Please read:
Picking Capacitors - Walter G. Jung and Richard Marsh
Also look for Cyril Bateman’s mulipart articles with in depth testing.
https://r.search.yahoo.com/_ylt=Awr...articles/RK=2/RS=wA.xzJR.EoGXUJBftmTyIR_1Isg-
Then you can make make an informed decision based on your budget. I work hard and buy film and foil Polypropylene for crossovers. Polystyrene and teflon for small size caps.
Picking Capacitors - Walter G. Jung and Richard Marsh
Also look for Cyril Bateman’s mulipart articles with in depth testing.
https://r.search.yahoo.com/_ylt=Awr...articles/RK=2/RS=wA.xzJR.EoGXUJBftmTyIR_1Isg-
Then you can make make an informed decision based on your budget. I work hard and buy film and foil Polypropylene for crossovers. Polystyrene and teflon for small size caps.
Last edited:
hi lojzek
please could you explain "impedance measurement setup" ? like a multi meter?
please forgive my amateur question, i'm just getting started in diy-hifi
gaz
I hope you will believe the science which is now coming your way, gaz! 😉 Sorry people, that's just a little in-joke!
Sorry for the tone of my earlier reply. Based on our short conversation on another thread, I was actually wondering if differences in ESR between different types of capacitor were of sufficient magnitude that, when taken into consideration along with the loudspeaker driver's impedance, would make any noticeable difference to the crossover frequency.
I recommend high voltage polypropylene capacitors for use in crossovers, except when the value is very large, when I'd go for a Mundorf ECap.
It is a VERY poor design that which can be improved by "improving" capacitors.
A good design is quite parts tolerance independent.
A bad design maybe one which is so close to the cliff edge that a small variation throws it into the abyss.
A good design is quite parts tolerance independent.
A bad design maybe one which is so close to the cliff edge that a small variation throws it into the abyss.
Try this site. He tests capacitors and describes each capacitors sound. Is he on to something or just full of it? You decide. Just keep scrolling down and you'll see dozens of caps tested, complete with his numerical score. Solen's are about a 6 if I remember right.
Humble Homemade Hifi - Cap Test
Humble Homemade Hifi - Cap Test
Maxwell House Coffe Nashville Science Comittee
Ok, the higher the voltage tolerance the worse the loss factor, you can add also pure metal foil versus metalized dielectric, double wounding... to increase the Q factor I surmise (it's how it seems to work on HumbleHomemadeHifi Coffee Shop). Ok we all understand a good designed filter is a good designed filter, but what are we learning about all of that choices with MKPs... Related to the monney vs the improvment I'm personnaly closer to the Lozjek point of view... am I wrong ?
Yes I understood it wasn't a question. So is it significant or not for you to be heard, better to give your point that kinding at me yet, no -if I understood your appologizes are ...not ? 😱
Does a MKT around 5K to 10KHz on may introduce enough loss to change the Q to be hearable ? Let's take an average 4 ohms impedance but a minimum 2.7 ohms in this frequencies windows : is it significant enough versus a MKP according to you ? Is the MKT ESR high enough there to introduce for illustration a slighty spl loss you may hear at those frequencies ? I'm thinking about the guy in an another thread which has a problem with a good Mundorf tweeter around 7K and asking himself about buying the same design with a Be dome. Could it be possible the bad bad npe or MKT to be an advantage here over a MKP for instance ?
I don't get your point about the crossover frequency, my simpliest though is its no significant enough related to the crossover point ony between several MKPs or did you refered or wanted to talk about some lossy NPE that have capacitance of few hundred nano Farad change according the frequency ?
Cause a said default may be an advantage at the service of a designer For illustration using a MKP or a lossy rough foil NPE to calm down a brighness cause a peak at high frequency or a too much flat spl curve ? Did you refer to hearable distorions related to loss factor ? I have to guess
Staying on the topic I will go myself with the Solens and keep the money for a next design on a better tweeter unit for instance... Am I wrong ? Why a Mundorf over a SCR smart cap serie or a SCR tin foil according to you ? Not sure I get your point...
