[Stereo75$$]

In the formula:  C=1/2x Pi x R x f ,  how do you find the value of f in the specs of the tweeter you are looking replace and use a signal capacitor.  I have 3 ways speakers with a 12 full range woofer,  8 in midrange with cap and previous RadioShack horn tweeter rated at 8 ohms 2000-25000Hz.  The filter on the old tweeter was a 4uf cap at 50 volts.  I replaced it with Morel CAT 378 and a 3.9 uF csa copper cap.  The problem I hear is at the very high notes the tweeter does not sound clear say you hearing a high key on a piano like in the Vince Guaraldi Christmas CD when the notes are at the higher octave there is a shrill, muffled not clear sound comping from both tweeters.  The old tweeters were clear sounding but one of them lost the voice coil hair size wire connection.  So I was trying to calculate simply the single high pass replacement for these tweeters
Thanks

[FullRangeMan]

Welcome to AC 

[Phil A]

Welcome!

[ArthurDent]

Greetings & Welcome to AC   

[FullRangeMan]

The F value may be the freq point tô the resistor operate but it could be other thing also.

[nlitworld]

Really hard to give solid advice here without more info on the speaker characteristics and current crossover design. Correct me of I'm wrong, but your muffled sound could be a gap in frequency response from where your new tweeter is supposed to cross over at since you used the same value cap. It could also be that your tweeter was accidentally wired out of phase. I would recommend finding the specs and measuring the mid woofer and seeing where to set your new crossover point at. There are plenty of online crossover calculators to help with values, but first order high pass are pretty easy.

[mboxler]

If you rearrange the equation for a 4uF (.000004F) capacitor and an 8 ohm tweeter you get...

F = 1 / 2ℼ * R * C

F = 1 / 2ℼ * 8 * .000004

F = 1 / .00020106176

F = 4974 hz

If you are going to use a 4uF capacitor the tweeter will need to handle a first order cross @ 4974 hz.

[Wayner]

Its basically a 5k cross-over point, but seems to be crouding the midrange. Midrange drivers need 3 octaves to work properly with the cap and inductor in the circuit, unless the mid is XO's with just a single cap.

@5000 XO point, 6db down is @ 2500, 12db down is @ 1250, 18db down is 625. 625 is below resonance frquency and is the reason your tweeter pukes. You need to either raise the XO point, or turn the XO into a 12db per octave slope, rather than the 6db. The circuit requires a cap in series and an inductor coil across the tweeter's inputs.

A second order Linkwitz-riley XO (12db per octave slope would require a 2uf in series and a .5mh coil across the tweeter coils. Then @2500 it would be 12db down, @ 1250 would be 24db down, @625 would be 36db down.

[Stereo75$$]

Wayner, so I am new to all of this .   A little hx my dad and I built these speaker cabinets in 1974.  As a kid at that time we just had the option of RadioShack. WE did a 3 way speaker design and did single high pass to the midrange and tweeter.  I have listened to the music via my Marantz 2250B .  The sound was great until I noticed one day the left channel was weaker.  I found out that it was the tweeter.  So I contacted Madison sound and checked what brand of tweeter was close to the old horn RadioShack tweeter   2000-25000Hz at 8 ohm ) now 1800-22000Hz.  I used a 3.9 uf csa filter/ v vs the old filter which was 4uf/50v.  Changed all wiring and marked all + speakers with Supraclassic 15 awg was about 22 g wire, Then I replaced the wires just coming out of the box to Binding post    The insulation is the same yellow and the port is 2 1/4 front port as before.

Wayner the main issue is the tweeter louder than the other speakers and the high notes played on the speaker like on a piano played more forcefully like with the Charlie Brown Christmas CD sounds distorted and not clean.   Its hard to listen to this for a while.   I checked the sound with my Bose headphones and it sounds clean so I feel it is both tweeters.   I feel that it is the high pass filter and Morel suggested using a higher uf high pass filter at 4.7 uf or 6.8 uf.  The issue I have is where do you stop.  I thought a formula would solve this but the value of f they always say pick the crossover point you want to achieve .  The issue is w/o special  gear how do you find that out w/o trial and error like I am doing. 
so again the issue is the highest point of the tweeter notes sound distorted like it has too much input going to this even at lower volumes and it sounds somewhat louder compared to the mid and woofer

[richidoo]

Welcome to AudioCircle!

Likely your interpretation of high notes and overtones on the piano are 3-5kHz. Above that is tinkly sibilant noise like cymbals, and above 10kHz most seasoned citizens can't hear anything anyway. 3-5kHz is where the Morel peaks a little if the cap is too small large  , so it will sound louder there. https://www.morelhifi.com/product/cat-378/
This could also be where the new tweeter is clashing with the midrange driver if the phase of each are not aligned.

It would distort if it was playing too low, but your 3.9uF cap is not letting it play too low, so that's not the source of distortion. If the mid and tweeter share crossover components like series crossover, etc maybe the mid is distorting.

Tweeter wouldn't distort at high frequency unless playing at extreme high SPL. The Morel specsheet shows maximal flat response with 5.6uF cap. Not knowing anything else about the particulars that is the best place to start. Buy cheap caps to experiment to find the right size. Then you can upgrade them to nicer sounding cap after you settle on the size. You can fine tune the size by parallelling caps together to get values between the available single cap values, but a single cap is close enough.

You say the RS tweeter was nominal 8 ohms, but without a measuring that is just a guess, or a marketing estimation. This Morel is actually 6 ohms. So there could be a difference there also. If the Morel impedance is lower than the RS tweeter, then it will draw more current and play louder. You said the Morel is too loud and the ears rule in the end, so you'll need to attenuate the tweeter with resistors. Use 5-10W metal oxide resistors.

Trial and error design by ear is difficult without some knowledge to interpret what your hearing and to understand what is the actual goal. If you're curious to learn more about crossover design to save some time and money I recommend reading Loudspeaker Design Cookbook as a good place to start for the basics.

If you just want to throw the dart and if the previous tweeters worked OK with just one cap, then I would try the 5.6uF cap on the Morels, then experiment with L-Pad resistor network to bring down the SPL to taste. "L-Pad" just means 2 resistors, one in series with the tweeter signal and the other in parallel with the tweeter. Together these allow you to finely adjust the overall SPL of the tweeter and the slope of the tweeter response.

Some useful tools to aid the process are SPL meter app for smart phone and test tones of various freqs, can be made for free with WinISD, or Audacity, et al.  OR the super easy way is OmniMic, which allows measuring and graphing the frequency responses of midrange and tweeter separately and together, also showing acoustic phase response to make sure they are singing together nicely in the range in which they overlap. Speaker design is a complex artform!

Start a new thread in Enclosures Circle if you want more detailed help with your project.
Rich

[Wayner]

Wayner, so I am new to all of this .   A little hx my dad and I built these speaker cabinets in 1974.  As a kid at that time we just had the option of RadioShack. WE did a 3 way speaker design and did single high pass to the midrange and tweeter.  I have listened to the music via my Marantz 2250B .  The sound was great until I noticed one day the left channel was weaker.  I found out that it was the tweeter.  So I contacted Madison sound and checked what brand of tweeter was close to the old horn RadioShack tweeter   2000-25000Hz at 8 ohm ) now 1800-22000Hz.  I used a 3.9 uf csa filter/ v vs the old filter which was 4uf/50v.  Changed all wiring and marked all + speakers with Supraclassic 15 awg was about 22 g wire, Then I replaced the wires just coming out of the box to Binding post    The insulation is the same yellow and the port is 2 1/4 front port as before.

Wayner the main issue is the tweeter louder than the other speakers and the high notes played on the speaker like on a piano played more forcefully like with the Charlie Brown Christmas CD sounds distorted and not clean.   Its hard to listen to this for a while.   I checked the sound with my Bose headphones and it sounds clean so I feel it is both tweeters.   I feel that it is the high pass filter and Morel suggested using a higher uf high pass filter at 4.7 uf or 6.8 uf.  The issue I have is where do you stop.  I thought a formula would solve this but the value of f they always say pick the crossover point you want to achieve .  The issue is w/o special  gear how do you find that out w/o trial and error like I am doing. 
so again the issue is the highest point of the tweeter notes sound distorted like it has too much input going to this even at lower volumes and it sounds somewhat louder compared to the mid and woofer

You'll need an L-pad to knock down the output while retaining the resistance of the tweeter circuit (which the XO is based on). I believe your new tweeter is around 92db/m/w, not knowning what the rest of your drivers are, but betting that they are not 92, perhaps more like 88 or 87. 2 types of L-pads can be tried. One is variable (the typical level control), the other is just 2 resistors, one in series with the tweeter, one across the coil. There are some free attenuation calculators, do a google search. I'd probably start with with a 5db reduction in the tweeter output. Perhaps you'd be better off with a level control (8 ohm) from Parts Express.

BTW, I've run a Morel tweeter MD-20 close to the resonance frequency with a 6db per octave XO and I could hear buzz from the speaker even at low levels. A careful listening test will verify one way or another.