Hi nicoch58,
Yes, as a matter of fact I do. Same as any manufacturer of a heavily fashion influenced product like audio equipment.
One - market acceptance and popular opinion in order to gain market share
Two - a price advantage with the capacitor brand.
These things sure as heck are not engineered to a specific brand and model. Engineering can identify a number of more than suitable candidates, but the marketing departments dictate what is used these days. In the 60's and early 70's the engineering departments properly selected components, procurement selected the vendor and component from a list of acceptable parts. Then cross-eyed marketing people gained control and we have the situation we have today. I've lived through the entire process and watched the shift first-hand.
If you have any other idea of why this is, you have been trained by popular press and are exactly the consumer they wanted to create. This stuff goes into your head to create expectation bias. Works like a charm whether you'll admit it or not. If you have any doubts about this, ask a psychiatrist This too is a science and heavily leveraged by marketing departments world-wide. Why? Because it not only works, it is extremely effective. YOu are walking prrof of this.
-Chris
Yes, as a matter of fact I do. Same as any manufacturer of a heavily fashion influenced product like audio equipment.
One - market acceptance and popular opinion in order to gain market share
Two - a price advantage with the capacitor brand.
These things sure as heck are not engineered to a specific brand and model. Engineering can identify a number of more than suitable candidates, but the marketing departments dictate what is used these days. In the 60's and early 70's the engineering departments properly selected components, procurement selected the vendor and component from a list of acceptable parts. Then cross-eyed marketing people gained control and we have the situation we have today. I've lived through the entire process and watched the shift first-hand.
If you have any other idea of why this is, you have been trained by popular press and are exactly the consumer they wanted to create. This stuff goes into your head to create expectation bias. Works like a charm whether you'll admit it or not. If you have any doubts about this, ask a psychiatrist This too is a science and heavily leveraged by marketing departments world-wide. Why? Because it not only works, it is extremely effective. YOu are walking prrof of this.
-Chris
Nichicon MUSE ES?
The ES are bipolar electrolytic and sound pretty good when bypassed with a film cap). I have a couple of banks of MUSE ES for a tube amp output cap (not recommended as they only go to 50V). If you need a cap that copes with reverse polarity then it’s a good starting point.
You could use any bipolar electrolytic that fits the requirements but I would listen to the amp too after measuring it!
Last edited:
Hi NickKUK,
Any good technician worth his salt always listens to the product as well as measures it. Only an idiot fails to measure the device once they are done.
These days I can tell pretty accurately how something will sound after performing all the measurements. But still I listen.
-Chris
Any good technician worth his salt always listens to the product as well as measures it. Only an idiot fails to measure the device once they are done.
These days I can tell pretty accurately how something will sound after performing all the measurements. But still I listen.
-Chris
Please note the scales used in the graph.
The absence of the test setup and applied signal level to the setup really makes this information a lot less useful than you present it to be. If you normalize the graph, the performance isn't that good - depending on the impedance level of the circuit the capacitor has been placed in.
So basically if you are trying to prove something, you have not. There are many more questions that should be asked than answered. That and the results are only valid for that set of circumstances, and since they are unknown at this point the information presented could be extremely misleading.
-Chris
The absence of the test setup and applied signal level to the setup really makes this information a lot less useful than you present it to be. If you normalize the graph, the performance isn't that good - depending on the impedance level of the circuit the capacitor has been placed in.
So basically if you are trying to prove something, you have not. There are many more questions that should be asked than answered. That and the results are only valid for that set of circumstances, and since they are unknown at this point the information presented could be extremely misleading.
-Chris
Just from my limited experience I was thinking - what does the input signal look like. Most sig gens will have harmonics and for THD most people seem to opt for a single oscillator specifically built for THD measurements. However without a description of the test setup it's questionable (even if an accurate result) - spending time around a number optical metrologists and scientists in my old job has made me much more jaded for measurements 😀
A separate history question.
Was there anything particularly good/interesting about either Samsung SSL caps (green) or Rubicon sky blue ones (seemed to be used to decouple some BB opamps of 1997 era)? If not I'll just replace them with modern panasonic/CDE/nichicon in an old CDP.
Was there anything particularly good/interesting about either Samsung SSL caps (green) or Rubicon sky blue ones (seemed to be used to decouple some BB opamps of 1997 era)? If not I'll just replace them with modern panasonic/CDE/nichicon in an old CDP.
