• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

About Tweeter Capacitor.

jst

Active Member
Joined
Aug 8, 2021
Messages
178
Likes
45
Just a quick question, can I use this capacitor for tweeter ? It doesn't look audio-y that's in my perception usually cylinder shaped and has yellow color on it, but it is also non polar.

6982fefe-bf3e-4a72-83a6-80eeb34434a7.png



Can I just use a 1 way tweeter crossover with default capacitor value 2.2uF and change the capacitor as I like to like 0.47 uF or 0.1uF or anything ? or does the capacitor value affect the coil/resistor value ?
 
Last edited:
That type will work fine, but do NOT change the value of the cap or the crossover will not function as designed. Color and shape are irrelevant, and "audio" capacitors are generally a waste of money.
 
Last edited:
That type will work fine, but do NOT change the value of the cap or the crossover will not function as designed. Color and shape are irrelevant, and "audio" capacitors are generally a waste of money.
What is usually marketed for the ones you mention as "audio" capacitors is tolerance and how that, for example, gives a good stereo matching between speakers.
I don't know if +/-1% or TS capacitors' +/-5% can create such big differences in FR? And more doubtful if it would be audible? There are probably already threads on that topic here on ASR.

These capacitors offer excellent build quality, long-term reliability, and astonishing sonic clarity, all at a 1% tolerance for improved stereo matching between speakers.

 
What is usually marketed for the ones you mention as "audio" capacitors is tolerance and how that, for example, gives a good stereo matching between speakers.
I don't know if +/-1% or TS capacitors' +/-5% can create such big differences in FR? And more doubtful if it would be audible? There are probably already threads on that topic here on ASR.

These capacitors offer excellent build quality, long-term reliability, and astonishing sonic clarity, all at a 1% tolerance for improved stereo matching between speakers.

Find me a pair of drivers that match to 1%. :cool:

WIMA caps are solidly built and reliable.
 
I'm not sure I would want to use one of their MKS (metallized PET) capacitors for a crossover, even those with high enough capacitance and voltage rating tend to have somewhat concerningly low impulse ratings. MKP4 series caps with 2.2µ/250V are a bit chonkier at 11 x 21 x 26.5 or 31.5 mm with an either 22.5 or 27.5 mm pin spacing. If you can get some of those, they should work fine. (In a low-power application, 100 V would probably work as well.) Crossover film caps tend to be regular, non-metallized polypropylene with somewhat lower capacitance, that's why they tend to be bigger.
 
Find me a pair of drivers that match to 1%. :cool:

WIMA caps are solidly built and reliable.
But that doesn't make it better, on the contrary, worse if you add it in. Does not necessarily mean that it will be particularly audible. It probably doesn't matter, especially considering what placement of the speakers in the listening room can do. :oops:

Design and size of the listening room, surrounding walls, furniture, carpets, windows and so on can cause differences in FR between the speakers. In fact, it is
probably the most important thing to concentrate on, which of course I know you know, but I mention it for others who read this thread and get concerned about capacitors and their deviations. Stereo but also in mono can, in the lower frequencies due to the aforementioned, trigger a 15 dB FR peak/ valley.
I took the first example that came up when I searched, for example:
JBLTHXamp.jpg
 
Last edited:
  • Like
Reactions: jst
I'm not sure I would want to use one of their MKS (metallized PET) capacitors for a crossover, even those with high enough capacitance and voltage rating tend to have somewhat concerningly low impulse ratings. MKP4 series caps with 2.2µ/250V are a bit chonkier at 11 x 21 x 26.5 or 31.5 mm with an either 22.5 or 27.5 mm pin spacing. If you can get some of those, they should work fine. (In a low-power application, 100 V would probably work as well.) Crossover film caps tend to be regular, non-metallized polypropylene with somewhat lower capacitance, that's why they tend to be bigger.
For a tweeter cap, no big deal. I've used them in lots of speakers I've built and over the past 30 years, zero failures.
 
But that doesn't make it better, on the contrary, worse if you add it in. Does not necessarily mean that it will be particularly audible. It probably doesn't matter, especially considering what placement of the speakers in the listening room can do. :oops:

Design and size of the listening room, surrounding walls, furniture, carpets, windows and so on can cause differences in FR between the speakers. In fact, it is the most important thing to concentrate on, which of course I know you know, but I mention it for others who read this thread and get concerned about capacitors and their deviations. Stereo but also in mono it can indeed, in the lower frequencies due to the aforementioned, trigger a 15 dB FR peak/ valley.
I took the first example that came up when I searched, for example:
View attachment 401473
As a practical matter, two caps out of the same batch will usually match to better than 1%. They may be as far as 5% off nominal (rare), but they will be close to one another.
 
