Having measured the contribution of various components to distortion and posted or published the results, let me add a bit.
I can easily measure capacitance, inductance and resistance to four or five digit accuracy.
So yes electrolytic capacitors do behave differently than film types.
Among film types dielectric absorption does vary and in some applications may hit perceptible levels.
The most telling test was the vibration table test! No sheez. Folks who understand that vibration affects all capacitors, either isolate or encapsulate their crossovers.
Wow what a surprise that some capacitors handle the vibration tests better than others! (Not!!!)
So in the article referenced at the beginning of this thread, my critique would be they measured the wrong thing. I would look at intermodulation distortion of the midrange or teeter when the woofer was shaking the crossover. Of course their capacitors under test were apparently not in a loudspeaker enclosure with a woofer.
I can easily measure capacitance, inductance and resistance to four or five digit accuracy.
So yes electrolytic capacitors do behave differently than film types.
Among film types dielectric absorption does vary and in some applications may hit perceptible levels.
The most telling test was the vibration table test! No sheez. Folks who understand that vibration affects all capacitors, either isolate or encapsulate their crossovers.
Wow what a surprise that some capacitors handle the vibration tests better than others! (Not!!!)
So in the article referenced at the beginning of this thread, my critique would be they measured the wrong thing. I would look at intermodulation distortion of the midrange or teeter when the woofer was shaking the crossover. Of course their capacitors under test were apparently not in a loudspeaker enclosure with a woofer.
Last edited:
Perhaps the best thing for anyone to do is test it out yourself in an audio circuit.
I may get shoot for saying this but music signals are not the same as test signals.
Cheers
I may get shoot for saying this but music signals are not the same as test signals.
Cheers
Which model? I've thought of picking one up to add to the collection - a 1689 possibly; I currently us a Wayne Kerr 6425 that I keep in cal.
I've used both extensively and the GR's have no practical advantage over the 6425. The latter is actually more flexible.
The 1657 is a workhorse! Mine (still works but retired now) saw so much use (parts sorting in a mfg environment) that all the labels are unreadable...
No, I wasn't planning on changing my daily driver, I just have never had any personal experience with a GR bridge but have heard a lot of good about them. Very impressed with the few pieces of GR equipment that I have and had been thinking about getting one just to have.
Your right test signals are more revealing and some are harder to reproduce than music, full power square waves, multitone (30 sines) signals etc.
Yet, at some point we may need to come to grips with capacitors in a way we haven't really done yet. They can be more complex than the simple models we use for them, especially electrolytics.
So we don't depend on models, simple or not... we build circuits and see what comes out. For audio, what matters is the sound - and if it truly sounds different than expected, then there will be measurable and repeatable differences in the output vs the input. That is just physics and it applies equally irrespective of our expectations. Not to say that someone may not truly hear a difference - but then that falls in the realm of the psychological and applies only to them in that specific instance. IMO, quite valid for them, just don't expect it to apply to others,.
Hal, with all due respect, how often do you measure for 2nd order effects such as, perhaps, time invariance of DA equilibrium, correlated noise, ESL, ESR, etc., as function of capacitor aging, and or as a function of adaptation to operating conditions, etc., for every brand, model, physical format of capacitor? And if you don't measure for such things how do you know they are not variables, know that capacitor manufacturers don't measure for such and or other things they may know about that you never even thought of? I already know some manufacturers know things they haven't told you in their literature.
Most of the time we measure for FR, HD, noise floor, and maybe a few other things. Maybe ESL, ESR, under certain limited test conditions, voltage coefficient, temperature coefficient, etc. What do you look for?
Also, how much do you know about psychological attribution? Studied that as much as you have studied electronics?
Most of the time we measure for FR, HD, noise floor, and maybe a few other things. Maybe ESL, ESR, under certain limited test conditions, voltage coefficient, temperature coefficient, etc. What do you look for?
Also, how much do you know about psychological attribution? Studied that as much as you have studied electronics?
Last edited:
Which model are you talking about. Those that dabble with simulators often have quite complex models. And those who actually want to build precision circuits will care about and characterise over their complete wanted parameters. But audio isn't precision design in any sense. We don't keep our DACs in thermal and humidity controlled environments. We don't calibrate before each listening session. And I bet most people don't limit their critical listening to the 1-2 hours of the morning when the ear is most sensitive.
BTW I would view 'Time invariance of DA equilibrium' as pure marketing twoddle to impress those who look at the gold embossed names on capacitors in an amplifer and say 'it must sound good it's got frauducaps in it'.
Why on earth would I do all of that? I stick with known brands and they have done their due diligence. And, even in extremely sensitive applications, most of the variables tend to average on the macro scale and really don't matter to the end result - like a freshly cut yard, each blade of grass may be jagged and of different lengths but once you back out to the normal viewing distance it looks perfectly smooth. And for me, the end result is what I am concerned with - in both my professional and private life I am concerned with precision metrology - I work with PPB and nanovolts and not just DC - most signals, even what most would call DC, are AC as far as I am concerned. Amplification, both for DC and AC, are integral to what I do and I am very focused on knowing precisely what comes out versus what went in. Yes, capacitors have numerous qualities that have a certain variance but most of them fall well below even what I look at as a noise floor and the ones that don't I am aware of and compensate for if necessary - and if there is something unexpected that occurs, I troubleshoot until I find out what caused it - usually I screwed up somewhere, but occasionally it may be a component fault. But, in the end, there will be nothing magic or unknown about any of it... it's just engineering and physics in action. Know how capacitors behave, whether with bias or different dielectrics and such, and things tend to go smooth with no unexpected results.
Hal
Hal
That phrase does have a certain smell to it, doesn't it 😉
Or as Bob Pease would say: Floobydust. A mixture of words with no semantic meaning.
To clarify though, I don't mean in any way to trivialize capacitor parasitics and fundamental differences - just that they are well known and easily researched so should only cause issues through lack of knowledge on the designer's part. Said knowledge is no small thing though... much is not taught in school and has to be learnt through experience and there are subtleties that are not obvious until you are burn by them. I really think that true knowledge is directly proportional to how much Magic Smoke (tm) you have managed to let escape 🙄
Actually I have a minor in Applied Psychology. It's been more or less completely useless other than to make me wonder why I wasted all that time and money on something I was never that interested in to begin with. I've been involved in electronics all my life but it is a relatively recent career choice. I originally graduated with a major in accounting and was a CPA for 15 years. God, I'm such a nerd...
- Status
- Not open for further replies.