Hi nicoch58,
... and that tells us exactly what?
NOTHING!
You are presenting zero information and it's less useful because people not accustomed to measurement procedures will assume things. Worst, they assume these findings are valid (they may or may not be). Worse yet, there is something called "read across" where one set of findings are applied to a different situation as if they are still valid. Airplanes have crashed killing hundreds due to this.
I'll be honest with you. This is irresponsible, what you just did. When measurements are taken of sound equipment, there are standard test setups and the equipment used is listed. You did this in high school in fact. So there is zero excuse for presenting data in this manner. In my case, I follow these procedures and levels to the letter, and I use standard test loads and test signal levels. My findings are repeatable by anyone with industry standard equipment. There are no standardized test setups for capacitor distortion measurements. Therefore all data must be presented with the test setup and description of how connections are made, plus signal levels and the equipment used (which you provided - I guess, afterwards). This is merely a cut and paste from an article without references.
I'm a certified calibration technician for one. I worked in test equipment calibration labs, ISO certified with standards traceable to NIST. I am also aware and was tested on test procedures - in other words, how to set up equipment to minimise variables, and quantify those variables to come to what is called an "error budget" so we can insure the TUR is less than 4:1 (industry standard limit). TUR stands for Test Uncertainty Ratio. My equipment is up to date.
The reason I mention this is to make you aware that I am speaking from a position of first hand knowledge, industry accepted. Not anecdotal, real first hand knowledge on the subject of electronic testing. I am going to bet that these tests were performed in a non-rigorous manner. Maybe not haphazard, but not conforming to an accepted lab procedure. Overall the results may be valid for that specific test setup, but we really don't know any more than that. Whether these results support any view, mine or anyone else's, they simply are not complete enough to warrant anything but further investigation at this point. Certainly not the way they are presented here.
The question NickKUK posed was fair, now we know it was an AP system, so as long as it was working properly and not loaded outside it's operating range by the test setup, we can assume low distortion. However a plot of the residual distortion of the system would have been expected. I do that just to ensure the test setup is functioning properly and it hasn't loaded anything down to far over the frequencies of interest.
Designing and testing the test setups is not trivial. This is also a very basic step. To present findings, you must be rigorous in your approach to make certain you are not introducing errors in your setup. In this case you would want to test also at various test signal levels, and impedances. That goes for the test setup, and each component you are testing. It takes time and effort to do this, plus a lot of thought and knowledge of the test fixture components and connectors.
Now for the real kicker. Do these measured results actually make any audible (measureable) effect given the system and various noise levels of the listening environment? Or, are we worried about things that are clearly not audible or detectible by the human body? I'm not at all saying these things don't matter. What I am saying is that they only matter when the aberrations from perfect are at a high enough level. Now there is the engineering question. Anything beyond that involves arm chair quarterbacks.
-Chris
... and that tells us exactly what?
NOTHING!
You are presenting zero information and it's less useful because people not accustomed to measurement procedures will assume things. Worst, they assume these findings are valid (they may or may not be). Worse yet, there is something called "read across" where one set of findings are applied to a different situation as if they are still valid. Airplanes have crashed killing hundreds due to this.
I'll be honest with you. This is irresponsible, what you just did. When measurements are taken of sound equipment, there are standard test setups and the equipment used is listed. You did this in high school in fact. So there is zero excuse for presenting data in this manner. In my case, I follow these procedures and levels to the letter, and I use standard test loads and test signal levels. My findings are repeatable by anyone with industry standard equipment. There are no standardized test setups for capacitor distortion measurements. Therefore all data must be presented with the test setup and description of how connections are made, plus signal levels and the equipment used (which you provided - I guess, afterwards). This is merely a cut and paste from an article without references.
I'm a certified calibration technician for one. I worked in test equipment calibration labs, ISO certified with standards traceable to NIST. I am also aware and was tested on test procedures - in other words, how to set up equipment to minimise variables, and quantify those variables to come to what is called an "error budget" so we can insure the TUR is less than 4:1 (industry standard limit). TUR stands for Test Uncertainty Ratio. My equipment is up to date.
