Guys, there are good caps for cheap, and an example is Philips/BC, so why use those lite green Lelon thinggies? 
This is not a critic to Greg, so please don't take me wrong but I'm just curious.
Because IMO and IME an amp (or any other device) that is completely insensitive to the PSU quality simply doesn't exist, whatever the PSRR figure it has.
This is not a critic to Greg, so please don't take me wrong but I'm just curious.
Because IMO and IME an amp (or any other device) that is completely insensitive to the PSU quality simply doesn't exist, whatever the PSRR figure it has.
jam said:
Jam I hope you are pulling my leg.
I thought I quoted what the difference were and to the contrary if you have to start talking about the whole list your amp is not in the hunt.
That was the point of the post in case you missed it (joke),
If an amp can disappear its the winner because there's nothing to talk about. I'm sure Nelson would appreciate my post...my last review he printed out and handed around the office as required reading so I think I'm on the money!
No I have not tried the bootstrapping mod at this stage.
I was expecting something more elaborate, so I got quite disappointed when I saw the schematic:
It speaks for itself.
Just consider that the gates in this circuit are current-driven directly from the LTP and that MOSFET input and reverse transfer capacitances change by an order of magnitude depending on Vds across the amplifier output voltage swing. Also consider that, in any MOSFET, Id is proportional to Vgs^2, thus requiring the highest and most non-linear gate drive AC currents at the lowest Id levels (during low level recording details) and vice-versa. Now imagine what happens when reactive loads are driven and zero-current transitions don't happen with the same Vds, gate capacitance or gate threshold where bias was adjusted... The battered input LTP is left with the chore of compensating for all this mess.
An externally hosted image should be here but it was not working when we last tested it.
It speaks for itself.
Just consider that the gates in this circuit are current-driven directly from the LTP and that MOSFET input and reverse transfer capacitances change by an order of magnitude depending on Vds across the amplifier output voltage swing. Also consider that, in any MOSFET, Id is proportional to Vgs^2, thus requiring the highest and most non-linear gate drive AC currents at the lowest Id levels (during low level recording details) and vice-versa. Now imagine what happens when reactive loads are driven and zero-current transitions don't happen with the same Vds, gate capacitance or gate threshold where bias was adjusted... The battered input LTP is left with the chore of compensating for all this mess.
Eva is already well known as a high experienced EE
I think Pinkmouse will not move her text.
regards,
Carlos
I think Pinkmouse will not move her text.
regards,
Carlos
Eva,
I don't think your criticisms are fair. The Gates are voltage driven from an emitter follower, as TubeDude remarks, and this EF is ideally loaded by a bootstrapped, shared resistor; further, the LTP is also bootstrapped and feeding the base of an EF which is an exceptionally high impedance. There is elegant economy here, far from the simplistic design you suggest.
This is a clever circuit; I cannot see your criticisms. OLG in the LTP is high, probably around 60-70dB, and a further roughly 30dB of gain comes from the common source configuration of the mosfet. I see huge capacitive and non-linearity issues in the mosfets, but this is true of any mosfet drive. The fact that this circuit dispenses with a formal VAS is, I think, its strength.
One concern I have relates to the LTP bootstrap cap. It holds about 1.3V across it, judging from the diagram, and by E = 1/2CVexp2, the energy content is small for bootstrapping. It would need to be very large, around 470uF, to do its job.
Feedback factor would be so high that distortion becomes almost immeasurable despite the non-linearities. An area of concern for any amp is always reactive loads, but I suspect the EF so effectively isolates the two voltage amplifying stages that even with difficult loads the phase shift would be uncritical.
A final consideration is switching of the outputs. Mosfets are susceptible to self-oscillation, particularly in Class AB, and I see no special provision to prevent short term instability at crossover. This is a concept circuit only; there will undoubtedly be some additional componentry to ameliorate this problem.
In any event, you have not heard this circuit, and neither have I. It would be premature to judge it until then......
Cheers,
Hugh
I don't think your criticisms are fair. The Gates are voltage driven from an emitter follower, as TubeDude remarks, and this EF is ideally loaded by a bootstrapped, shared resistor; further, the LTP is also bootstrapped and feeding the base of an EF which is an exceptionally high impedance. There is elegant economy here, far from the simplistic design you suggest.
This is a clever circuit; I cannot see your criticisms. OLG in the LTP is high, probably around 60-70dB, and a further roughly 30dB of gain comes from the common source configuration of the mosfet. I see huge capacitive and non-linearity issues in the mosfets, but this is true of any mosfet drive. The fact that this circuit dispenses with a formal VAS is, I think, its strength.
One concern I have relates to the LTP bootstrap cap. It holds about 1.3V across it, judging from the diagram, and by E = 1/2CVexp2, the energy content is small for bootstrapping. It would need to be very large, around 470uF, to do its job.
Feedback factor would be so high that distortion becomes almost immeasurable despite the non-linearities. An area of concern for any amp is always reactive loads, but I suspect the EF so effectively isolates the two voltage amplifying stages that even with difficult loads the phase shift would be uncritical.
