Wahab,
Not half so ecologically incorrect as driving your car less than 2kms each day.....how many Kw-hours in a litre of gasoline or diesel? It's around 8.9Kw-hours per litre, and that would run a very inefficient Class Amp for a long, long time. Profligate use of automobiles uses FAR more energy than running a politically incorrect Class A amplifier, yet no one comments on our outrageous mobility.
Hugh
Not half so ecologically incorrect as driving your car less than 2kms each day.....how many Kw-hours in a litre of gasoline or diesel? It's around 8.9Kw-hours per litre, and that would run a very inefficient Class Amp for a long, long time. Profligate use of automobiles uses FAR more energy than running a politically incorrect Class A amplifier, yet no one comments on our outrageous mobility.
Hugh
Last edited:
Hi wahab,
I don't normally consider 0.05% THD to be very good, even at 20 kHz. With distortion levels in that range, the only way it is OK is if the spectrum of the distortion products is really benign.
Cheers,
Bob
as an exemple; let s take a design i consider to be a total
failure, nelson s SON OF zen...
each channel draw 375 w at "iddle" , thus heading for
about 1 KW for both channels with the supply losses..
cost here would be about 0.1 euro/hour..
since i let my amp be on at least 8 hours a day, even i listen to
nothing, it s about 0.8 euro/day or 24 euros/month just for
the amplifier...
by the way, this is the cost of 24 litres of gasoil in europe, where
it is heavily taxed, wich is enough for a 400KM ride,almost the average
kms of a city car in a month..
wahab
agree that it s not particularly low although, as mentionned
it was measured in a 200KHZ bandwith...
sure that if it was restricted to a few harmonics (since some pretend that
we don t hear more than 20khz) , the figure would be better, but it
would be intellectually speaking, dishonesty..
also, the harmonics voltage where simply linearly added, as if they
were in phase , although it s not rigourous, i prefer this methodology that give a clue about the worst case figure..
of course, enhancing the input stage gain, thd can be reduced
by almost 15db, but stability would suffer...
the requirement is that it must work without LR output filter
and whatever is the capacitive part of the load...
i don t know what is your definition of a benign spectrum,
but in this matter, there s no miracle...
although the harmonics goes lower as their rank is higher,
the products extend well outside the audio band..
i suppose that it s typical of any class AB amp...
regards,
wahab
Thanks to AKSA for reminding me that one journey to my lady friend (15miles each way in 2litre petrol car) is approximately the same pollution/energy as running KSA100 for about 60hrs or a pair of 27W F5 for 250hrs.
Maybe that's why I have 4 ClassA channels for audio and another 10channels of ClassAB.
But, will she let me see her less?
Maybe that's why I have 4 ClassA channels for audio and another 10channels of ClassAB.
But, will she let me see her less?
Wahab,
Bad example. Most Class A amps do not dissipate 1kW. Let us assume 300W, that's more like it, and just 8 euros a month.
This is perhaps three cups of good coffee a month? Given the pleasure of music, this is a trivial cost. Your ecological aspirations are not realistic, particularly when weighed against the huge mobility of most people in the western world in motor vehicles, which are around 22% thermally efficient at best. Diesel seldom exceeds about 35% - even the best French diesels.
There will always be a market for ecologically unsound choices. The V12 engine lives on - MB, Audi and BMW all make them, and Ferrari have made them since WW2. It is merely a matter of budget, not of guilt!
When we aim for efficiency, particularly where art is concerned, we usually lose something......
As for your arguments about the tragedy of wonderful contributors, there is always you. You have some very good circuits in your archives. You explain things well, perhaps you can make the forum more interesting, n'est pas?
Andrew,
The question is not whether you will see less of your lady friend, but rather more of her.....
Hugh
Bad example. Most Class A amps do not dissipate 1kW. Let us assume 300W, that's more like it, and just 8 euros a month.
