Nordic said:
But which one for an EF output stage ?
Example :
The Parasound JC-1 runs 9 pairs of 2SA1295/2SC3264 in the output stage.
Average Cob of the Sankens is 250pF, 9 in parallel already amount to 2250pF.
The SR spec of the JC-1 is 130V/uS, means having to push close to 600mA into 4500pF to reverse-charge the output devices fast enough.
But Alas Yorick, in reality the driver stage of the JC-1 is biased at a fraction of the 600mA figure, basically what the driver stage is for is to reduce the bias needs of the output stage.
(at least that's what my schematic of the Halo JC-1 says, and i tricked JC into answering the bias question.
)lumanauw said:
I have done this and found that, apart from the heat, class A is better in every way but would also only suggest using it if the rest of the system can reproduce it's advantages.
To me the main advantage of class A is mainly in detail and low level sound reproduction (ambience).
I'm quite sceptical about single ended class A designs, a complementary output stage can run on half the current or allow for higher output power.
But not if it sounds better.tschrama said:
The piece of wire with gain. Been there, built that, have bought that, and listened to that. And I don't like it--- not when I have heard how much "better" you can go. The single ended input stage adds predominantely even harmonic distortion, which the ear likes ( mine certainly do ). Noise was mentioned as well. I used too look at signal to noise ratios as an important parameter, but have long since come to the conclusion it's relatively unimportant. Not hum and PSU problems of course, I mean good old hiss. Does the ear use hiss as a form of "dither" to extract more detail and to stop it subconsiously "searching" for low level detail that is not there ?
Mooly said:
If your looking for richness of tone then you might want to consider a valve amplifier or even a valve pre amp.
I built a valve preamp mixer for my disco and guitar.
It uses 1/2 an ECC83 as mixer and one of the input channels has the other 1/2 of the ECC83 for extra gain. I use the extra gain channel for my guitar and it seriously rocks.
nigelwright7557 said:
If you want to listen to a sine wave, stay with class AB, if you want more detail and ambiance from the music, try class A. My amplifier can switch between Class A (up to full power) and AB (class A up to quarter power), when I want to listen to music it will be class A.
Mooly said:
I did indeed. It was called the Trimodal amplifier, and it appeared in Electronics World in June & July 1995.
You can get the circuit from The Audio Power Amplifier Handbook,
See: http://www.dself.dsl.pipex.com/ampins/books/book.htm
or Self on Audio
See: http://www.dself.dsl.pipex.com/ampins/books/book2.htm
The Signal Transfer Company still sell PCBs for it.
See: http://www.signaltransfer.freeuk.com/
I still have the prototype; it runs in Class A in winter and Class B in summer.
Oh, and all three modes sound fine.
nigelwright7557 said:
For one, with large dynamic range and difficult classical junk, the temperature of a Class AB power amp is constantly altering.
Class AB biasing may eliminate XO distortion and optimal-AB biasing will put delta-V across the emitter resistors in the sweet range, but does nothing to stabilise the driver and output stage's operating temperature.
Which is where the big Class AB hunks enter.
Current production Halo JC-1 amps have 0.15 Ohm resistors in the output stage (says JC), and the 9 pairs of output devices of the Halo are biased in the lower part of the optimal AB range. With +/- 90V rails, that puts out so much quiescent heat that the power amp will have to be driven really hard for the operating temperature to change significantly.
In effect, a high parallel number output stage mimics the thermal behaviour of a Class A output stage up to levels above the Class A crossing point.
(some folks built EC Class AB power amp designs in the midst of the '80s, to revert back to heavy quiescent level output stages later on)
Hi,
if one were to optimally bias a 9pair output stage using Re=0r1 & Vre=20mV then each device is biased to 200mA and the complete output stage has 1.8A of ClassAB bias.
This gives a peak ClassA output current of 3.6Apk and about 52Watts of ClassA power into an 8ohm load.
If this same output stage were powered from ~+-65Vdc power rails then it has the potential to develop a maximum of ~200W into into 8r0 and if the PSU is up to it ~400W into 4r0.
The majority of "loud" listening is going to stay in the ClassA region of power delivery.
If the amp is properly designed, then the bass performance will be exemplary, no question.
Jacco & Andre are on the ball, keep it up.
if one were to optimally bias a 9pair output stage using Re=0r1 & Vre=20mV then each device is biased to 200mA and the complete output stage has 1.8A of ClassAB bias.
