KSTR said:Jan, you are disappointing me!
[snip]- Klaus
I'm sorry, I did try my best 😉
KSTR said:[snip]For an power amp output cap the AC situation is exactly the same as for PSU caps (note that the brigde and Xformer are "non-existent" for most of the AC mains cycle). Both are passive bridge legs, and are effectively in series with the load. When there is no feedback loop, then any series effects of the caps (and wiring etc) are fully effective. With FB, it depends on how well it is implemented (incl. from where the reference point is taken etc) to what amount the caps are still in the signal path.
- Klaus
Yes they are both in series with the load. But even without any feedback there is still a difference. Even without feedback, the amp output before going into a load series cap, is considered the output voltage. Any voltage across that series cap as a result of non-linearities in the cap is effectively in series with Vout, and thus directly modifies the speaker signal.
With the power supply caps, this is of course not the case. Any voltage across the cap as a result of the non-linearities modifies the supply voltage, NOT the output voltage.
Of course, that modified supply voltage will in some way leak through to Vout, but attenuated by the PSRR.
But you know all this. I think you're just pulling my leg.
Jan Didden
Reading this it suggests that any noise or modulation across the caps will appear directly across the load. However it doesn't. The supply's noise coupling is reduced by the relationship of the transistors or tubes to the load. And using a follower the supply's influence is much reduced compared to the voltage at the input to the follower. It's a second or third order issue. And using a full bridge it is further reduced since the return path isn't through the ground.
Essentially, your looking at two paths for this to get into the system, one through modulations to the ground or between the ground reference and the return for the load, the other through the power terminal on the transistor or tube. In a follower that is pretty small.
I would suggest an obsessive understanding of the current flows through the systems grounds would yield the most benefit, looking for ways for that noise to insert itself between ground reference points.
The article referenced on filament supplies highlighted the issue with the hidden ground on the supply pumping energy back across the measuring instrument. Using a balanced probe isolated the hidden current path and allowed the actual noise to be seen.
Essentially, your looking at two paths for this to get into the system, one through modulations to the ground or between the ground reference and the return for the load, the other through the power terminal on the transistor or tube. In a follower that is pretty small.
I would suggest an obsessive understanding of the current flows through the systems grounds would yield the most benefit, looking for ways for that noise to insert itself between ground reference points.
The article referenced on filament supplies highlighted the issue with the hidden ground on the supply pumping energy back across the measuring instrument. Using a balanced probe isolated the hidden current path and allowed the actual noise to be seen.
For better understanding, it is easy to put 100 ohm resistor into collector of an emitter follower. The output impedance of the follower is of course NOT 100 ohm, but several ohm, depending on the part, quiescent current etc. We cannot simply say that power supply is in series with the load.
1audio said:Reading this it suggests that any noise or modulation across the caps will appear directly across the load. However it doesn't. The supply's noise coupling is reduced by the relationship of the transistors or tubes to the load. And using a follower the supply's influence is much reduced compared to the voltage at the input to the follower. It's a second or third order issue. And using a full bridge it is further reduced since the return path isn't through the ground.
Essentially, your looking at two paths for this to get into the system, one through modulations to the ground or between the ground reference and the return for the load, the other through the power terminal on the transistor or tube. In a follower that is pretty small.
I would suggest an obsessive understanding of the current flows through the systems grounds would yield the most benefit, looking for ways for that noise to insert itself between ground reference points.
The article referenced on filament supplies highlighted the issue with the hidden ground on the supply pumping energy back across the measuring instrument. Using a balanced probe isolated the hidden current path and allowed the actual noise to be seen.
Fully agree. The first time I ran into the ground path issue was when doing the super reg article with Walt Jung in '96 or so. That grounding issue became the limiting factor, after the best opamp, best reference, etc was selected.
@Pavel: The statement was, that the load current also flows through the supply caps. That, of course, is true.
