Re-voicing my Peerless Kevlar and Vifa Textile dome ATL speakers

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,822
Points
113
Location
Mumbai
I will be re-voicing my 5.25" Kevlar mid-woofer and Vifa Textile dome ATL speakers this month and after that will post if for sale in the forum. Will use my experience gained for the past 5 years in the crossover design. I had this speaker for 2 years since 2013 and after buying the tube amp did not use them as they were around 88dB SPL not able to get driven enough by my puny tube amp. Will re-polish them just before shipping. I would pay 5% of my sales proceedings to hfv as a gesture of appreciation if sold.



Thanks for looking.
 
Last weekend was able to revoice the speaker and this is what i get after measuring. The synthesis method is cool and it gave me just 3 passive components in the crossover (series type) with just one capacitor and one inductor for the filters and one resistor for level matching the tweeter with the woofer. That's it. Checkout the measurements -


For those who are keen on knowing what a series crossover schematic look like - here it is for this speaker system
 

Attachments

  • series.png
    series.png
    86.9 KB · Views: 43
Setting up a series crossover take up too many iterations and is quite dependent on the drivers too, right ?

To my amusement, I saw a huge debate on series crossovers. It was both love for some and hate relationship for others !!!!

I first designed with parallel crossover and later converted it to series. IMO the series crossover gives a better step and impulse response as its a by-pass type filter. In parallel crossover the filter elements are in series with the driver and in seriesvcrossover they are in parallel with the driver. Also its difficult to tweak a series crossover only by listening unless you know what you are doing. Hence synthesis is easier than analysis using series crossover and that way its much faster to design & implement. Filter phase works differently in series and parallel crossover (dont know how though).

Offlate my minimalist approach is fetching better dividends that using too much of compensation network. But for this the driver quality should be good.
 
Will post simulated response in the evening, dont remember now.
The reason why I asked is I have worked with 6 ohm vifa tweeters many times before. A 10uf and above in series sounds really good, as you are sending vocals to the tweeters which is the fastest driver in your box now. So peppy decay is expected. But in all cases, I ended up breaking the tweeters when driven louder. After that, the tweeters lost the top airneyss and mids through them got aggressive. I screwed up multiple tweeters this way. In the beginning the vocals sound amazing as every upper mid detail used to be there. This is why even though it's nice at lower volume manufacturers are reluctant to add larger capacitors across tweeters.


But, if it works for your case, it's really great. Have you tried listening them louder? My tweeters we're vifa d26 and scan speak h26 series. Both partially (Ferro fluid came out of its place due to lower frequency vibration) broke with 10uf playing louder. They were comfortable with 4 uf playing louder.
 
The reason why I asked is I have worked with 6 ohm vifa tweeters many times before. A 10uf and above in series sounds really good, as you are sending vocals to the tweeters which is the fastest driver in your box now. So peppy decay is expected. But in all cases, I ended up breaking the tweeters when driven louder. After that, the tweeters lost the top airneyss and mids through them got aggressive. I screwed up multiple tweeters this way. In the beginning the vocals sound amazing as every upper mid detail used to be there. This is why even though it's nice at lower volume manufacturers are reluctant to add larger capacitors across tweeters.


But, if it works for your case, it's really great. Have you tried listening them louder? My tweeters we're vifa d26 and scan speak h26 series. Both partially (Ferro fluid came out of its place due to lower frequency vibration) broke with 10uf playing louder. They were comfortable with 4 uf playing louder.
Ok, thanks for sharing your experience with the Vifa tweeter. This particular tweeter has a resonance of around 1.1KHz with FF. This tweeter does not have a resonance chamber and the magnet is opened to the ambient. What i have done is - added a 3" PVC tube that exactly fits the magnet size and the baffle and stuffed this chamber heavily with recron polifill to damp out the resonance. So my tweeter resonance is around 600Hz with this modification. Due to this i am able to drive the tweeter and cross it much lower. Also the Q at resonance is damped very much and hence there is no high current demand at resonance due to this. Also i have added a resistance in series to level match the mid-woofer level.

