About Tweeter Capacitor.

[solderdude]

Still 8 ohm when a filter is used in both drivers.
When just filtering the tweeter at 14kHz the impedance will still be around 8ohm as the impedance of the FR will rise at higher frequency and the tweeter will have some extra resistance (because of the filter impedance at audio frequencies).
When just filtering the tweeter at say 4kHz and not filtering the FR driver the impedance will become a bit below 8 ohm (but not 4 ohm) for higher frequencies but this is not a problem as there is not much 'energy' above 4kHz relatively to the mids.

 

[jst]

Ok thx again.

 

[mhardy6647]

Thank you all, I now have a bit of understanding about XO and apply a custom one to the drivers I use, I've tried 3 and not satisfied with how they sound.

--
One more thing, if my tweeter is 8 ohm and fullrange also 8 ohm, when coonected like the diagram, is it then 4 ohm or still 8 ?

With the proviso that the nominal impedance of any driver is just that (nominal*), most crossover strategies maintain system impedance. In other words, two 8 ohm drivers connected to their amplifier source through a "normal" passive XO will still present a nominal 8 ohm load to the amplifier.

EDIT: Sorry -- @solderdude already answered your question
I'll still leave my post here, though, FWIW.
__________________
* In reality, the actual impedance of (almost) any loudspeaker driver or system varies as a function of frequency of the applied signal. There are three components to impedance: resistance (which is, or at least should be, frequency independent), capacitive reactance, and inductive reactance. The nominal value for the impedance of any given driver is just that, and does NOT necessarily represent the value that should be used for the best possible crossover design! That's why some other comments you've gotten in this thread reflect that best case XO design is a nontrivial task. That said, the nominal values may get one close enough to be happy with the result!

 

[jst]

With the proviso that the nominal impedance of any driver is just that (nominal*), most crossover strategies maintain system impedance. In other words, two 8 ohm drivers connected to their amplifier source through a "normal" passive XO will still present a nominal 8 ohm load to the amplifier.

EDIT: Sorry -- @solderdude already answered your question
I'll still leave my post here, though, FWIW.
__________________
* In reality, the actual impedance of (almost) any loudspeaker driver or system varies as a function of frequency of the applied signal. There are three components to impedance: resistance (which is, or at least should be, frequency independent), capacitive reactance, and inductive reactance. The nominal value for the impedance of any given driver is just that, and does NOT necessarily represent the value that should be used for the best possible crossover design! That's why some other comments you've gotten in this thread reflect that best case XO design is a nontrivial task. That said, the nominal values may get one close enough to be happy with the result!

Yea, all this time I thought it becomes 4 ohm because parallel, kinda misleading watching many videos saying 4 ohm while it's true without any filter since noone uses speaker system without filter unless maybe fullrange or mid like you lol.

So if I add a 8ohm woofer with a 1 way passive xo it will still stay 8 ohm then ?

What if it's a mixed of different ohm like 8ohm tweeter 8 ohm Fullrange and 4 ohm woofer ? ( in case I switch to 2.0 without sub or just out of curiosity )

 

[solderdude]

When you make a passive 3 way system with 8Ω nominal rated speakers and use a proper filter the nominal impedance of the resulting 3 way speaker is still 8Ω nominal.
In reality the impedance will vary a lot probably somewhere between 6ohm and 30ohm or so but will be around 8Ω at 1kHz.

When you add a woofer of 8Ω (with a low pass filter) but do not high pass the FR driver the nominal impedance below the XO point of the nominal impedance around the lower frequencies can drop to 4Ω but... most speakers have a resonance peak between say 40Hz and 100Hz or so and in practice means that aside from the lowest subbass frequencies the impedance will not drop down to 4Ω.

When you would add a 4Ω woofer the same happens but at subbass frequencies the impedance will drop as low as 2.5Ω or so.

Also it is important to use a woofer that has the same sensitivity (dB/2.83V) so not recommended to do that.

When you use proper filtering when making the 3-way speaker and use a 4Ω woofer + 8Ω mid + treble the lowest impedance the amp will see (for subbass and high bass, depending on the XO frequency) would simply be 4Ω and above the low XO frequency would rise to 8Ω.

Note that passive filtering a subwoofer type of arrangement, when done properly, requires a bit more complex filter when XO-ing at low frequencies where there is an impedance hump in one or both speakers.

 

[mhardy6647]

Yea, all this time I thought it becomes 4 ohm because parallel, kinda misleading watching many videos saying 4 ohm while it's true without any filter since noone uses speaker system without filter unless maybe fullrange or mid like you lol.

So if I add a 8ohm woofer with a 1 way passive xo it will still stay 8 ohm then ?