Jan Didden
I realize that a lot has been said about power supply interaction with amps and preamps here. One thing that surprises me is WHY power supply caps, especially in power amps, can take so long to 'break-in'. I can't be sure, but as my JC-1 power amp takes many hours to sound consistently the same, I have attributed this to the extra high quality Nichicon power supply caps that we use. I have been told that IF I would only change out the caps to Mallory power supply caps of the same value, I would not have this 'break-in' problem. Anyone else have any opinion on this?
janneman said:
Good point. I'll weasel out by saying that it is not 100% applicable to all people. I accept that the limit of where the difference cannot be heard anymore is different for different people.
However, when you take a large enough sample, you get close. In a few months, sampling 1000 americans will let you predict pretty accurate how the other 150 million or so will vote. It will not predict how an individual person will vote, but how the cohort will vote.
If you do a DBT that consistently correctly identifies a BT against an AYRE, you have pretty well establed that they have an audible difference, period.
If not, you have pretty well established that there is no audible difference between the two *in that particular setup*. If you then make sure that that particular setup is better than most, you can conclude that most people will not be able to hear a difference between a BT and an AYRE, based on sound alone (and this is an important qualifier).
Everything is about statistics, including science.
Jan Didden
proofAl Gore won the 2000 Presidential election.
The sample said so!
Anyway.
The dbt would only show what an average person might hear +/- some % margin of error.
A sample of 1000 people cannot reveal what the 1% might be able to hear.
What is 1% of the population?
If you are designing mass-fi for the 99%, the results may be fine, but not for the 1%!
Another flaw is that the 1000 tested may not know what they are listening for. If they did, 99% might notice the sound difference.
john curl said:Of course, you are correct, KBK, but my issue is WHY ABX testing 'Throws the baby out with the bathwater'.
Normally, when discussing parts, etc with a fellow designer, like Charles Hansen, Demian Martin, or Jam, by telephone, we have no trouble accepting that one emitter resistor will often sound better than another. We might even note it for further comparison of our own.
What is amazing to me is how often we come up with essentially the same conclusions, while working completely independently, and competitively with each other.
It is like when race car teams find the same tire to be better, over many other similar brands.
Now John,
You know perfectly well that a test in a K-Mart parking lot of 1000 drivers is
a perfect as a predictor of the next Indy 500 race!

Just trust the statistics!
I'm still catching on on some old posts.
Jan, that is just a load of pure nonsense.
"Criticizing is *always* done for good reasons"???
GMAFB. I don't think there is a person alive that hasn't criticized for poor reasons at least once. And most of the critical posts on this thread are for very poor reasons ....
janneman said:And criticizing is *always* done for good reasons, although they may not appear good reasons for the one not agreeing to it.
Jan, that is just a load of pure nonsense.
"Criticizing is *always* done for good reasons"???
GMAFB. I don't think there is a person alive that hasn't criticized for poor reasons at least once. And most of the critical posts on this thread are for very poor reasons ....
Charles Hansen said:I'm still catching on on some old posts.
Jan, that is just a load of pure nonsense.
"Criticizing is *always* done for good reasons"???
GMAFB. I don't think there is a person alive that hasn't criticized for poor reasons at least once. And most of the critical posts on this thread are for very poor reasons ....
Charles, I don't exactly remember to which post my reaction was meant. But what I meant was that even criticism that looks to others like it has no reason, still is done for a very good reason *for the criticizer*. There's always a strong motive and reason. Now that I read this, it looks like I'm kicking in an already open door, but I'm sure when I posted it I had a good reason 😉
Edit: What's GMAFB?
Jan Didden
Only slightly OT:
NS has introduced opamps with built-in EMI resistance. Not exactly your top audio opamp, but the development is interesting. The devices are: LMV831, LMV832, LMV834.
The app notre describes also the measurements they did to characterize EMI suseptibility of opamps and is probably more of interest to designers here: http://www.national.com/an/AN/AN-1698.pdf
Jan Didden
NS has introduced opamps with built-in EMI resistance. Not exactly your top audio opamp, but the development is interesting. The devices are: LMV831, LMV832, LMV834.