TBH, i have not driven this driver at very high level ( i anyway have only a 6w + 6W tube amplifier), but i could not notice any tweteer fatigue even at 2 o'clock position volume. May be the damping trick is working on the tweeter though its too early to conclude.
 
What is the frequency at which the tweeter is crossed?
The crossover freq is 1650Hz.

Why would you want to route vocals through the tweeter ?

I came to know about series crossover through this:

View attachment 28976

Series crossovers do need fine quality drivers and are tuned by ears. Swears the person who designed crossovers for pureaudioproject for their Open Baffle speakers.
I believe crossover design are driver dependent and cannot be generalized imo. I may plan to add the resistor across the woofer which allows to lower the Qts for the woofer for a TL design. But in an OB this would not be required as it will lower the Qts of the driver. Again i believe 3.9 ohm is too low a value and will affect SPL & amplifier damping. I will prefer between 15 Ohms to 47 Ohms.
 
Why would you want to route vocals through the tweeter ?

I came to know about series crossover through this:

View attachment 28976

Series crossovers do need fine quality drivers and are tuned by ears. Swears the person who designed crossovers for pureaudioproject for their Open Baffle speakers.


"Why would you want to route vocals through the tweeter ? " The vocal range is normally extended to around 3khz . In a speaker box, the tweeter diaphragm is the lightest massed unit. it can respond faster and stop faster. Also since lower 'bass frequencies' like anything under 500hz is not coming to the tweeter, the partial vibration is relatively low compared to woofer.

By using them to handle much more lower frequencies you are having the advantage of making vocals to sound more energetic, more complex passage would sound dynamic. Many brands like Spendor, Harbeth and so on do this on their two way designs. But, those tweeters are designed to handle the speakers. So till what point can a tweeter handle lows,? it can handle anything provided, so little power reaches so that that little suspension of the dome does not loose its ability to pull it back. Also, till ferrofluid leaks out which is one of the crucial elements making them to move in a controlled way.

So rule of thumb of crossover design is to leave an octave between the fs and crossover point for a tweeter. This is to avoid the suspension break on the tweeter by getting it to the strain faster than the woofer. The voice coil may not break , but may be you will end up with ferrofluid leak or failure of the suspension.

Again, "

"Why would you want to route vocals through the tweeter ?" to make it faster, less coloured. By doing so, the tweeter power handling before breaking is considerably lowered. Manufacturers tend to play safe by crossing it high. For some people like me , its a eureka moment, when I found I can make a better crossover than the manufacturer. -all until I realized it was there for a reason.

" This tweeter does not have a resonance chamber and the magnet is opened to the ambient. " - Can you post the model number of the tweeter? I have used this one before : https://www.madisoundspeakerstore.c...-bc25tg15-04-1-textile-dome-tweeter/?mobile=0

From the tweeters design point of view that tweeter SHOULD not change the sound depending on what you do behind it. The front dome is totally isolated from the rear in this case. Why? To install in boxes without further isolation to protect it from the air movement from other drivers. Can say it is sealed in one way and the only sounding part is the dome. Are you sure the lower handling is not due to your box's baffle? If the pipe structure alone, is making a sound difference, outside the box, can you explain how ended up deciding the length of the pipe to be used? Sorry for the long questionaire. I also had noticed change in sound of treble before in the boxes, but always it had to do some thing with the cabin itself. If this is the case, I can start consider trying it.

From the tweeters design point of view that tweeter SHOULD not change the sound depending on what you do behind it. The front dome is totally isolated from the rear in this case. Why? To install in boxes without further isolation to protect it from the air movement from other drivers. Can say it is sealed in one way and the only sounding part is the dome. Are you sure the lower handling is not due to your box's baffle? If the pipe structure alone, is making a sound difference, outside the box, can you explain how ended up deciding the length of the pipe to be used? Sorry for the long questionaire. I also had noticed change in sound of treble before in the boxes, but always it had to do some thing with the cabin itself. If this is the case, I can start consider trying it.[/QUOTE]
 