Ok, the higher the voltage tolerance the worse the loss factor, you can add also pure metal foil versus metalized dielectric, double wounding... to increase the Q factor I surmise (it's how it seems to work on HumbleHomemadeHifi Coffee Shop). Ok we all understand a good designed filter is a good designed filter, but what are we learning about all of that choices with MKPs... Related to the monney vs the improvment I'm personnaly closer to the Lozjek point of view... am I wrong ?
Yes I understood it wasn't a question. So is it significant or not for you to be heard, better to give your point that kinding at me yet, no -if I understood your appologizes are ...not ? 😱
Does a MKT around 5K to 10KHz on may introduce enough loss to change the Q to be hearable ? Let's take an average 4 ohms impedance but a minimum 2.7 ohms in this frequencies windows : is it significant enough versus a MKP according to you ? Is the MKT ESR high enough there to introduce for illustration a slighty spl loss you may hear at those frequencies ? I'm thinking about the guy in an another thread which has a problem with a good Mundorf tweeter around 7K and asking himself about buying the same design with a Be dome. Could it be possible the bad bad npe or MKT to be an advantage here over a MKP for instance ?
I don't get your point about the crossover frequency, my simpliest though is its no significant enough related to the crossover point ony between several MKPs or did you refered or wanted to talk about some lossy NPE that have capacitance of few hundred nano Farad change according the frequency ?
Cause a said default may be an advantage at the service of a designer For illustration using a MKP or a lossy rough foil NPE to calm down a brighness cause a peak at high frequency or a too much flat spl curve ? Did you refer to hearable distorions related to loss factor ? I have to guess
Staying on the topic I will go myself with the Solens and keep the money for a next design on a better tweeter unit for instance... Am I wrong ? Why a Mundorf over a SCR smart cap serie or a SCR tin foil according to you ? Not sure I get your point...
Last edited:
I apologised as I thought you may have considered my tone to be argumentative - of that I was obviously correct! Let me be clear that I seek no argument. From what you have said, I am reassured that the difference in ESR between modern capacitors of different types is negligible when it comes to crossover design. Thanks for your comprehensive reply.
Could you please enlight me and answer to my questions above please ?
While it's negligible (but my questions to you really aren't as well) , in this little gap, some find it big enough to spend monney. I would like to understand those differences and if I'm correct in the way ESR or loss factor difference can explain it. It may help me as the original poster. Taking the example above, according to you, no difference in sound whatever the cap ? I'm not talking of designing a filter but in the original poster context. If the sound changed for some : Lee from HHMH cap test, Troels, enthusiasts here that work hard for Jantzen and Mundorf, I mean it should be related to science : that's what I would like to understand. Is it possible to hear 0.001 loss factor over 0.00001 over 0.8 at a given frequency as the reviewers write in caps test sites ?
Let say we are talking just for tweeter, not taking caps measurement at 100 Hz or 1 K Hz. But more where the tweeter lives : 2500 Hz to.... (write your age here 😀) ?
While it's negligible (but my questions to you really aren't as well) , in this little gap, some find it big enough to spend monney. I would like to understand those differences and if I'm correct in the way ESR or loss factor difference can explain it. It may help me as the original poster. Taking the example above, according to you, no difference in sound whatever the cap ? I'm not talking of designing a filter but in the original poster context. If the sound changed for some : Lee from HHMH cap test, Troels, enthusiasts here that work hard for Jantzen and Mundorf, I mean it should be related to science : that's what I would like to understand. Is it possible to hear 0.001 loss factor over 0.00001 over 0.8 at a given frequency as the reviewers write in caps test sites ?
Let say we are talking just for tweeter, not taking caps measurement at 100 Hz or 1 K Hz. But more where the tweeter lives : 2500 Hz to.... (write your age here 😀) ?
For clarity, low and high are relative here to circuit context. 'The lower the better' doesn't apply. I understand you are referring to Voltage rating and dissipation factor..
Do you have something (a link) to show this?
Of course once this is known, a designer will either measure and compensate or at least not be surprised if it is an issue. It will be seen as a circuit error in need of correction and thus, not applicable to the question of whether the capacitors actually sound different.
Among many, one of the means to measure impedance is to use Artalabs LIMP module. The schematic is attached, and the sound card used was a lower quality old Creative. If you inserted a capacitor instead, the curve would be mirror imaged, impedance falling down with higher frequencies. R (real part), jwL (imaginary part) of the complex number Z (impedance).
Attachments
Thank you AllenB for sharing your knowledge and long experience.