This is perhaps three cups of good coffee a month? Given the pleasure of music, this is a trivial cost. Your ecological aspirations are not realistic, particularly when weighed against the huge mobility of most people in the western world in motor vehicles, which are around 22% thermally efficient at best. Diesel seldom exceeds about 35% - even the best French diesels.
There will always be a market for ecologically unsound choices. The V12 engine lives on - MB, Audi and BMW all make them, and Ferrari have made them since WW2. It is merely a matter of budget, not of guilt!
When we aim for efficiency, particularly where art is concerned, we usually lose something......
As for your arguments about the tragedy of wonderful contributors, there is always you. You have some very good circuits in your archives. You explain things well, perhaps you can make the forum more interesting, n'est pas?
Andrew,
The question is not whether you will see less of your lady friend, but rather more of her.....
Hugh
Last edited:
hugh,
i agree that the exemple is somewhat extreme, as this
amp is not exactly zen, rather a fat obese cat...
300 W TDP is still huge , 50W/channel seems to me an optimal
and also maximal value , since we don t need a lot of power
for home listening...
in fact, 10W rms output per channel is vastly enough..
i measured rarely more than 1W/channel at my habitual
listening levels, my speakers being 90dbm/W efficency..
as an indicator, typical consumation of a family of five
is about 60 euros/month , about 600KWH, that makes
833W as permanent consumation..
now, put the 300W TDP class A in this perspective..
if everybody was to use such amps, the households consumation
would rise by a whopping 30% , meaning about 15% more
electricity production...
of course, in third world countries, monthly consumation of
a household is badly lower, perhaps no more than 100KWH/month
in most cases...
i agree that the exemple is somewhat extreme, as this
amp is not exactly zen, rather a fat obese cat...
300 W TDP is still huge , 50W/channel seems to me an optimal
and also maximal value , since we don t need a lot of power
for home listening...
in fact, 10W rms output per channel is vastly enough..
i measured rarely more than 1W/channel at my habitual
listening levels, my speakers being 90dbm/W efficency..
as an indicator, typical consumation of a family of five
is about 60 euros/month , about 600KWH, that makes
833W as permanent consumation..
now, put the 300W TDP class A in this perspective..
if everybody was to use such amps, the households consumation
would rise by a whopping 30% , meaning about 15% more
electricity production...
of course, in third world countries, monthly consumation of
a household is badly lower, perhaps no more than 100KWH/month
in most cases...
Wahab,
I agree that 20 kHz THD should be measured in a 200 kHz measurement bandwidth, and that is how I do it. That is the basis on which I suggested that 0.05% THD20 is not that great. If the spectrum is very benign, like that of many tube amplifiers, then 0.05% THD20 is very good. For most solid state amplifiers that operate in class AB, it is usually not very good.
There is no requirement that amplifiers work without an output L-R network, and to do so involves compromises in risk. Even John Curl agrees that 1 uH is probably sonically transparent if the output coil is carefully implemented. No inexperienced designer on the planet should design without an output coil unless they have complete control over the kind of load the amplifier will ever see.
Cheers,
Bob
Carefully implemented ouput coil
Hello Bob
What do you mean by carefully implemented output coil is this a reference to the build quality, can you elaborate.
Regards
Arthur
Hello Bob
What do you mean by carefully implemented output coil is this a reference to the build quality, can you elaborate.
Regards
Arthur
I guess he means that it's designed in such a way that it fulfills the purpose for it being there i.e. preventing the amplifier from oscillating / ringing with any load and also damping the downstream ringing as much as possible with all loads.
Downstream ringing which would have no harmonic relation to the music is effectively adding noise into the signal and noise is what most of us are usually trying to reduce
Downstream ringing which would have no harmonic relation to the music is effectively adding noise into the signal and noise is what most of us are usually trying to reduce
Or maybe do what Naim did and sell the punters some special speaker cable to make up for the inductor
LOLTrev
Charles Hansen has developed an output circuit that does not need coat hanger inductors.
Charles Hansen from DIYaudio:
"As I have noted many time before, in my experience the best output stage is the "T-Circuit" (basically a triple emitter follower) developed by Bart Locanthi at JBL in 1966. Leach used it in his amplifiers (and has a copy of one of Locanthi's original articles on his website). I have used that output stage in all of my power amplifier designs for the last eight years. The only improvement I have been able to make is to stabilize it so that no output inductor or Zobel network is required for stability with a capacitive load.
Building an emitter-follower that is stable into capacitive loads without using an output inductor is definitely possible. Like I said, it took me over a month of full-time work. But maybe someone else out there is more clever and can come up with a solution more quickly. "
Charles Hansen from DIYaudio:
"As I have noted many time before, in my experience the best output stage is the "T-Circuit" (basically a triple emitter follower) developed by Bart Locanthi at JBL in 1966. Leach used it in his amplifiers (and has a copy of one of Locanthi's original articles on his website). I have used that output stage in all of my power amplifier designs for the last eight years. The only improvement I have been able to make is to stabilize it so that no output inductor or Zobel network is required for stability with a capacitive load.
Building an emitter-follower that is stable into capacitive loads without using an output inductor is definitely possible. Like I said, it took me over a month of full-time work. But maybe someone else out there is more clever and can come up with a solution more quickly. "
John Linsley hood also managed to avoid o/p inductors in his highly regarded mosfet power amp of 1985 which he also thought was worthy of note - so no o/p inductors seems like a worthy goal.
In that design he used an 8.2R & 0.1uF Zobel followed by a 0.22R o/p resistor.
But this is a DIY site so if we build an amp we can control what the load is going to be so we have a much easier task. It's only if we want to sell an amp that this issue becomes more critical.
For me, I concluded that it's best to have separate amps for bass below about 100 Hz which makes the problems even easier to deal with.
In that design he used an 8.2R & 0.1uF Zobel followed by a 0.22R o/p resistor.
But this is a DIY site so if we build an amp we can control what the load is going to be so we have a much easier task. It's only if we want to sell an amp that this issue becomes more critical.
For me, I concluded that it's best to have separate amps for bass below about 100 Hz which makes the problems even easier to deal with.
Hi Arthur,
The coil should be of air core design - definitely no ferrite or steel. It should be of adequate gauge and should not be wrapped around something like a reservoir capacitor (like someone actually did). The coil should be away from steel materials like a chassis, and the coil should be away from magnetic sources that can radiate into it (like rail lines) and away from things that its magnetic field can radiate into (like a feedback path or an input line).
I actually built an air-core toroid output coil one time. However, it was 3 uH; it is rather difficult to make an air core toroid that is only 1 uH.
Cheers,
Bob
These are good points that Charles has made, especially in regard to the T circuit. As I said, no inexperienced designer on the planet should build an amp without an output coil unless he has full control of the load to be driven (as in a powered loudspeaker where the amplifier is connected to a known loudspeaker). Charles obviously does not fall into that inexperienced category.
The feat that Charles has described is a difficult one, and more power to him if he has done that without seriously compromising other aspects of output stage performance and can really drive ANY capacitance (including low ESR). BTW, the coild guards against both local output stage parasitic oscillation and instability of the global negative feedback loop. Charles does not use global negative feedback, so his main challenge is to stabilize the Triple against local parasitic oscillations.
Cheers,
Bob
Yes, but not as good because its construction is more open and because there is no iron to more strongly guide the magnetic field lines. Nevertheless, there is no "open-end" where the field lines must escape.
Cheers,
Bob
For the record, none of the amps that I have made for the last 20 years, for Parasound, uses any output coil at all. Go figure! ;-) Still the coat hanger concept is a good one, in many ways, cheap, effective, and we all know that even wires made with coat hanger material cannot be heard in an ABX test, so why not?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.