Last edited by a moderator:
" This tweeter does not have a resonance chamber and the magnet is opened to the ambient. " - Can you post the model number of the tweeter? I have used this one before : https://www.madisoundspeakerstore.c...-bc25tg15-04-1-textile-dome-tweeter/?mobile=0


From the tweeters design point of view that tweeter SHOULD not change the sound depending on what you do behind it. The front dome is totally isolated from the rear in this case. Why? To install in boxes without further isolation to protect it from the air movement from other drivers. Can say it is sealed in one way and the only sounding part is the dome. Are you sure the lower handling is not due to your box's baffle? If the pipe structure alone, is making a sound difference, outside the box, can you explain how ended up deciding the length of the pipe to be used? Sorry for the long questionaire. I also had noticed change in sound of treble before in the boxes, but always it had to do some thing with the cabin itself. If this is the case, I can start consider trying it.
Typically the resonance of the dome tweeter are lowered by damping the rare chamber of the tweeter. There is no calculation done by me, but just added a pvc pipe of 6" length and 3" dia and stuffed heavily with polyfill. This helped to lower the resonance of the tweeter to 600Hz. Also the Q @ resonance was much lowered by this. You can try this as this cost nothing.

Also its a known fact if the Q at resonance is high, the current drawn by the tweeter will be higher at this resonant frequency. Since the resonance and Q both is lowered by the cavity damping, these effects are minimized. Also now, my cross-over frequency 2.5 times the resonance frequency and hence it allows a first order filter without the issue of cone breakup mode at lower frequency.

If you are trying this for your tweeter - i will suggest you measure the Z of the tweeter before and after the mod for your own comparision and this will allow you to add stuffing inside the chamber.
 
Typically the resonance of the dome tweeter are lowered by damping the rare chamber of the tweeter. There is no calculation done by me, but just added a pvc pipe of 6" length and 3" dia and stuffed heavily with polyfill. This helped to lower the resonance of the tweeter to 600Hz. Also the Q @ resonance was much lowered by this. You can try this as this cost nothing.

Also its a known fact if the Q at resonance is high, the current drawn by the tweeter will be higher at this resonant frequency. Since the resonance and Q both is lowered by the cavity damping, these effects are minimized. Also now, my cross-over frequency 2.5 times the resonance frequency and hence it allows a first order filter without the issue of cone breakup mode at lower frequency.

If you are trying this for your tweeter - i will suggest you measure the Z of the tweeter before and after the mod for your own comparision and this will allow you to add stuffing inside the chamber.

The ability for a tweeter to handle the lows is only depending on how much can the voice coil handle + the capability of the suspension to break + the ferrofluid inside - incase of vifa /scanspeak/seas. These are mechanical limitations. Doing whatever in the cabin however does not imrove these weaknesses for sure. Still cannot digest the idea of dampening the magnet from where the sound never comes out in a tweeter. (woofer , may be) But it will be one new thing learned. Certain brands like B & W does this, but its done because the diaphragm is not isolated from the rear mechanically.

and the resonance chamber extends to the rear.

http://www.bowers-wilkins.com/Discover/Discover/Technologies/Aluminium_Tweeters.html


But in case of regular vifa/scanspeak entry levels(costing upto 15k ) the resonance chamber is usually in the middle and isolated from the rear. its meant for easy integration into boxes without further need of another chamber. I took a pic from a scanspeak to see how deep it is. But, again, I will not jump into conclusions before trying out. Just my might is not digesting things fast!
 
This tweeter must not be influenced by anything behind it. Just took out the scanpspeak H26TG series. There isnt a way for the tweeters waves to reach the rear after installing in a baffle. The assembly is same for the flat plate instead of horn, See the screw mounts. Vifa / scanspeak / seas follows the same design on entry level to make it easy for their machines to build them Only thing some reputed companies does is to order from them, with a custom voice coil and faceplate. So, the posted magnet assembly is the 3 screw version. The 4 screw ones is similar. Either a horn can be locked or a faceplate can be screwed in.