Just a quick question, can I use this capacitor for tweeter ? It doesn't look audio-y that's in my perception usually cylinder shaped and has yellow color on it, but it is also non polar.

6982fefe-bf3e-4a72-83a6-80eeb34434a7.png



Can I just use a 1 way tweeter crossover with default capacitor value 2.2uF and change the capacitor as I like to like 0.47 uF or 0.1uF or anything ? or does the capacitor value affect the coil/resistor value ?
Wire 4 of them in parallel, maybe 5 of them, to get close to the original 2.2uF required by the crossover. They will work fine that way. If you use only one of them, the crossover will not function correctly.
 
Wire 4 of them in parallel, maybe 5 of them, to get close to the original 2.2uF required by the crossover. They will work fine that way. If you use only one of them, the crossover will not function correctly.
I see, good to know but I was just experimenting with a range of capacitor value, from 6.8uF , 4.7uF, 3.3uF, 2.2uF, 1uF , but then I thought whether capacitor below 1uF exist, and found that one. even to 0.22 and 0.1uF, but below 0.47 seems the volume is too low and sometimes unheard so I think it exceeds 20khz ?

I can't decide which one I like though, so I keep changing them at random time lol.


I wonder how you guys tailor your tweeter when using diy speakers, using measurement or just tweak and audition it until you like how it sounds.
 
You can simulate the filter and calculate the required XO.
The filter frequency depends on the mid XO point, power rating and sensitivity or the tweeter as well as the impedance of it and which order the XO must be.
1st, 2nd or 3rd order.

Then you can use a UMIK microphone for some close-up measurements to check if you have the right level and if it needs some compensation network or not.
 
I see, good to know but I was just experimenting with a range of capacitor value, from 6.8uF , 4.7uF, 3.3uF, 2.2uF, 1uF , but then I thought whether capacitor below 1uF exist, and found that one. even to 0.22 and 0.1uF, but below 0.47 seems the volume is too low and sometimes unheard so I think it exceeds 20khz ?

I can't decide which one I like though, so I keep changing them at random time lol.


I wonder how you guys tailor your tweeter when using diy speakers, using measurement or just tweak and audition it until you like how it sounds.
Careful not to get too large with the capacitance. That would allow more lower frequency content through to the tweeter, which could overpower and ruin it.
 
  • Like
Reactions: jst
Speaking of which, you generally want something a little more complex than just a single capacitor for your tweeter crossover. A 6 dB slope is very shallow, made worse by the tweeter's impedance bump at its fundamental resonance. So you'll have a large messy crossover region and an easily overloaded tweeter unless it's crossed very high.

If you want to get your feet wet in speaker simulation, two popular packages are (a) VituixCAD and (b) Boxsim.
 
  • Like
Reactions: jst
Speaking of which, you generally want something a little more complex than just a single capacitor for your tweeter crossover. A 6 dB slope is very shallow, made worse by the tweeter's impedance bump at its fundamental resonance. So you'll have a large messy crossover region and an easily overloaded tweeter unless it's crossed very high.

If you want to get your feet wet in speaker simulation, two popular packages are (a) VituixCAD and (b) Boxsim.
Agree.
 
  • Like
Reactions: jst
I see, good to know but I was just experimenting with a range of capacitor value, from 6.8uF , 4.7uF, 3.3uF, 2.2uF, 1uF , but then I thought whether capacitor below 1uF exist, and found that one. even to 0.22 and 0.1uF, but below 0.47 seems the volume is too low and sometimes unheard so I think it exceeds 20khz ?
1a) capacitors "exist" (i.e., are purposefully manufactured) into the range of low picofarad values (1 uF = 1E6 pF).
1b) This is a capacitor that can exist even if not purposefully manufactured (i.e., for-profit manufactured). :)
1730135746200.jpeg


2) It's all relative! I.e., a very high (i.e., ridiculously high) first order XO isn't as absurd as it might seem. For example, the "supertweeters" on my Altec-esque loudspeakers have first order XO using a 1 uF Sprague "Atom" capacitor. That is a nominal XO frequency of 20 kHz for an 8 ohm load! The drivers are very sensitive, and the slope of the XO is only 6 dB per octave, so there's plenty of output (- 6dB, nominally) at 10 kHz ... and even below.

EDIT: Here's a handy first order XO calculator:

1730136110531.png

 
  • Like
Reactions: jst
I see, good to know but I was just experimenting with a range of capacitor value, from 6.8uF , 4.7uF, 3.3uF, 2.2uF, 1uF , but then I thought whether capacitor below 1uF exist, and found that one. even to 0.22 and 0.1uF, but below 0.47 seems the volume is too low and sometimes unheard so I think it exceeds 20khz ?
You might be right!