Sorry my typo, read MKT not MKP.
So we are talking about the negligible diference for precision to the readers:
At reading some MKT (so Polyester aka Mylar caps), I noticed differences between manufacterers
Some gives the dissipation factor at 1 K Hz or 100 Hz, some at 10 K or 100 K and I saw difference, for instance a Vishay MKT is much better at 10k than an Intertechnik -aka Audyn Caps. OEM for some Mundorf, npe in particular?
I also saw some used to choose MKT for aluminium tweeter dome, is it for cost they don't use mkp, I hear myself a less open sound with MKT vs MKP... (while I doubt of my objectivity, I prefer to try to understand about these little difference. For illustration at 10 K hZ an Intertechniq MKT has 8% DF and a 250V MKP SCR has 0,1%....and expensive brands with better construction, double wounded plain foil with higher internal damping, oil, bear wax or tin measure even better 0,01 or 0,001 %... on the paper difference seems a little "huge" (while still negligible and not price justified imo)
About the voltage rating, yes I saw the ESR of a MKP (metalized Polypropylene) is slighty better at higher tolerance voltage increasing the Q, so increasing (making it worse) slighty its cunterpart loss factor. David Lee honestly asking himself if what he hear is not due here to psychological factor, a simple MKP from Solen or Mundorf has the same subjective 6 notation, same caps at 400V instead 250 has a 6+...easiest difference being a little better esr so a slight more loss from the DF... hence my question to Galu fellow member, as he advices highest voltage.... ? Again we are talking about negligible difference maybe just young ears focus on by lack of experience (musicality is not always corralated to be able to hear the breath of the singer by more clarity)
Sorry my typo, read MKT not MKP.
So we are talking about the negligible diference for precision to the readers:
At reading some MKT (so Polyester aka Mylar caps), I noticed differences between manufacterers
Some gives the dissipation factor at 1 K Hz or 100 Hz, some at 10 K or 100 K and I saw difference, for instance a Vishay MKT is much better at 10k than an Intertechnik -aka Audyn Caps. OEM for some Mundorf, npe in particular?
I also saw some used to choose MKT for aluminium tweeter dome, is it for cost they don't use mkp, I hear myself a less open sound with MKT vs MKP... (while I doubt of my objectivity, I prefer to try to understand about these little difference. For illustration at 10 K hZ an Intertechniq MKT has 8% DF and a 250V MKP SCR has 0,1%....and expensive brands with better construction, double wounded plain foil with higher internal damping, oil, bear wax or tin measure even better 0,01 or 0,001 %... on the paper difference seems a little "huge" (while still negligible and not price justified imo)
About the voltage rating, yes I saw the ESR of a MKP (metalized Polypropylene) is slighty better at higher tolerance voltage increasing the Q, so increasing (making it worse) slighty its cunterpart loss factor. David Lee honestly asking himself if what he hear is not due here to psychological factor, a simple MKP from Solen or Mundorf has the same subjective 6 notation, same caps at 400V instead 250 has a 6+...easiest difference being a little better esr so a slight more loss from the DF... hence my question to Galu fellow member, as he advices highest voltage.... ? Again we are talking about negligible difference maybe just young ears focus on by lack of experience (musicality is not always corralated to be able to hear the breath of the singer by more clarity)
Last edited:
Thanks for answering my simple technical question.
I wonder what was so challenging about that one question that you were moved to ask TEN of me in return.
I obviously have a long way to travel on the path to enlightenment! 🙂
So with actual good caps, a designer may be able with some tweeter to notice an hearable difference when designing a new filter, then will compensate with some other factors, esr of an inductor and so on ? Not talking about the topic where it's about a ready made speaker filter, does a MKT or a non polar electrolytic can be chose to introduce some Z for instance to qvojd a little resistor value ? Less money spent for the manufacturer in mind ?
So being modern, different dielectric will not be negligible here as important enough to be heard and correct something the designer wisch to?
Absolutely, I asked a simple question regarding the relative magnitudes of the ESRs of different types of modern capacitors - it had nothing to do with how a capacitor 'sounds'.
I based the question on my knowledge that old electrolytics had significant ESRs which have to be compensated for when substituting a modern capacitor in a loudspeaker crossover.
Sorry if my question was 'off topic'.