Tweeter is : https://www.madisoundspeakerstore.c...2606/9200-horn-loaded-1-textile-dome-tweeter/
If you look at https://www.aliexpress.com/item/NEW-DANMARK-VIFA-D26TG-05-TWEETER/1096613136.html its the same with the faceplate wihout the horn. In both cases the front and back are isolated.
 

Attachments

  • 1_1.jpg
    1_1.jpg
    200.5 KB · Views: 22
  • 2_2.jpg
    2_2.jpg
    185.7 KB · Views: 25
  • 3_3.jpg
    3_3.jpg
    241.1 KB · Views: 24
  • 4_$.jpg
    4_$.jpg
    302.1 KB · Views: 21
Last edited:
Since we are discussing tweeter integeration with a mid, is there a way to measure the impedance of a combo, say when I just top off a tweeter with midrange driver using just a cap ? I mean how do I know what kind of impedance the amp would be looking at ?

Why I ask is, my tube amp is happy driving 8 ohm impedance loads and I wish to experiment with some tweeter - midrange combos to play above 120hz (thru miniDSP) and wish to know if it’s safe to drive thru my tube amp.
120Hz is not too high a freq - your amp should be able to handle it when used only a capacitor. The best way is to use first a simulator software after simulating the driver FRD & ZMA files using the TS parameters and play around with it before actually connecting it physically. If you use Xsim freeware you can also see the Power response and the component power dissipation vs Freq. You can then purchase the required components. Alsso the amplifier power & impedance can be changed for you to get the correct value.

This tweeter must not be influenced by anything behind it. Just took out the scanpspeak H26TG series. There isnt a way for the tweeters waves to reach the rear after installing in a baffle. The assembly is same for the flat plate instead of horn, See the screw mounts. Vifa / scanspeak / seas follows the same design on entry level to make it easy for their machines to build them Only thing some reputed companies does is to order from them, with a custom voice coil and faceplate. So, the posted magnet assembly is the 3 screw version. The 4 screw ones is similar. Either a horn can be locked or a faceplate can be screwed in.

Tweeter is : https://www.madisoundspeakerstore.c...2606/9200-horn-loaded-1-textile-dome-tweeter/
If you look at https://www.aliexpress.com/item/NEW-DANMARK-VIFA-D26TG-05-TWEETER/1096613136.html its the same with the faceplate wihout the horn. In both cases the front and back are isolated.
I can only say this - The proof of the pudding is in the eating. I have done the rare chamber for the Vifa and have both lowered resonance and Q at resonance. I would open the second box this weekend and may post the impedance curve of the vifa tweeter with and without the chamber to keep the discussion simple.
 
Hari, my few tweeters have 4 to 8 ohm impedances and extended range drivers are also 4 to 8 ohm impedances.

I can measure TSP of the mid-extended range drivers but what would measuring TSP of tweeters reveal ? And in no way I can find out Fs of tweeters.

My measuring tool is DATS V2.

Further my desired XO between tweeter and mid-extended range would be 3khz and mostly 7-8khz.

I wish to use just a cap. Some experiments would be coaxially mounting the tweeter against the extended range driver.

Please advise. Thanks.
Connecting only the tweeter at 3KHz or 7KHz-8KHz is not safe and it will damage the amplifier as the impedance will roll-off at lower frequency. With the TSP you can simulate the FRD & ZMA files of the tweeter and then use it in simulation software for further evaluation. Check out subsim software for further details.
 
No, both would be connected to the amp. The XO point between the two would be 3khz and 7-8khz.

The tweeters TSP is unknown and they are not standard tweeters. The TSP are not available any more.
Try measuring the TSP using known volume method. In case of tweeter it will be approximately 0.
 
The ability for a tweeter to handle the lows is only depending on how much can the voice coil handle + the capability of the suspension to break + the ferrofluid inside - incase of vifa /scanspeak/seas. These are mechanical limitations. Doing whatever in the cabin however does not imrove these weaknesses for sure. Still cannot digest the idea of dampening the magnet from where the sound never comes out in a tweeter. (woofer , may be) But it will be one new thing learned. Certain brands like B & W does this, but its done because the diaphragm is not isolated from the rear mechanically.
Even l cannot .If back chamber of tweeter is totally closed,nothing should come out from back.ln case of woofer,basket is open on backside,so resonance is created in cabinet.If you hold hands around tweeter,you will hear different sound mostly thicker.Back side l dont understand this physics. I feel still chance of damaging tweeter on loud volume
 
I can only say this - The proof of the pudding is in the eating. I have done the rare chamber for the Vifa and have both lowered resonance and Q at resonance. I would open the second box this weekend and may post the impedance curve of the vifa tweeter with and without the chamber to keep the discussion simple.