A final consideration is switching of the outputs. Mosfets are susceptible to self-oscillation, particularly in Class AB, and I see no special provision to prevent short term instability at crossover. This is a concept circuit only; there will undoubtedly be some additional componentry to ameliorate this problem.
In any event, you have not heard this circuit, and neither have I. It would be premature to judge it until then......
Cheers,
Hugh
A current driven emitter follower produces a current output, there is no magic here. The base resistor of the "emitter follower" is intentionally bootstrapped to simulate a current source and to boost apparent open loop gain at the expense of an almost infinite base drive impedance and current drive to the gates.
I would expect somebody called tube dude to make such a mistake by accident, but it's quite disappointing to see such a degree of misunderstanding of the bipolar transistor behaviour from you, AKSA
In fact, I'm just going to replace a couple of current sources that drive emitter followers in one of my circuits by emitter followers that self-bootstrap their bases, because this is going to produce the same almost infinite (very high) output impedance of the current source with a lower part count.
By the way, I won't discuss sound because there is already a lot of people that does that quite well. I'm strictly focused on inner circuit workings and circuit design instead.
I would expect somebody called tube dude to make such a mistake by accident, but it's quite disappointing to see such a degree of misunderstanding of the bipolar transistor behaviour from you, AKSA
In fact, I'm just going to replace a couple of current sources that drive emitter followers in one of my circuits by emitter followers that self-bootstrap their bases, because this is going to produce the same almost infinite (very high) output impedance of the current source with a lower part count.
By the way, I won't discuss sound because there is already a lot of people that does that quite well. I'm strictly focused on inner circuit workings and circuit design instead.
AKSA said:
Hugh,
A very valid point..the world is full of arm chair engineers, interlectuals and skeptics but they are not the ones that keeping the wheels turning! Pity we do not have more Leonardo's..
At this point it would be useful to read from others who have actually built the SKA and their impressions.
Eva said:
That's all very nicely put in a confusing and introverted way - for a non tach savvy such as myself.
But were back to the whole schematic vs. sound debate again aren't we?
I've been listening to the amps for a week now and I still really like what I hear. Best amp I've had in my system so far.
I guess we needed something to offset all the positive comments and for that I appreciate your post Eva.
ShinOBIWAN said:
I know the tale of the naked king quite well. There are thousands of megabytes of it in the forum archives (it's told again and again everytime a new kit is put into the market 😀 )
Also I'm curious about where the "30dB" MOSFET "gain" figure that AKSA quoted comes from. MOSFET are transconductance devices, they produce an output current proportional to the square root of Vgs. How a gain could be defined there if the input voltage follows a quadratic law and the output voltage is proportional to the load impedance at that particular frequency?
Eva,
I cannot imagine why you would be disappointed. You have not met me, you've not heard an AKSA, by your own admission you focus solely on the engineering.
It's been clear for some time you have a engineering education in this area, although your definition of 'current driven' is moot indeed. A gate is voltage driven too, you know, this is the root of the term 'transconductance'. But the vaulting arrogance implicit in your quote points to astonishing hubris which blinds you to the perceptions of others.
In contrast, I'm not disappointed at all!
Hugh
Eva said:
Yes , the bootstrap effectively increase the input impedance driving the follower , but the same exist in 99% of amp designs , that use a traditional common emitter VAS stage as the output of this stage is also a CCS.
And nobody claim that a bipolar follower after a VAS stage , drives de output Mosfets in current mode...I'm afraid..
Anyway R11 and R12 , have a virtual ground in their interconnection point , also lowering the driving impedance of the mosfets...
Re: I perceive here that there are a very sensitive problem
Very true.
Its entirely personal and was the reason I tried to avoid a direct comparison initially. It doesn't serve any real purpose for anyone but the individual doing the comparison.
I suspect the biggest flaw's aren't actually in amplifier's in the ilk of the AKSA, SKA, Aleph, Symasym etc. but in the perception and preference of sound that varies between individuals. So its we who are flawed because our own hearing-brain mechanisms are so inaccurate and misunderstood that trying to quantify amplifiers is impossible to all but the beholder.
destroyer X said:
Very true.
Its entirely personal and was the reason I tried to avoid a direct comparison initially. It doesn't serve any real purpose for anyone but the individual doing the comparison.
I suspect the biggest flaw's aren't actually in amplifier's in the ilk of the AKSA, SKA, Aleph, Symasym etc. but in the perception and preference of sound that varies between individuals. So its we who are flawed because our own hearing-brain mechanisms are so inaccurate and misunderstood that trying to quantify amplifiers is impossible to all but the beholder.
Eva said:
I can only guess that you took offense to my post. It does work both ways 😉
It kinda makes your technical analysis mute when you've got a few folks speaking from a point of view of actually hearing and really enjoying it, so I can see where the spite comes from. Since we listen to amps rather than listen to folks that debate the technicalities of amps, I saw your post as irrelevant. No offense intended.
