Workhorse said:Go for Balanced single ended with active current source loading Class-A using mosfets.
I would go for class A push-pull mosfet ( like IRFP240 / IRFP9240 )
If we look at Nelson Pass ... the mosfet guru of this forum
he started out with Single Ended resistor loaded MOSFETs ( balanced )
But later on, he has frequently used PUSH-PULL Class-A.
Like for example in his F4 amplifier,
which is a schoolbook example how to make a very good Class A output
http://www.firstwatt.com/products/f4.htm
I have my own taste of balanced single ended in class-A using all N-channel mosfets[no p-channel crap]I dont like complementary mosfets, because they are not complementary in reality due to vast difference in specs.
Even all my class-AB and Class-H amps also feature All N-channel output stage only
Even all my class-AB and Class-H amps also feature All N-channel output stage only
"A+C+C is close to class D, and is the best compromise between power consumption and sound quality I ever found. "
I would tend to agree, what I am calling A+G is what you would call A+C.
AB+C+C is how the original Carver Cube (M400) ran; ±25V, ±50V, ±80V. The lowest voltage and size of the heatsink (none!) prevented higher than AB bias. Efficiency on a resistive load was about 87.5% (as analyzed by Hsu when he was at MIT), and improved driving a reactive load line.
Most people don't need 80V rails, 50V rails may be adequate for all but the largest rooms or most inefficient speakers (it is only about 3dB less in any event).
±12V biased to about 1A would allow about 10W in class A (6R nominal) before transitioning to the higher rails. For average listening the amplifier may never leave class A, and would still have 10dB of headroom for demanding program material.
The 25W of bias would require about a 1°C heatsink for low temperatures, but this is no worse than what a typical 100W class AB would need (and 25W is what the 100W Hafler DH200 ran for bias, but only would deliver 0.5W/6R in class A).
I would tend to agree, what I am calling A+G is what you would call A+C.
AB+C+C is how the original Carver Cube (M400) ran; ±25V, ±50V, ±80V. The lowest voltage and size of the heatsink (none!) prevented higher than AB bias. Efficiency on a resistive load was about 87.5% (as analyzed by Hsu when he was at MIT), and improved driving a reactive load line.
Most people don't need 80V rails, 50V rails may be adequate for all but the largest rooms or most inefficient speakers (it is only about 3dB less in any event).
±12V biased to about 1A would allow about 10W in class A (6R nominal) before transitioning to the higher rails. For average listening the amplifier may never leave class A, and would still have 10dB of headroom for demanding program material.
The 25W of bias would require about a 1°C heatsink for low temperatures, but this is no worse than what a typical 100W class AB would need (and 25W is what the 100W Hafler DH200 ran for bias, but only would deliver 0.5W/6R in class A).
Class A sound better than class AB
Also simple circuits use to sound better.... small number of stages... single ended amplifiers..... even having low efficiency, they sound great.
Each component will play something into the final quality....if you could find a "wire with gain", this would be the best solution, unfortunatelly we have not that wire with gain.
If one stage can do the job to a headphone, this will be better than other amplifiers... and the quality difference is great...very big...but you will have to use the headphone...not confortable...and the quality will depend on the headphone.
A single class A stage sounds special into a headphone... and if you couple your headphone in series with the transistor colector, using small current and small power into this class A, small power stage, then you will have something close to the perfection.
Audio chain must be simple, having few parts, low transistor count, all them class A biased, deeply class A biased.... hot....very hot if your environment can hold the heat...of course you can have 100 sub circuits around... auxiliary circuits, alike CCS, current sinks, regulators, stabilizers, mirrors....but keep the audio chain clean and simple.
It is good that you want to try and research amplifiers... a long road in front of you.... enjoy and be happy.
regards,
Carlos
Also simple circuits use to sound better.... small number of stages... single ended amplifiers..... even having low efficiency, they sound great.
Each component will play something into the final quality....if you could find a "wire with gain", this would be the best solution, unfortunatelly we have not that wire with gain.
If one stage can do the job to a headphone, this will be better than other amplifiers... and the quality difference is great...very big...but you will have to use the headphone...not confortable...and the quality will depend on the headphone.
A single class A stage sounds special into a headphone... and if you couple your headphone in series with the transistor colector, using small current and small power into this class A, small power stage, then you will have something close to the perfection.
Audio chain must be simple, having few parts, low transistor count, all them class A biased, deeply class A biased.... hot....very hot if your environment can hold the heat...of course you can have 100 sub circuits around... auxiliary circuits, alike CCS, current sinks, regulators, stabilizers, mirrors....but keep the audio chain clean and simple.
It is good that you want to try and research amplifiers... a long road in front of you.... enjoy and be happy.
regards,
Carlos
I agree, there is an alure to the sound of class A that I find so relaxing and alive... Strangley enough it sometimes feels that some class AB amps give more details (probably just a product of harmonics), actualy now that I started I do not know how to put it in words... erm... lets say my AB amps does better with complex music, and the class A amps I tried gave the best performances with solo singers, quartets etc... less complex more organic music...
I am working up towards building Nelson's F5 now, according to reviews this seems to also add the good of a quality class AB amp to a class A topology.
At the end of the day, I think the DIY croud and equipment can be seperated into two groups... High efficiency speakers, with low powered, high quality amps, and low efficiency powered by higher powered amps...
I have yet to hear a high powered amp that can provide the truthfull reproduction of a quality high efficiency system.
I think anyone operating outside of these two groups, or mixing them, has a system that performs at less than the sum of the parts.
Of course I have been doing this for a relatively short time and one could say I have not sampled enough to get to any kind of statistical conclusion, but so far this holds water for me...
And of course I am quick to change positions in the light of new data or experience...
I am working up towards building Nelson's F5 now, according to reviews this seems to also add the good of a quality class AB amp to a class A topology.
At the end of the day, I think the DIY croud and equipment can be seperated into two groups... High efficiency speakers, with low powered, high quality amps, and low efficiency powered by higher powered amps...
I have yet to hear a high powered amp that can provide the truthfull reproduction of a quality high efficiency system.
I think anyone operating outside of these two groups, or mixing them, has a system that performs at less than the sum of the parts.
Of course I have been doing this for a relatively short time and one could say I have not sampled enough to get to any kind of statistical conclusion, but so far this holds water for me...
And of course I am quick to change positions in the light of new data or experience...
Dear Dx and all,
Thanks for vlble info. There are lots to be consider about the sond quality. single ended sonically good but efcncy is bad other hand push pull sonically not well but efcncy is good. yes its run very hot.
i have question about the vocal to DX, i find class A vocal is more open then AB and some high frq is more realistic (life) then AB. i listen carefully and fund this.
why class Ab music is not very open then A? dose it cause by crossover distortion?
Thanks for vlble info. There are lots to be consider about the sond quality. single ended sonically good but efcncy is bad other hand push pull sonically not well but efcncy is good. yes its run very hot.
i have question about the vocal to DX, i find class A vocal is more open then AB and some high frq is more realistic (life) then AB. i listen carefully and fund this.
why class Ab music is not very open then A? dose it cause by crossover distortion?
My experience with the DX Precision one did show great promise in haveing huge dynamic headroom, in fact it did some things better than any of the amps that came before it in the series, but overall I was still not sold.
So I concur that the 3rd option you mentioned is used, but I think mostly in grossly mismatched speakers and amps as is common under the crowd that only looks at how many zeros there are before the big W, when makeing purchasing choices.
I have limited experience / exposure to these setups so I will allow that the possibility exists...
I think to ge back to the topic, he needs to make a desicion on the full system, the efficiency of his current drivers, wether he intends keeping them in future etc...
So I concur that the 3rd option you mentioned is used, but I think mostly in grossly mismatched speakers and amps as is common under the crowd that only looks at how many zeros there are before the big W, when makeing purchasing choices.
I have limited experience / exposure to these setups so I will allow that the possibility exists...
I think to ge back to the topic, he needs to make a desicion on the full system, the efficiency of his current drivers, wether he intends keeping them in future etc...
what those amplifier desiners say
and also those listen to them is:
10 Watt pure Class A
will equal or beat a very good ~100 Watt Class AB
.. in sound quality
those small watts, first watts are very important
those watt from 0-2 Watt RMS output, is where most sound is
so, a good designer will put 100% quality into those First Watts
www.firstwatt.com
Even in Class AB first watts are as important
those higher peaks into 50-100 watts are not many
when listening as a normal person
Nordic,
is point to a matter: Sensitivity SPL of speakers
We can not talk of good or bad amplifier, if not tell what impedance and SPL we have connected.
4 Ohm vs. 8 Ohm .. substancial difference
88 dB vs. 93 dB ... real substancial difference!
and also those listen to them is:
10 Watt pure Class A
will equal or beat a very good ~100 Watt Class AB
.. in sound quality
those small watts, first watts are very important
those watt from 0-2 Watt RMS output, is where most sound is
so, a good designer will put 100% quality into those First Watts
www.firstwatt.com
Even in Class AB first watts are as important
those higher peaks into 50-100 watts are not many
when listening as a normal person
Nordic,
is point to a matter: Sensitivity SPL of speakers
We can not talk of good or bad amplifier, if not tell what impedance and SPL we have connected.
4 Ohm vs. 8 Ohm .. substancial difference
88 dB vs. 93 dB ... real substancial difference!
It is not necessary that class AB amp sounds worse on 1'st watts than class A one. Usually current amplifying devices connected as emitter followers are biased using stable bias voltage. As the result, slight variations of signal and in load impedance cause small bias shift that lead to significant current variations. Also, usual thermal feedback is represented by a complex function that has more than one pole with stability problems resulting in non-stable bias current. Keep in mind character of devices you use and make them happy; arrange topology for used devices instead of trying to fit devices to wrong but many times repeated and described in literature topology: this is the simple secret of gorgeous sound reproduction.
space2000,
you are a musician playing the piano and the violin, familiar with the harmonics of the acoustic instruments, which you want to get reproduced, right? If so, you should avoid topological finesses like push-pull and GNF. Any approach dictated by demands on efficiency and high power handling costs sound quality.
you are a musician playing the piano and the violin, familiar with the harmonics of the acoustic instruments, which you want to get reproduced, right? If so, you should avoid topological finesses like push-pull and GNF. Any approach dictated by demands on efficiency and high power handling costs sound quality.
Wavebourn said:
IMHO, the only way to achieve this with an amp correctly designed in classB (D Self definition) is to put may output devices in parallel.
For an output stage made with BJT's, you can get only 0.6W rms for one peers if you use Re=0R1, so you need at least three peers for 2W class A.
The max classB obtained with this OPS topology is limited only by dimensions of the heatsink.
My RMI-FC100 amp give 2W classA and 80W classB and the sound in glamorous at the normal listening levels (91dB SPL speakers, 22m2 room area)
Cheers,
Mihai
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.