Technically it would be weird to see that: because : Qes / Qms as per formular depends on Fs , Cms and Mms. Fs depends on Cms and Mms alone.

In a sealed tweeters case, the Cms ( Compliance of the driver's suspension ). In the case of Mms(Mass of the diaphragm/coil, including acoustic load ) with woofers it changes as the acoustic load depends on the free movement behind the driver. For a sealed tweeter, the Mms, must remain the same, - meaning a cabin should not alter the tweeters sound, but a front baffle shape can affect. Looking at your box, the wave guide like part can act like a horn to make it go lower. Or the pipe structure can reduce the overall free space inside the box, resulting in changing the acoustic load on the woofer. If the tweeter is taken out of the box and measured with and without the pipe, and if they are different, then that is something new and good to know.
No, both would be connected to the amp. The XO point between the two would be 3khz and 7-8khz.

The tweeters TSP is unknown and they are not standard tweeters. The TSP are not available any more.
Even l cannot .If back chamber of tweeter is totally closed,nothing should come out from back.ln case of woofer,basket is open on backside,so resonance is created in cabinet.If you hold hands around tweeter,you will hear different sound mostly thicker.Back side l dont understand this physics. I feel still chance of damaging tweeter on loud volume
A dome tweeters waves go quite well in all directions. So by holding your hands on the side or behind, you are changing the waves that travel through the surface there by changing the sound. BUT once you put it in a cabin, even if you keep you hands behind it should not change the sound. The waves across the surface of the face plate has no way to reach your hands but instead it will be bouncing off on the surface of the box's baffle.


I am not convinced about haris theory, I won't say it won't happen until I try. Theoretically it should not. If electrical mechanical specs of a closed tweeter gets altered depending on installation, it is quite a difficult thing to design the box.

There is no gap between the dome and magnet, if yes the tweeter will be distorting like hell due to the air from the woofer getting there. There is no gap between the dome and faceplate which connects to the interior of the box. If yes, the air from cabin would easily escape through this gap and act like a port for woofer. So, there is no way, by which the cone which is the only sounding thing on the tweeter is exposed to anything behind. The woofer may produce small amounts of sound from the magnet due to structural vibrarions due to the lower freq it is handling.

So now we can assume the tweeter unit safely as a sealed box which can be installed into any box, and what only can change its sound will be the front baffle design.

Now if you look at Wikipedia, as the FS and Q (es and ms) depends on Mms and Cms. Both are constant in our case, as we are not tampering with the dome. Acoustic load is a constant as the tweeter is sealed. Re is a constant as we are not changing the voice coil. Mathematically it doesn't seem to happen. But I don't know further things to be considered which I am ignorant of.
 

Attachments

  • 20180725_112339.jpg
    20180725_112339.jpg
    640.4 KB · Views: 4
  • 20180725_112353.jpg
    20180725_112353.jpg
    639.4 KB · Views: 4
Technically it would be weird to see that: because : Qes / Qms as per formular depends on Fs , Cms and Mms. Fs depends on Cms and Mms alone.


A dome tweeters waves go quite well in all directions. So by holding your hands on the side or behind, you are changing the waves that travel through the surface there by changing the sound. BUT once you put it in a cabin, even if you keep you hands behind it should not change the sound. The waves across the surface of the face plate has no way to reach your hands but instead it will be bouncing off on the surface of the box's baffle.
.
Exactly same thing I was saying.Only certain tweeters are developed with the extension tube like B&W do with specific purpose.
 
Get the Award Winning Diamond 12.3 Floorstanding Speakers on Special Offer
Back
Top