I wasn't the first to tout the SKA BTW, I jumped in a month or two later 🙂 I'm far too tight to put money down on something unproven so I waited for a couple of other opinions. Thanks to Macka and Hubert & friend I decided to take the plunge.
Again sorry if I put hearing above over-analysis.
Tube_Dude said:
That's because the output of the VAS is either capacitively loaded to ground or miller compensated, thus producing an output impedance inversely proportional to frequency. Even without such capacitors, the inherent modulation of the hFE and Vbe of the VAS transistor across the output voltage swing simulate a load that decreases output impedance. In SKA there is not such a compensation, so we can talk about current gate drive in all the frequency range (there is not other way to get enough open loop gain with just two transistors, as loading reduces gain).
The funny part comes when we are complementarily current-driving two capacitances whose values change wildly and are usually an order of magnitude apart. It produces sound by accident (have you ever asked why your circuit still produces sound when you have just noticed that you have put one transistor backwards).
Also, you have said it: It works much in the same way as a traditional VAS, but with some cosmetical changes (remember all those J50/K135 kits sold in the 1980s?). It's a circuit aimed at low parts count at the expense of performance. By the way, technically these kits were far superior because at least the source follower connection ensured symmetric Vgs drive and consistent bias, while the SKA features random uncontrolled dynamic bias during normal operation (except when it's biased into 4A class A or so).
Re...
So when did parts-count dictate performance?
Look at the Aleph, start thinking ZEN...😎
Arne K
Eva said:
So when did parts-count dictate performance?
Look at the Aleph, start thinking ZEN...😎
Arne K
Where did MUSIC go???
I few points about recent posts here.
NO ONE has really talked about listening to MUSIC!! All the technical points are totally irrelevant.
The single best comment that means something is that the Pass amp "disappeared" !
Compare the numbers between the PASS and the SKA!! So what do the numbers tell you?? NOTHING!!!
Macka, if you want to scrape your jaw off the floor and you like the Pass, go put a really good 845 or 211 SET amp in front of your same test setup!!!
This is also the ultimate living proof that most numbers are total BS when it comes to enjoying MUSIC!!
Sorry but it is easy to PROVE NONE of you can hear 0.1% 2nd harmonic distortion BUT what does 0.1% 7th harmonic sound like?
Stop being so naive about technology!!!
cheers
I few points about recent posts here.
NO ONE has really talked about listening to MUSIC!! All the technical points are totally irrelevant.
The single best comment that means something is that the Pass amp "disappeared" !
Compare the numbers between the PASS and the SKA!! So what do the numbers tell you?? NOTHING!!!
Macka, if you want to scrape your jaw off the floor and you like the Pass, go put a really good 845 or 211 SET amp in front of your same test setup!!!
This is also the ultimate living proof that most numbers are total BS when it comes to enjoying MUSIC!!
Sorry but it is easy to PROVE NONE of you can hear 0.1% 2nd harmonic distortion BUT what does 0.1% 7th harmonic sound like?
Stop being so naive about technology!!!
cheers
Re: Re...
Buy a car without wheels. I'm sure that the part reduction will improve performance (it can still move if you drag it) 😀
Cobra2 said:
Buy a car without wheels. I'm sure that the part reduction will improve performance (it can still move if you drag it) 😀
Re: Re: Re...
In a thread about opinions on listening impressions:
What about such an opinion stripped of the subjectivity of hearing and that then focus's soley on the technical aspects (it still carries weight but only if you drag it) 😀
Eva said:
In a thread about opinions on listening impressions:
What about such an opinion stripped of the subjectivity of hearing and that then focus's soley on the technical aspects (it still carries weight but only if you drag it) 😀
Re: Where did MUSIC go???
"Listening impressions" are pseudo random and self induced. Some days I even enjoy square waves while others I can't stand any sound 😀 (I happen to build crossovers by ear and I suffer a lot these effects).
The only way to get good "listening impressions" is to keep building/buying new stuff every year and reading good reviews about it. It has nothing to do with the gear itself. A square wave geneator with good reviews will sound nice to most people 😀
Well, I'm sorry if I have been too harsh, and I hope to have made somebody laugh or reconsider things a bit. I'll move back to more technical lands where performance is some non-subjective repeatable thing that you can see in the screen of an oscilloscope. (I was just on holidays that night 😀😀😀 )
hifryer said:
"Listening impressions" are pseudo random and self induced. Some days I even enjoy square waves while others I can't stand any sound 😀 (I happen to build crossovers by ear and I suffer a lot these effects).
The only way to get good "listening impressions" is to keep building/buying new stuff every year and reading good reviews about it. It has nothing to do with the gear itself. A square wave geneator with good reviews will sound nice to most people 😀
Well, I'm sorry if I have been too harsh, and I hope to have made somebody laugh or reconsider things a bit. I'll move back to more technical lands where performance is some non-subjective repeatable thing that you can see in the screen of an oscilloscope. (I was just on holidays that night 😀😀😀 )
- Status
- Not open for further replies.