|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: Re: Aksa compared to Leach Amp ? posted by djk on November 08, 2002 at 18:15:19:
OK djk, you're the guru.I must be confused as to what D13, D14, R11 and Q5 are doing on the Leach Amp Version 4.5. I thought cascode Q5 was part of the ccs. Maybe we are talking about different versions of the Leach and AKSA circuit diagrams.
Anyway, I suggest anyone following this thread should have a look for themselves and come up with their own conclusions.
Follow Ups:
"OK djk, you're the guru."Nawww, I just really like these things.
"I must be confused as to what D13, D14, R11 and Q5 are doing on the Leach Amp Version 4.5. I thought cascode Q5 was part of the ccs. Maybe we are talking about different versions of the Leach and AKSA circuit diagrams."
Q5 is the cascode, you can remove it if you like.
http://users.ece.gatech.edu/~mleach/lowtim/graphics/ckt.pdf
R11 is the load resistor for the diff pair. D13,14 with C2,4 provide a stable and clean supply for R15 which provides the current for the long tailed pair on the negative rail. This supply also biases the cascode on the positive rail (if you want it). A look at how simple it can be (schematic on page 2):
http://users.ece.gatech.edu/~mleach/papers/lowtim/feb76feb77articles.pdf
With modern high gain transistors you could delete the pre-drivers (Q8, 9, and R26, 27), and adjust the bias accordingly.
Another variation:
https://www.onsemi.com/pub/Collateral/AN1308-D.PDF
Thanks djk for the further information, the ccs issue makes a little more sense to me now. Looks like I can only spot a ccs when it is drawn like the text books. :)Looking back on my original post I should have said this....
"I haven't heard a Leach amp so I can't comment how they compare sonically, but technically the design philosophies are different. The Leach amp version 4.5 being a more complex design, while the AKSA 55 version 1.6 has been designed using the minimalist approach. Some people say the extra components have a negative impact on the sonics while many notable amp designers believe the impact to be indetectable. I don't know?"
vladimir original question was: "Is it (the ASKA) comparable or is Leach a superior dessign?"
On the technical level, I suppose what I was trying to point out is the people who like more complex amps, like the Leach version 4.5, may not like minimalist amps like the AKSA and visa versa. So to answer vladmir's question is: it's just a matter of opinion.
Do you think that is a valid statement?
How's that for an answer?Let's look at the differences here.
If you build the original low TIM amp without the pre-drivers, the bias and output stage will be identical to the AKSA. The front end will have three more transistors.
A single transistor will have less noise than a diff pair will. My VanAlstine MOSFET 120 was a good example of this minimalist design philosophy. It makes the AKSA look complex.
But a dual diff pair will have less noise than a single transistor and is better able to reject power supply noise.
The second stage of voltage gain, sometimes called the Vas, is just as important. The big problem here is that a single ended stage slews faster in one direction than the other, this makes for poor linearity. Most amplifiers use an active current source here and this looks good on paper, but doesn't always sound the best. McIntosh has used a simple resistor here without even a bootstrap capacitor, my MC2120 sounds great. They get around the problems by using a regulated supply for the diff pair and a very high voltage supply with a very large value resistor to go with the single Vas transistor.
If you use a dual diff input you can drive a second Vas transistor and skip the bootstrap cap and not have to worry about fancy supplies.
Using off-the-shelf unmatched transistors and no feedback at all, the Leach topology is capable of about 0.5%THD and 40Khz bandwidth. If you match the transistors you can get that down to 0.05%THD, if you think that matters.
With a modest amount of feedback applied to an already good working design it only gets better.
The added cost; about $0.50, and complexity; three small plastic transistors, are well worth it for the sonics (IMO).
If you use the published board layout you can always add the short circuit protection later if you find you need it.
If you need 500W at 2 ohms for PA all you need to do is add one more pair of outputs and build it as published. I've never heard a better sounding amplifier for PA.
If you think you need more power there is always the Super Leach.
djk,Are you saying that the Leach and the AKSA use the same basic topology?
Regards,
They both use the JBL 'T' output circuit, a fully compementary symmetry output stage.Most amps do these days.
The big difference is the front end, the voltage gain stages. The Leach is symmetrical, like your left hand is a mirror of your right hand. The AKSA is missing one hand (so to speak). The other hand in the AKSA is replaced by a bootstrap capacitor current source. This would be similar to putting a weight onto the stump of the missing hand to balance the body so as to keep moving in a straight line when walking.
The symmetrical front end requires three more inexpensive transistors, a total of about $0.50 extra.
DJK,A colorful analogy - like counting one's friends on a hand of amputated fingers!!
The symmetry argument concerning amplifiers is not, in my view, too relevant, although it is pleasing as all humans love symmetry.
Asymmetrical treatment of positive and negative half cycles leads to even order harmonic distortion of the fundamental, which is primarily musical, at least to the fourth harmonic. Symmetrical distortion of the half cycles, while far harder to spot and arguably of lower total magnitude in a fully symmetrical design like the Leach, leads to identical distortions on each half cycle which defines as odd order harmonic distortion of the fundamental. This is not so musical, and by the 5th harmonic, actually quite objectionable. It is essentially this order of distortion which creates so many nasties in the sonic presentation of solid state amplifiers, though tubes can do it too.
I hope my post does not offend.
Cheers,
This post is made possible by the generous support of people like you and our sponsors: