Hi again,
I had a few more questions on caps for passive crossovers:
- Is there any major downside to running a network of parallel or series caps to get to a specific value rather than a single cap? If not, does that hold true regardless of how many you had to use (for instance, five small film caps of different values in parallel)?
- How much worse off are you for using a large value NP electrolytic to make up the bulk of the capacitance and then a small value non-electro in parallel?
- < less important question for less important crossovers
> Relatively old hi-pass caps test within range on an LCR meter, and series resistance looks OK on an ESR meter. Do you assume that those caps are in good working order, and if not, under what conditions will they show their age? I've got a scope and a frequency generator and would love to actually see whatever these distortions are that I'm trying to avoid.
Much Thanks!
George
I had a few more questions on caps for passive crossovers:
- Is there any major downside to running a network of parallel or series caps to get to a specific value rather than a single cap? If not, does that hold true regardless of how many you had to use (for instance, five small film caps of different values in parallel)?
- How much worse off are you for using a large value NP electrolytic to make up the bulk of the capacitance and then a small value non-electro in parallel?
- < less important question for less important crossovers
> Relatively old hi-pass caps test within range on an LCR meter, and series resistance looks OK on an ESR meter. Do you assume that those caps are in good working order, and if not, under what conditions will they show their age? I've got a scope and a frequency generator and would love to actually see whatever these distortions are that I'm trying to avoid.Much Thanks!
George
ONE cap sounds better, always
If in need of bigger one than available, like say 150uf, the TWO filmcaps100uf+50uf is better than a single 150uf bipolar
Whether 75uf+75uf film cap is even better, I dont know, but some will claim that, other the opposite
But ONE is better, so....
But I dont say that nothing else wont work
Design is the most important of all
Then comes xover
Next is drivers
Then the built
And so forth
Xover components is pretty much last on the long list
And how you use them is even more important than the component itself
If in need of bigger one than available, like say 150uf, the TWO filmcaps100uf+50uf is better than a single 150uf bipolar
Whether 75uf+75uf film cap is even better, I dont know, but some will claim that, other the opposite
But ONE is better, so....
But I dont say that nothing else wont work
Design is the most important of all
Then comes xover
Next is drivers
Then the built
And so forth
Xover components is pretty much last on the long list
And how you use them is even more important than the component itself
Last edited:
Damn. Not what I wanted to hear.
I'll have to stock up on some larger values from somewhere. I found some recommendations for Vishay (MKT maybe) and ordered a bunch of them, but there was a minimum of ten each, so I thought I would be better off going with 1 and 1.5uF, then a couple sub .5 values to fine tune. To get higher values, I'd have to order a whole batch (15 or 20 bucks worth) of something just for a couple caps and I still couldn't get the exact values in one unit. None of my target values are large. I think 5.5uF is the highest one right now.
Thanks!
George
PS- Still wondering on the second two questions if anyone knows.
<edit>- (regarding that NP electro/film combo question) If it matters much, those higher valued ones would be serving mid/low frequency duty. My tweeter stuff looks like it's all going to be <2uF.
Last edited:
Oh
with smaller values I never care too much about the exact value, but rather prefer a single cap, the closest standard value
Though one might need to determine whether going to the lower closest or higher closest value, if difference is close to the same
My, thats tricky in foreign language
Mind you, personally I have no doubt what I prefer
To me, multiple caps have more negative sideeffects than just using a standard value being slightly off value
You or others may have different experience/opinion
In time you may not think much about it either way, but will get accustomed to the sound you hear
Last edited:
I will have to disagree on this. Multiple small caps will decrease the impedance and inductance of the caps and this is a good thing. I have a personnal preference for only using one value when // caps. That is unless the other cap is a bypass. Using multiple caps is fine and should lead to improvements.
Terry
Terry
Why, ok it shouldn't come up very often, the values are small anyway, but why not. You decrease both inductance and resistance.
I agree in absolute terms if the caps are of different values, different values tend to cause smear, otherwise there may be advantages. Try both and see. It is a small point since I have seldom tried it.
Terry
The main problem, single or parallel, is believing the value printed on the cap - channel differences cause the "smear".
If you are on a tight budget, ie no measuring tools and cheap caps with larger tolerances, switching in parallel is your only hope - if you are lucky, tolerances will cancel each other to some degree.
If you are on a tight budget, ie no measuring tools and cheap caps with larger tolerances, switching in parallel is your only hope - if you are lucky, tolerances will cancel each other to some degree.
I have used lots of the oval Monacor
They have been very accurate
But I have heard differences between small caps of same measured value
I can only address this to manufactoring accuracy, tightness of "rolling"
Those different sounding caps actually does have sligtly different physical size
On most other caps using a standard size holster, you cant see this difference, but difference might still be there
Pit, the channel differences you mention I usually find is related to inductor matching
And sorry to say, but you cannot rely on delivered inductors to be of exact correct value
But I suppose you know that
They have been very accurate
But I have heard differences between small caps of same measured value
I can only address this to manufactoring accuracy, tightness of "rolling"
Those different sounding caps actually does have sligtly different physical size
On most other caps using a standard size holster, you cant see this difference, but difference might still be there
Pit, the channel differences you mention I usually find is related to inductor matching
And sorry to say, but you cannot rely on delivered inductors to be of exact correct value
But I suppose you know that
Pulse Rise Time
If you connect caps in parallel:-
(1)-use only equal value caps, eg:- 4.7//4.7 = 9.4uF.
(2)-use only same type of dielectric for both, and preferably the same brand and model for both.
Because, capacitors with different internal lengths -{that is the width of their foils}- have different Pulse Rise Times, and that seems to be what causes the smearing and other audible effects that others have Posted in this Thread about hearing.
Capacitors with different dielectrics, even if same uF and internal length, have different pulse rise times also, and the different dielectrics cause differences in sound.
Some listeners like the sound effects caused by mixing capacitors of different dielectrics; different internal lengths; etc ... but it is a sound effect and is not accurate passage of the signal's wave-form.
Inductance and resistance are important, but are secondary priorities to Pulse Rise Time for sound, and especially for high frequency sound if one wants accuracy.
If possible, one cap is best for high frequencies.
I think this is likely a result of the way the leads of parallel connected caps are soldered by some people - eg:- twisting the leads together for a distance and then soldering over them causes Impedance mismatches to occur between each cap of the pair.
Some audio engineers will laugh and ridicule what I have said, but if you look at critical RF circuitry you wil not see parallel connected caps with twisted together leads, except perhaps in parts of the Power Supply section.
Differences between samples' internal assembly are likey the cause of audible effects also,
such as tinitus Posted about in this Thread.
If you connect caps in parallel:-
(1)-use only equal value caps, eg:- 4.7//4.7 = 9.4uF.
(2)-use only same type of dielectric for both, and preferably the same brand and model for both.
Because, capacitors with different internal lengths -{that is the width of their foils}- have different Pulse Rise Times, and that seems to be what causes the smearing and other audible effects that others have Posted in this Thread about hearing.
Capacitors with different dielectrics, even if same uF and internal length, have different pulse rise times also, and the different dielectrics cause differences in sound.
Some listeners like the sound effects caused by mixing capacitors of different dielectrics; different internal lengths; etc ... but it is a sound effect and is not accurate passage of the signal's wave-form.
Inductance and resistance are important, but are secondary priorities to Pulse Rise Time for sound, and especially for high frequency sound if one wants accuracy.
If possible, one cap is best for high frequencies.
I think this is likely a result of the way the leads of parallel connected caps are soldered by some people - eg:- twisting the leads together for a distance and then soldering over them causes Impedance mismatches to occur between each cap of the pair.
Some audio engineers will laugh and ridicule what I have said, but if you look at critical RF circuitry you wil not see parallel connected caps with twisted together leads, except perhaps in parts of the Power Supply section.
Differences between samples' internal assembly are likey the cause of audible effects also,
such as tinitus Posted about in this Thread.
Last edited:
Alan, "rise time" - you´re bonkers. If it doesn´t show when measured with a static signal it´s not there, Would not look good in the ads, so it´s superstition. Go and wash your mouth with something .
And tinitus - Monacor. I keep waving their flag. I very much doubt you can get better value for your pennies elsewhere in Euroland. Any problems you have, you get an email back within half an hour, too.
And tinitus - Monacor. I keep waving their flag. I very much doubt you can get better value for your pennies elsewhere in Euroland. Any problems you have, you get an email back within half an hour, too.
Alan is a very clever guy, no doubt about it
Im sure he is on to something, in between
But I wouldnt use those "minor" theories as a cure fore all
Even if there is some truth to all of it, its not like a magic cure
But Im sure we all agree on using any good component, and get the job done
Keep it simple as possible
Use whats needed and no more
"Tricks" doesnt really work
Some of what have been said might sound strange
But its actually the other way round
Think about it
Im sure he is on to something, in between
But I wouldnt use those "minor" theories as a cure fore all
Even if there is some truth to all of it, its not like a magic cure
But Im sure we all agree on using any good component, and get the job done
Keep it simple as possible
Use whats needed and no more
"Tricks" doesnt really work
Some of what have been said might sound strange
But its actually the other way round
Think about it
Last edited:
Manufacturers' Specifications
Well Pit,
unfortunately yours is one of the types of responses I expected to get,
however if you obtain Specification Sheets from major manufacturers of capacitors for Industrial and similar purposes you will see the Pulse Rise Times specified for their various caps of different foil widths and for different dielectrics.
