That problem was from the smps of the computer and it appear in the measurement chain.
The theory, as I understand it, is that a fully symmetrical and perfectly matched balanced system should cancel all even order harmonics.
Much the same effect is found in BTL coupled designs, where 2 individual amplifiers play the same but inverted signal.
\\\Jens
Are they about tubes, or also transistor based transformer coupled amplifiers ?
And why transformers belong only to history ?
.
Almost certainly to provide symmetrical drive to the OPS.
When you drive heavy loads, the balanced, capable drive is useful. As I remarked earlier, with blameless drive you need to consider this issue. One option is to use an EF3 and the other to ensure the VAS current is sufficiently high - I've seen a few designs around on the web with 2-3 mA VAS standing current and an EF2. I've simmed balanced designs that show sub 10'ppm into 3 ohms.
The simulations are very accurate if u know what you do. At SARA, my design,push-pull EF-VAS the simulated result of the THD 1khz 50W 8ohm it was 0.000235% . The real THD was 0.00025%. Almost the same results at 20khz between sims and real.
At the new future project based on the SARA,with an circuit update, show me a simulated THD = 0.000040 % at 1khz in the same conditions.
The biggest problem under the 0.001% it is the parasitic components. Magnetic field of the cables and components can ruin a very good schematic. Here is the real problem to solve.
At the new future project based on the SARA,with an circuit update, show me a simulated THD = 0.000040 % at 1khz in the same conditions.
The biggest problem under the 0.001% it is the parasitic components. Magnetic field of the cables and components can ruin a very good schematic. Here is the real problem to solve.
Impressive numbers. How do you measure such low levels IRL? The AP I have is not capable of these levels.
Mine is about 5ppm... so it will be difficult to verify greierasul's amp on my setup 🙁
Does anyone have descriptions on how to extend the range?
I notice that the noise floor on the generator comes up with higher levels - do you use the internal generator or an external one?
Best regards
Jens
The important words being "a fully symmetrical and perfectly matched balanced system" which is going to be pretty hard to build.
Sansui's explanation is, in their words"...the function of this section is complete push-pull operation by which even harmonic distortion is extremely decreased."
Unfortunately life ain't that simple. As Chapter 8 of APAD6 shows with pitiless clarity, a push-pull VAS, implemented in any of four different ways, doesn't extremely decrease anything, but gives MORE second-harmonic distortion than a single-ended VAS.
The best example of push-pull action genuinely cancelling second harmonic distortion must be a differential input stage with properly balanced collector currents.
I really don't know why you keep saying this. At a minimum, APAD6 describes five kinds of VAS and five kinds of output stage. (Class-B, Class-AB, Class-A, Class-G, Crossover Displacement) which by my reckoning gives 25 different amplifier configurations. How many more do you want?
I think the only major missing topic is CFA, and since I see no need for slew-rates of four hundred volts per femtosecond, it didn't make the cut. Books are not infinitely expandable.
How about some details, like model numbers? How about linking to the Stereophile reviews, (all on-line, I believe) so we can see for ourselves?
Transistors. I don't do valves.
And why transformers belong only to history ?
.
Briefly: big, heavy, expensive, non-linear, poor freq response, subject to magnetic interference. Otherwise fine. 🙂
It is interesting to note that what you call class S, the RF world calls a Doherty amplifier, and that in that space they are very popular with the cell phone BTS crowd (Narrow band stuff so playing transmission line games for the load pull is easy).
Some commentry on continious Vs peak power requirements would have been good in the section on thermal design, in particular I often find that (In the sound reinforcement game) an amp thermally limited to maybe a couple of hunded watts but with sufficient voltage and SOA headroom to deliver an extra 10dB at a 5% duty cycle is a MUCH more useful animal then a amp rated for double the power but which runs out of voltage swing at that power.
This a prehaps a good argument for class G or some form of envelope tracking power supply.
I am well aware that this is a poor fit for the single number 'RMS Power' that everyone in marketing likes to quote.
Regards, Dan.
Some commentry on continious Vs peak power requirements would have been good in the section on thermal design, in particular I often find that (In the sound reinforcement game) an amp thermally limited to maybe a couple of hunded watts but with sufficient voltage and SOA headroom to deliver an extra 10dB at a 5% duty cycle is a MUCH more useful animal then a amp rated for double the power but which runs out of voltage swing at that power.