The reason I mention this is to make you aware that I am speaking from a position of first hand knowledge, industry accepted. Not anecdotal, real first hand knowledge on the subject of electronic testing. I am going to bet that these tests were performed in a non-rigorous manner. Maybe not haphazard, but not conforming to an accepted lab procedure. Overall the results may be valid for that specific test setup, but we really don't know any more than that. Whether these results support any view, mine or anyone else's, they simply are not complete enough to warrant anything but further investigation at this point. Certainly not the way they are presented here.
The question NickKUK posed was fair, now we know it was an AP system, so as long as it was working properly and not loaded outside it's operating range by the test setup, we can assume low distortion. However a plot of the residual distortion of the system would have been expected. I do that just to ensure the test setup is functioning properly and it hasn't loaded anything down to far over the frequencies of interest.
Designing and testing the test setups is not trivial. This is also a very basic step. To present findings, you must be rigorous in your approach to make certain you are not introducing errors in your setup. In this case you would want to test also at various test signal levels, and impedances. That goes for the test setup, and each component you are testing. It takes time and effort to do this, plus a lot of thought and knowledge of the test fixture components and connectors.
Now for the real kicker. Do these measured results actually make any audible (measureable) effect given the system and various noise levels of the listening environment? Or, are we worried about things that are clearly not audible or detectible by the human body? I'm not at all saying these things don't matter. What I am saying is that they only matter when the aberrations from perfect are at a high enough level. Now there is the engineering question. Anything beyond that involves arm chair quarterbacks.
-Chris
Look at "Overlays" in #2,428
1uF 50V
Pana Bi-Polar
+0V Bias
100R/1K@1V
Yes, that is a VERY odd place to note the conditions.And if I am reading it right this is a 1,592Hz cut-off (hi-pass, lo-pass?? makes a difference) which is a classic place to NOT use any electrolytic cap.
I could postulate that this cap with a 10k resistor (100X bigger) would move to 16Hz and -130dB. But also wonder why they didn't measure that way....
Hi PRR,
Absolutely correct.
There isn't enough information to spend enough time to decipher what is really going on. However it is a great example of how test results can be presented and possibly misrepresent what is actually going on.
Absolutely correct.
There isn't enough information to spend enough time to decipher what is really going on. However it is a great example of how test results can be presented and possibly misrepresent what is actually going on.
Hi Tony,
I agree with you entirely ... until the truth is distorted and then it becomes a lie. Free enterprise assumes the truth is being told so the target audience can make a decision based on the facts. If the truth is not being presented, it is no longer "free enterprise".
You know, preferences are fine, opinion is fine as long as it doesn't contradict the facts. I am all about the truth as you well know. That's all, real simple.
For example, if someone prefers the sound of a single ended tube amp, that's fine. However as soon as they try to claim it has lower distortion, there is a big problem because that is demonstratively not the case. See what I mean?
Now when "data" is presented as proof for any argument, it must be presented in a way that people can interpret it in a way that is true and factual. Anything less makes that "data" into a lie. Something put out there to mislead people. It may not be intentional, or it may be. That doesn't change the fact that what is offered misrepresents what is actually true. People tend to accept technical appearing reports and graphs as correct and factual. That is where I feel the presenter must be careful and actually know and understand what they are presenting. The same is true for a story that isn't complete. The complete story may totally change what someone understands from what is said. That is a lie of omission. This is what the above graph is.
-Chris
I agree with you entirely ... until the truth is distorted and then it becomes a lie. Free enterprise assumes the truth is being told so the target audience can make a decision based on the facts. If the truth is not being presented, it is no longer "free enterprise".
You know, preferences are fine, opinion is fine as long as it doesn't contradict the facts. I am all about the truth as you well know. That's all, real simple.
For example, if someone prefers the sound of a single ended tube amp, that's fine. However as soon as they try to claim it has lower distortion, there is a big problem because that is demonstratively not the case. See what I mean?
Now when "data" is presented as proof for any argument, it must be presented in a way that people can interpret it in a way that is true and factual. Anything less makes that "data" into a lie. Something put out there to mislead people. It may not be intentional, or it may be. That doesn't change the fact that what is offered misrepresents what is actually true. People tend to accept technical appearing reports and graphs as correct and factual. That is where I feel the presenter must be careful and actually know and understand what they are presenting. The same is true for a story that isn't complete. The complete story may totally change what someone understands from what is said. That is a lie of omission. This is what the above graph is.
-Chris
I have post back in the thread the info https://linearaudio.nl/cyril-batemans-capacitor-sound-articles