What if it's a mixed of different ohm like 8ohm tweeter 8 ohm Fullrange and 4 ohm woofer ? ( in case I switch to 2.0 without sub or just out of curiosity )

What @solderdude said
In terms of the 'on paper' effect of mixing different impedance drivers, most of the on-line calculators allow one to input the nominal impedance of the tweeter & the woofer.
There are probably on-line simulators to show the 'theoretical' frequency response of a given XO configuration. I don't have one handy, though!

 

[jst]

Thx it clears the confusion I had all this time. So it's fluctuative depending on frequency and not simply a fixed 4 or 8, and measured at 1khz for spec-ing (kinda like dB with 1w or 2.83v/m, which I also learned here some time ago)

 

[solderdude]

yep

 

[mhardy6647]

yep

I was going to say yup, but I believe that you & I hail from different regions.

 

[solderdude]

yep

 

[jst]

Hahah.

Anyway I keep thinking about the harshness when i crank the volume quite loud and it's dominantly the high, so it must be like @solderdude said, that the sensitivity is higher and can be reduced with a resistor. (Would like to use an l-pad but can't find it locally)

So I added a resistor to my tweeter, 5W10ΩJ . I don't know how many dB is reduced by adding it but I'm quite satisfied with how it sounds now, like more balanced.

If my amp is 50 watt do I have to match it ? can't seem to find it.

 

[solderdude]

adding a resistor in series also changes the xo frequency (and increases the impedance for the treble)
get the level correct first.
Then when the level is correct you can either change the capacitor value or calculate the values for a proper attenuator ( consists of 2 resistors)

 

[mhardy6647]

adding a resistor in series also changes the xo frequency (and increases the impedance for the treble)
get the level correct first.
Then when the level is correct you can either change the capacitor value or calculate the values for a proper attenuator ( consists of 2 resistors)

yup.


This is why "we" brought up the L-pad mentioned earlier. There are actually several different ways to attenuate without changing the load impedance (which is why only adding a single resistor in series with the load input changes the XO frequency) -- but they all require (at least) two resistors, as @solderdude notes above).

Know Your L- and T-Pads, September 1967 Radio-Electronics

Potentiometer values are chosen to maintain a fairly constant input and output impedance throughout the attenuation adjustment range.

www.rfcafe.com

L-pad

Z sub s = source impedance; Z sub L = load impedance
R1 is the series resistor (resistance) of the L-pad; R2 is the parallel resistor (resistance)

T-pad

uses three resistors (R1, R2, and R3)

H-pad (for balanced circuits, on the right)

Aren't you glad you asked, @jst? Actually, you didn't ask per se, but this is all good stuff for you to know about! Again, the equations that relate the values of the various "R" to the impedance of the source, load, and the desired attenuation are all easy to find on the internet, and there are calculators for all of them.
The L-pad topology is fine for your needs.

 

[egellings]

Hahah.

Anyway I keep thinking about the harshness when i crank the volume quite loud and it's dominantly the high, so it must be like @solderdude said, that the sensitivity is higher and can be reduced with a resistor. (Would like to use an l-pad but can't find it locally)

So I added a resistor to my tweeter, 5W10ΩJ . I don't know how many dB is reduced by adding it but I'm quite satisfied with how it sounds now, like more balanced.

If my amp is 50 watt do I have to match it ? can't seem to find it.
View attachment 403478

"J" is for 5% tolerance.

 

[levimax]

Lots of good stuff here.... now if you REALLY want to get into it there is a great free crossover CAD program called VituixCAD, the learning curve is steep but not impossible if you are stubborn enough and have the time. https://kimmosaunisto.net/

 

[solderdude]

10Ω on an 8Ω tweeter means the attenuation is 7dB.
Also your XO frequency for the tweeter shifted a little over an octave lower.
So.. you either use half the capacitance you have in the there now (so 1.5uF) or use the same capacitor but use 2 resistors.
Where R1 = 3.9Ω and R2 = 6.8Ω.
The tweeter is in parallel to R2 and the capacitor is connected to R1 (on the left side)

 

[jst]

"J" is for 5% tolerance.

Oh ok.

Lots of good stuff here.... now if you REALLY want to get into it there is a great free crossover CAD program called VituixCAD, the learning curve is steep but not impossible if you are stubborn enough and have the time. https://kimmosaunisto.net/

Thx I'll keep that in mind for when I want to go deeper.

yup.


This is why "we" brought up the L-pad mentioned earlier. There are actually several different ways to attenuate without changing the load impedance (which is why only adding a single resistor in series with the load input changes the XO frequency) -- but they all require (at least) two resistors, as @solderdude notes above).