We don't know the details of your crossover network, or if there's just a capacitor for the tweeter.

Capacitors are almost always marked and resistors (if any) are usually marked or color coded. But inductors are often unmarked and unknown so that makes it more difficult to analyze or reverse-engineer the crossover.

The simplest RC filter where "R" is an 8-Ohm tweeter, the cutoff is at 9kHz (3dB down) with 2.2uF. The cutoff is inversely proportional to capacitance so with 1.1uF you're at 18kHz.

With a lower capacitance the tweeter will kick-in* at a higher frequency and you'll get a hole in the response between where the woofer (or midrange) falls-off and the tweeter kicks-in.

With a higher capacitance the tweeter will kick-in at a lower frequency and the woofer (or midrange) will overlap. That should make a bump in the response but there may also be a phase mismatch so you could end-up with a bump and a dip!

And when you increase capacitance, more power (wattage) goes to the tweeter for two reasons: The more octaves of sound you send to the tweeter, the more energy. And, there is more energy in the lower octaves. If the capacitance value is too high (or if you bypass it altogether) you can end-up burning-out the tweeter.




* Kick-in is rather bad terminology because there is a filter-slope. ;)
 
  • Like
Reactions: jst
Careful not to get too large with the capacitance. That would allow more lower frequency content through to the tweeter, which could overpower and ruin it.
Yea, highest I ever tried was 10uF, and I just tried in the calculator posted here, that would produce around 2khz (9.94uF), which is still in the range of the tweeter frequency.

1a) capacitors "exist" (i.e., are purposefully manufactured) into the range of low picofarad values (1 uF = 1E6 pF).
1b) This is a capacitor that can exist even if not purposefully manufactured (i.e., for-profit manufactured). :)
View attachment 402254

2) It's all relative! I.e., a very high (i.e., ridiculously high) first order XO isn't as absurd as it might seem. For example, the "supertweeters" on my Altec-esque loudspeakers have first order XO using a 1 uF Sprague "Atom" capacitor. That is a nominal XO frequency of 20 kHz for an 8 ohm load! The drivers are very sensitive, and the slope of the XO is only 6 dB per octave, so there's plenty of output (- 6dB, nominally) at 10 kHz ... and even below.

EDIT: Here's a handy first order XO calculator:

View attachment 402258
Interesting calculator, I don't actually use a woofer since the bass is from a 2.1's sub, but I have the tweeter connected parallel to a 4 inch full range, also 8 ohm, with no crossover and open baffle. I'm currently using a 6.8uF capacitor and judging from the measurement of the calculator, the frequency is around 3khz and if I want 20khz, 0.9uF can be used. Good to know.
====

I only use a capacitor with no resistor and inductor/coil, what resistor and inductor actually do ? resistor for reducing the dB and inductor for cutting the high ? Why would I want to cut the high since it's a tweeter that's supposed to play high frequency ?

BTW with the current capacitor that I use, 6.8uF which shows 3khz in the calculator, is 3khz too low, or depends since it's a crossover frequency and not actually a flat 3khz, since measurement with mic usually show a range and a peak and a downslope until higher frequency but lower dB.
 
Last edited:
When using a single capacitor (first order filter) I would recommend not filtering too low in frequency as that would lead to increased power in the tweeter and you can blow it up more easily.

When you want to use the lowest frequency the tweeter is specified for I recommend to use either 2nd or 3rd order filters (cap + inductor = 2nd order, cap-inductor-cap = 3rd order.

6.8uF on a tweeter (1st order) is a bit on the high side of things (too much energy in the mids enter the tweeter heating up the voicecoil but not adding more upper mids).

It is not recommended to make use of the recommended frequency range using a 1st order filter.
Best would be to remain above that frequency and use higher order filters too.
You may have to do the same for the (mid)woofer in some cases.
 
  • Like
Reactions: jst
This is the 4 inch full range that I use ,found a youtube video someone doing a measurement on it, I dont use a crossover on it and I dont put this speaker driver into a box. it has a 2khz highest peak and start to go down at 14.7khz. Maybe I should not use tweeter since it has enough high and eliminate complications?

this is the tweeter, local brand ACR 9kh
Diameter 3.7 Inch / 9 cm
Height 3,2 cm
Cone Type Cloth Transparant
Max Power = 150 W
Impedance = 8 Ohm
Magnet Weight 8.0 Oz
Frequency Response 2 KHz - 20 KHz
SPL 1 W / 1 M 91 dB
VC dia 2.54 Cm / 1 Inch
 
Last edited:
Back
Top Bottom