The app notre describes also the measurements they did to characterize EMI suseptibility of opamps and is probably more of interest to designers here: http://www.national.com/an/AN/AN-1698.pdf
Jan Didden
PMA said:Try Google, the first hit fits.
Ahh yes. Then, in brackets is says (polite form). The mind reels at what the un-polite form would be... 😉
Jan Didden
The sample said so!
I'm sorry to say so, but a lot of people do not know statistics and show utter ignorance.
To every correct statistic an error estimate must be given to be valid. The smaller the sample the larger the (possible) error. On the contrary, if the sample is sufficiently large, the error becomes - especially in non-border cases - insignificantly small.
The same for double blind testing: you need a rather large number of tests to be able to make a statistically significant statement - that means that one can separate the result well from plain guessing.
Of course that's a lot work and if there's no money behing it, nobody does it.
Without checking the error, I would think that one would need to repeat double blind testing at least 10 times to be on the safe side (same circumstances, same equipment).
Have fun, Hannes
h_a said:
I'm sorry to say so, but a lot of people do not know statistics and show utter ignorance.
To every correct statistic an error estimate must be given to be valid. The smaller the sample the larger the (possible) error. On the contrary, if the sample is sufficiently large, the error becomes - especially in non-border cases - insignificantly small.
The same for double blind testing: you need a rather large number of tests to be able to make a statistically significant statement - that means that one can separate the result well from plain guessing.
Of course that's a lot work and if there's no money behing it, nobody does it.
Without checking the error, I would think that one would need to repeat double blind testing at least 10 times to be on the safe side (same circumstances, same equipment).
Have fun, Hannes
Hannes, I guess there is a relationship between the number of subjects that do the test and the number of test runs? Like, you do 20 runs with 10 subjects, or you do 10 runs with 20 subjects? Would that be equivalent, from a statistics pov?
What would the numbers be for comparing two amps, with an error estimate, say 5% (if that is a sensible number)?
Jan Didden
janneman said:
So, we use this as a control. Double blind, with whatever other controls you would like. The BT and the 100$ Sony are correctly identified almost all the time, statistically significant.
So, we know that 'the system' is able to show audible differences. Now we replace the 100$ Sony with a 2000$ Sony.
result: correctly identified just above the statistical limit.
Now replace the 2000$ Sony with a 10,000$ AYRE.
Test results: correct identification not different from tossing coins, 50%.
Would that be an acceptable type of test, where you would conclude that this test setup can reveal differences between a BT, a 100$ Sony and a 2000$ Sony, but not between a BT and a 10,000$ AYRE.
Jan Didden
The idea is appealing, but enlights a general difficulty in developing a control that is sufficient for a difference with unknwon audibility.
So, you as an experimentator has to find out which way to get the most sensitive listeners. That means to observe the learning curve and test on different sensitivity levels.
And of course include a negative control as well.
Unfortunately there is no easy answer for all the possible questions; and, not to forget, after all there is the golden rule of testing: ´the more controlled a test is, the less the practicability of results´ (maybe the translation does not catch all the meaning).
So it might be, that the best test method is to hide from the listeners that they are participating in a somewhat ´official´ test.
We talked about that earlier, if you present just two different units, (same look, but different circuits inside), and ask just for some help in the decision which one to prefer, then listeners are doing what they are doing normally to compare two units.
In the end you´ll see, if there is significant difference.
I tried that method once and was quite happy with the results (of course you have to randomly change the presentation order, marking and you have make sure that the units aren´t different for a look inside and so on).
Apart from the mentioned difficulties, it has several advantages, because it let the listeners do what they normally would do; they choose the music, they choose how long to listen to each sample, they choose the level, and so on. No need for any level adjustment as random errors will be averaged out.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Solid State
- John Curl's Blowtorch preamplifier