Know Your L- and T-Pads, September 1967 Radio-Electronics

Potentiometer values are chosen to maintain a fairly constant input and output impedance throughout the attenuation adjustment range.

www.rfcafe.com

L-pad

View attachment 403515

Z sub s = source impedance; Z sub L = load impedance
R1 is the series resistor (resistance) of the L-pad; R2 is the parallel resistor (resistance)

T-pad
View attachment 403517
uses three resistors (R1, R2, and R3)

H-pad (for balanced circuits, on the right)
View attachment 403519

Aren't you glad you asked, @jst? Actually, you didn't ask per se, but this is all good stuff for you to know about! Again, the equations that relate the values of the various "R" to the impedance of the source, load, and the desired attenuation are all easy to find on the internet, and there are calculators for all of them.
The L-pad topology is fine for your needs.

I cannot find L-Pad in local store, only normal potentiometer. So I guess I'll just use resistor for now. The idea of just using knob to adjust the sound level is very user friendly lol.

10Ω on an 8Ω tweeter means the attenuation is 7dB.
Also your XO frequency for the tweeter shifted a little over an octave lower.
So.. you either use half the capacitance you have in the there now (so 1.5uF) or use the same capacitor but use 2 resistors.
Where R1 = 3.9Ω and R2 = 6.8Ω.
The tweeter is in parallel to R2 and the capacitor is connected to R1 (on the left side)
View attachment 403550

If I want to reduce only 3dB, what size should I use ? Does the Watt matter in there , there are 5 to 30w with increment of 5, above that the look is different like what ppl use for dummy for test load.

You said half capacitance, as of now I'm using 1 uF+0.22uF+0.1uF. So now I'm using half from 1.32uF ? So I should double that to like 2.64uF (I can use 2.2+0.47=2.67uF) ? or use two resistors.

Or use 2 capacitors with 3.9 and 6.8 size.

Here is the speaker that I need to add resistors to


So I would need to do it like this ? Or the capacitor should be BEFORE the 3.9ohm resistor ? you said on the left side, so maybe before the resistor.

Also if the resistor is put on + and - of the speaker , is it safe and not connecting the + and - ?



=====
This site has resistor/l-pad calculator for the dB but it doesn't include the capacitor.

L Pad Calculator + Speaker Attenuation Guide

Don't let decibels and Ohms intimidate you. My L-pad calculator simplifies designing an L-pad resistor network for your tweeters. I've done the math for you!

soundcertified.com

Is there an all in one calculator that calculates the capacitor+inductor like mhardy6647 posted earlier and include the resistor for when I want to reduce a custom dB.

 

[solderdude]

Ah you are not using the tweeter what it is designed for but rather as a super tweeter.
Your XO point calculates at -3dB at 15kHz so the SPL is already 3dB down there.
Brightness is in the 3-6kHz range which the tweeter does not do so any perceived brightness is in the FR driver.
That one can fix with a LCR filter.

> 10kHz is the sharpness range. So if you experience sharpness only when the tweeter is engaged you can pad it down a notch.

In your current experiments if I were you and not make use of the tweeter qualities opposite those from the FR I would simply grab a few different resistors mount a different value, add a switch so you can switch it in and out and try to determine what works for you.

I assumed you XO'ed at 6kHz.

You can use the same calculator for the tweeter but just use impedance + chosen resistor for the 'R'.

The impedance of the FR driver is rising towards the upper treble anyway so the L in series probably also isn't at the exact frequency anyway.

 

[jst]

Ah you are not using the tweeter what it is designed for but rather as a super tweeter.
Your XO point calculates at -3dB at 15kHz so the SPL is already 3dB down there.
Brightness is in the 3-6kHz range which the tweeter does not do so any perceived brightness is in the FR driver.
That one can fix with a LCR filter.

> 10kHz is the sharpness range. So if you experience sharpness only when the tweeter is engaged you can pad it down a notch.

In your current experiments if I were you and not make use of the tweeter qualities opposite those from the FR I would simply grab a few different resistors mount a different value, add a switch so you can switch it in and out and try to determine what works for you.

I assumed you XO'ed at 6kHz.

You can use the same calculator for the tweeter but just use impedance + chosen resistor for the 'R'.

The impedance of the FR driver is rising towards the upper treble anyway so the L in series probably also isn't at the exact frequency anyway.

I see. I'm still ordering the inductor/coils, the 0.21mH still on process, I want to try using 6khz later when it arrives.

Ok, so I don't need to change anything then.

Just curious, is this kind of connection safe ? isn't that like connecting + and - ?

 

[solderdude]

Without the red arrows that is the way to connect the Lpad. The capacitor in series with Rseries. Does not matter where that capacitor is (on the left or right side in the schematic) as it is in series.

The + and - on the left of the schematic kind of replace the + and - or the unattenuated driver.

5W to 10W is fine.
Don't ever do this trick with woofers or FR drivers.