
In non-switching class AB, the output transistors conduct current continuously, but when the current through one output transistor becomes large, the current through the other output transistor approaches some fixed minimum current.
Ian Hegglun's square and cube law amplifiers do something in between.
people believe class A is better because the transistors are always on and therefore do not enter the nonlinear turn on / off phase at 0.7v.
with proper stability and mosfet outputs it is impossible to tell the difference sonically between full class A ( 3+ amps bias) and class AB ( 100ma of bias).
Class D can also outperform Class A as the resolution is highest at lower power levels
If you look at a Pass Amp, you will see that Nelson kept the parts count low and open loop gain low so that small amounts ( if any ) NFB is used. Tube amps are very similar in this regard.
If you are comparing a high NFB amp to a low NFB amp, you will be able to hear a difference between them.
with proper stability and mosfet outputs it is impossible to tell the difference sonically between full class A ( 3+ amps bias) and class AB ( 100ma of bias).
Class D can also outperform Class A as the resolution is highest at lower power levels
If you look at a Pass Amp, you will see that Nelson kept the parts count low and open loop gain low so that small amounts ( if any ) NFB is used. Tube amps are very similar in this regard.
If you are comparing a high NFB amp to a low NFB amp, you will be able to hear a difference between them.
Last edited:
Hmmm Someone Really needs to get out more 😉.. absolutely so, if one believes that.
Ohh I see you just edited that Drivel
Ohh I see you just edited that Drivel
High-performance class D amplifiers typically use enormous amounts of negative feedback. If you would make a class A amplifier with fifth-order feedback, it should easily be able to outperform any class D amplifier. Of course there is no reason why this would be necessary if you simply want to listen to music.
- Status
- Not open for further replies.