I have the Roedestein -{ERO, now Vishay/Roedestein}- specifications for their various different internally constructed Polypropylene caps and their Polycarbonate caps -{no longer manufactured}- and their Polyester caps.
I have similar for the Rifa manufactured caps, and the pulse rise times are somewhat differently specified to EROs, though I've misplaced the Rifa and will have to find it.
*******************************************************
"Q",
I am interested to learn why you think a different rise time would be a different capacitance ?
I do mean "Pulse Rise Time", and not "Time Constant".
Please explain, because there might be something that I have not considered.
However, in Roedestein's specifications there is different pulse rise times for example for 4.7uF of different polypropylene internally constructed types, polycarbonate, polyester-{MKT}.
Yes, all for a measurable 4.7uF cap, and for all the other capacitance sizes.
*******************************************************
Hi tinitus,
keep trusting your hearing - you seem to be onto something !
I can't state for absolute certainty it is the differences in pulse rise times we are hearing, but it seems possible from a lot I have read and though carefully about.
We do hear differences between caps with different Dielectric Absorptions -
eg:- polypropylene types versus MKT versus electrolytic, etc ...
but now we are hearing differences between combinations such as:-
3.3uF//6.8uF versus 5.1uF//5.1uF, or versus a single 10uF cap,
when all are selected from samples to give exactly the same measured capacitance, and when all are from exactly the same Series from the same manufacturer.
I have done this experiment and listened.
It probably can be measured if one has a suitable resolution, high frequency measurement device.
Well Pit,
unfortunately yours is one of the types of responses I expected to get,
however if you obtain Specification Sheets from major manufacturers of capacitors for Industrial and similar purposes you will see the Pulse Rise Times specified for their various caps of different foil widths and for different dielectrics.
I have the Roedestein -{ERO, now Vishay/Roedestein}- specifications for their various different internally constructed Polypropylene caps and their Polycarbonate caps -{no longer manufactured}- and their Polyester caps.
I have similar for the Rifa manufactured caps, and the pulse rise times are somewhat differently specified to EROs, though I've misplaced the Rifa and will have to find it.
*******************************************************
"Q",
I am interested to learn why you think a different rise time would be a different capacitance ?
I do mean "Pulse Rise Time", and not "Time Constant".
Please explain, because there might be something that I have not considered.
However, in Roedestein's specifications there is different pulse rise times for example for 4.7uF of different polypropylene internally constructed types, polycarbonate, polyester-{MKT}.
Yes, all for a measurable 4.7uF cap, and for all the other capacitance sizes.
*******************************************************
Hi tinitus,
keep trusting your hearing - you seem to be onto something !
I can't state for absolute certainty it is the differences in pulse rise times we are hearing, but it seems possible from a lot I have read and though carefully about.
We do hear differences between caps with different Dielectric Absorptions -
eg:- polypropylene types versus MKT versus electrolytic, etc ...
but now we are hearing differences between combinations such as:-
3.3uF//6.8uF versus 5.1uF//5.1uF, or versus a single 10uF cap,
when all are selected from samples to give exactly the same measured capacitance, and when all are from exactly the same Series from the same manufacturer.
I have done this experiment and listened.
It probably can be measured if one has a suitable resolution, high frequency measurement device.
FYI Rise time is specified for caps to be able to calculate the AC current capabilty of said part/s and these are usually called pulse caps. It's largely a function of the surface area of the ends of a cap and the method and quality of the connection system to the foil/plate edges. I = C*dV/dt where rise time is dV/dt .
Jadis
I dont think you will get a definitive answer mainly because there is none. As always it depends on the final application. Ie passive crossovers shunts, zobel and series pass elements are differing.
IMO I really Don't see a problem with paralleling film and electros for zobels with electros being the dominate value. Series pass caps for mids and tweets are more stringent and its better if you stick to using at least 75-90% of the total value in one dielectric case/package. Publish your schematic for a specific recomendation.
I dont think you will get a definitive answer mainly because there is none. As always it depends on the final application. Ie passive crossovers shunts, zobel and series pass elements are differing.
IMO I really Don't see a problem with paralleling film and electros for zobels with electros being the dominate value. Series pass caps for mids and tweets are more stringent and its better if you stick to using at least 75-90% of the total value in one dielectric case/package. Publish your schematic for a specific recomendation.
Last edited:
Thanks, I know I am
Its all based on listening experiences
I have very little theoretical knowledge about anything
I have learned that there is no magical stick or simple formula
And I dont really need one
Not quite so if you build speaker fore a living
Then you need to be able to repeat it over and over again
Question is, how good will that be
Many of the things we talk about, you wouldnt notice any of it on measurements
Mind you, I didnt say "not visible", I said "not noticed"
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.