This gives a peak ClassA output current of 3.6Apk and about 52Watts of ClassA power into an 8ohm load.
If this same output stage were powered from ~+-65Vdc power rails then it has the potential to develop a maximum of ~200W into into 8r0 and if the PSU is up to it ~400W into 4r0.
The majority of "loud" listening is going to stay in the ClassA region of power delivery.
If the amp is properly designed, then the bass performance will be exemplary, no question.
Jacco & Andre are on the ball, keep it up.
There's another important thing that classA have (that classAB don't) besides xover distortion properties. Power supply ripple. ClassA can sound like classA because in it's PS rails the ripple is low and constant. ClassAB don't have this feature. This ripple will enter the audio CCT one way or the other. This is why when we make "smart" output stage (that eliminates xover distortion or never turnoff), the sound can never be the same with real classA.
Jacco, what is your email?
Build an amplifier with adjustable bias. That's the only way to compare the "sound of classes".
I have built an amp (OK, it was mono only) with adjustable bias from no bias (class B/C) to high bias (around 0,3A, that means class A at normal listening levels).
The second adjustment was feedback factor going from ~30dB high open loop bandwidth to ~60dB low open loop bandwidth.
Both were adjustable with user-friendly pots and I asked several people not knowing what they adjust to find the best sound point.
Most (not all) have heard the difference in bias and have set the more the better (no class AB "sweet spot" exists in real life listening conditions) and nearly none have heard the difference in feedback. Those who did not say "it doesn't do anything" have set the minimum feedback.
That's the way reliable comparisons should be so that one can draw general conclusions.
Regards,
Adam
I have built an amp (OK, it was mono only) with adjustable bias from no bias (class B/C) to high bias (around 0,3A, that means class A at normal listening levels).
The second adjustment was feedback factor going from ~30dB high open loop bandwidth to ~60dB low open loop bandwidth.
Both were adjustable with user-friendly pots and I asked several people not knowing what they adjust to find the best sound point.
Most (not all) have heard the difference in bias and have set the more the better (no class AB "sweet spot" exists in real life listening conditions) and nearly none have heard the difference in feedback. Those who did not say "it doesn't do anything" have set the minimum feedback.
That's the way reliable comparisons should be so that one can draw general conclusions.
Regards,
Adam
Mooly said:
Now that could be seen as a missed marketing opportunity, a weather sensing Class ?????
That would be simple enough to do with an LM35 temp sensor and comparator, but it the sensor would have to be outside the amplifier box. The original article (part 2) did not mention weather sensing, but it did say this:
"One interesting extension of the ideas presented here is the Adaptive Trimodal Amplifier. This would switch into Class-B on detecting device or heatsink overtemperature, and would be a unique example of an amplifier that changed mode to suit the operating conditions. The thermal protection would need to be latching; flipping from Class-A to Class-B every few minutes would subject the output devices to unnecessary thermal cycling."
Personally, I think a weather-sensing trimodal amplifier would be super-neat.
Put some high-watt devices on and hook your AC thermocouple to it with a fan.
It would be best that the fan was pulling air from inside the room and pushing it through the heatsink and then through the air ducts. However, I would not recommend completely relying on this as a heating solution.
I will say that on a fanless heatsink, a whole lot of heat gathers in a very restricted area. This is simply unnecesssary. Why turn on the AC when you could just set a nice quiet fan by your class-A amp? You would have lower heating costs, and (thoeretically) better sound. You could also set your amplifier right in front of an air duct, so you wouldn't need a fan. If one were so amitious as to put the amlifier inside the air duct, you would have a faraday cage formed by the metal enclosing!
The resourcefulness of the general republic nowadays I think needs a primer.
It may be weird, but it's worth it. Someone could make money!
- keantoken
Put some high-watt devices on and hook your AC thermocouple to it with a fan.
It would be best that the fan was pulling air from inside the room and pushing it through the heatsink and then through the air ducts. However, I would not recommend completely relying on this as a heating solution.
I will say that on a fanless heatsink, a whole lot of heat gathers in a very restricted area. This is simply unnecesssary. Why turn on the AC when you could just set a nice quiet fan by your class-A amp? You would have lower heating costs, and (thoeretically) better sound. You could also set your amplifier right in front of an air duct, so you wouldn't need a fan. If one were so amitious as to put the amlifier inside the air duct, you would have a faraday cage formed by the metal enclosing!
The resourcefulness of the general republic nowadays I think needs a primer.
It may be weird, but it's worth it. Someone could make money!
- keantoken
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.