This a prehaps a good argument for class G or some form of envelope tracking power supply.
I am well aware that this is a poor fit for the single number 'RMS Power' that everyone in marketing likes to quote.
Regards, Dan.
Hi Doug,
There are a vast amounts of topics covered in the book, even though Cfa didn't make the cut. Given some time to think about it and my experience in starting out with the blameless years back it is definately a universal topology. Anyone can get it to work well, it's easy to build and requires little more than matching the ltp beta to get good results at minimum.
I do love the Cfa push pull for my home system, I've never heard better grain free natural sounding mids to high resolution than with Cfa, but admittedly it takes much more care in matching to get these results and low dc offset. This level of attention to detail is prohibitive to making money for the average consumer audio manufacturer. The Lin Bailey, blameless style will generally give consistent large production spread success with unmatched parts other than the ltp and still sound good and remain easy for anyone who can use a solder iron and dmm to build.
Colin
There are a vast amounts of topics covered in the book, even though Cfa didn't make the cut. Given some time to think about it and my experience in starting out with the blameless years back it is definately a universal topology. Anyone can get it to work well, it's easy to build and requires little more than matching the ltp beta to get good results at minimum.
I do love the Cfa push pull for my home system, I've never heard better grain free natural sounding mids to high resolution than with Cfa, but admittedly it takes much more care in matching to get these results and low dc offset. This level of attention to detail is prohibitive to making money for the average consumer audio manufacturer. The Lin Bailey, blameless style will generally give consistent large production spread success with unmatched parts other than the ltp and still sound good and remain easy for anyone who can use a solder iron and dmm to build.
Colin
Douglas, you are quite capable of doing this research yourself. Just type 'Stereophile' into Google and go to the reviews section.
And no, 25 options is not enough if you insist that some types are sub par when patently they have been used in the industry for decades with success (means they work physically and they work commercially).
CFA is not just about slew rate, though if you insist that non-symmetrical is a virtue, I doubt I or anyone else will convince you on CFA.
Since this is your thread, and I don't want to seem argumentative, I will step out after this post.
Your books have been invaluable as teaching tools and calling to task the subjectivist fraternity, but ease up a bit and accept that there are ways of doing things that may not fit with your world view. That does not invalidate them.
🙂
And no, 25 options is not enough if you insist that some types are sub par when patently they have been used in the industry for decades with success (means they work physically and they work commercially).
CFA is not just about slew rate, though if you insist that non-symmetrical is a virtue, I doubt I or anyone else will convince you on CFA.
Since this is your thread, and I don't want to seem argumentative, I will step out after this post.
Your books have been invaluable as teaching tools and calling to task the subjectivist fraternity, but ease up a bit and accept that there are ways of doing things that may not fit with your world view. That does not invalidate them.
🙂
As part of todays designwork also includes thinking about energy consumption, I think a "bang for the buck" approach would be nice.
E.g. how do we make the best sounding amp without spending 100W idle? Bearing in mind that the average domestic amp only puts out a few mW 90% at the time.
Can you please clarify if the SE vs. PP evaluation is made using the same Bias current in VAS.
I know that the the bulk of the DIY audio members might not care, but this information is invaluably for us working with this stuff every day.
Thanks
Jens
Last edited:
Have you designed transformer amplifiers ? Are there data available on them ? Which transformers have you used ?
.
Doug,
Nobody's questioning the value of your thorough analysis of push-pull VAS problem. In fact I decided to buy 6th edition just to read that part. (I shall wait till Bob's second edition to buy it together because I want to save some money on postage and packing).
I am just saying that excellent sounding amplifiers are made using push pull VAS and that push pull VAS probably has some advantages too, even if they are not obvious from your research.
We should not automatically disregard it especially if user's reactions are, in fact, quite good.
Nobody's questioning the value of your thorough analysis of push-pull VAS problem. In fact I decided to buy 6th edition just to read that part. (I shall wait till Bob's second edition to buy it together because I want to save some money on postage and packing).
I am just saying that excellent sounding amplifiers are made using push pull VAS and that push pull VAS probably has some advantages too, even if they are not obvious from your research.
We should not automatically disregard it especially if user's reactions are, in fact, quite good.
- Status
- Not open for further replies.