Hi,
Because of my recent frequency response measurements, I have now concluded
that in my speakers the crossover for the tweeter is right
(despite it was fully tuned by ear without measurements) .
Therefore if someone wants
to have a very simple crossover for the tweeter and wants to build a similar
waveguide than in my speakers (look at my speaker forum for more details),
simply one 2.2 uF polypropylene capacitor (and depening on the other
drivers a resistor in seris) in series with the tweeter
will do the trick assuming that the circular waveguide is exactly the same shape than
mine. Waveguide simplifies the crossover a lot. The boost to the midrange
area with the waveguide is in other words quite exactly 6 dB / octave towards
lower frequencies and therefore the 6 dB/octave high pass will straighten the
response to flat.
So if you need a decent tweeter without ferrofluid and that can be crossed as low as
at 1.5 kHz without any problems (without sacrificing distortion, clarity etc.),
I think the Seas 25TAC/GW is a good choice if accompanied
with the waveguide. The width of the waveguide is 17 cm and depth is
about 5 cm. You can go to my loudspeaker forum to look the picture.
There must not be any gap between the waveguide and the tweeter, so
be sure to mount it tightly.
I wonder why only few DIY speaker projects utilize waveguide since
their benefits are quite evident once you hear how it sounds like.
After that you may not ever want to go back to a waveguideless speaker again...
The secret of the soundstage and accuracy in my speakers is the low
crossover point and the waveguide. Without them, they would be just yet
another speakers without anything really much impressive,
but with the waveguide, they are pretty damn good.
Best Wishes,
Karoliina
Because of my recent frequency response measurements, I have now concluded
that in my speakers the crossover for the tweeter is right
(despite it was fully tuned by ear without measurements) .
Therefore if someone wants
to have a very simple crossover for the tweeter and wants to build a similar
waveguide than in my speakers (look at my speaker forum for more details),
simply one 2.2 uF polypropylene capacitor (and depening on the other
drivers a resistor in seris) in series with the tweeter
will do the trick assuming that the circular waveguide is exactly the same shape than
mine. Waveguide simplifies the crossover a lot. The boost to the midrange
area with the waveguide is in other words quite exactly 6 dB / octave towards
lower frequencies and therefore the 6 dB/octave high pass will straighten the
response to flat.
So if you need a decent tweeter without ferrofluid and that can be crossed as low as
at 1.5 kHz without any problems (without sacrificing distortion, clarity etc.),
I think the Seas 25TAC/GW is a good choice if accompanied
with the waveguide. The width of the waveguide is 17 cm and depth is
about 5 cm. You can go to my loudspeaker forum to look the picture.
There must not be any gap between the waveguide and the tweeter, so
be sure to mount it tightly.
I wonder why only few DIY speaker projects utilize waveguide since
their benefits are quite evident once you hear how it sounds like.
After that you may not ever want to go back to a waveguideless speaker again...
The secret of the soundstage and accuracy in my speakers is the low
crossover point and the waveguide. Without them, they would be just yet
another speakers without anything really much impressive,
but with the waveguide, they are pretty damn good.
Best Wishes,
Karoliina
Karoliina,
I've also become a believer in Waveguides after a recent set of tests I've done. Very simple crossovers do indeed work well.
For someone looking for a full set of measurements and construction techniques, I have this available. It uses a Lexan horn converted into a waveguide. It's pretty easy to modify for waveguide use.
http://www.zaphaudio.com/hornconversion.html
I've also become a believer in Waveguides after a recent set of tests I've done. Very simple crossovers do indeed work well.
For someone looking for a full set of measurements and construction techniques, I have this available. It uses a Lexan horn converted into a waveguide. It's pretty easy to modify for waveguide use.
http://www.zaphaudio.com/hornconversion.html
Oops, I thought this was a new thread, but now I see that it was a year old and just brought to the top with a reply.
I need to check those original posting dates. I feel like a dork now. Karoliina is probably not even paying attention to this forum anymore.
Paul, you must have had to dig deep to find this one.
I need to check those original posting dates. I feel like a dork now. Karoliina is probably not even paying attention to this forum anymore.
Paul, you must have had to dig deep to find this one.
John, yes I've been doing some searching on a number of different topics, I'm on a learning curve with some ideas I've been working on. I saw Karoliina's waveguides a long time ago, and then more recently yours. Some time ago I turned up some crude attemps at waveguides for my Vifa dome tweeter, but I didn't have the means to do any measuring and couldn't pick a difference that I liked. Its more recently I've learnt that it only works with some eq, where the waveguide reduces excursion at the low end, hence reducing distortion, increasing power handling and potential output in addition to directivity control. The advantages are so many that it makes me wonder why it is not done more often.
waveguides
I have built a test waveguide driven by a 25mm. dome tweeter.
This is essentially a conical horn with a flared mouth. The mouth flare was as far as I know first mentioned by Keele in a AES preprint, and allows constant directivity down to a lower frequency with a given mouth diameter. The flare has the secondary effect of reducing finite apperture diffraction that occurs at the upper end of the frequency range.
Geddes has published a proposal for a mathematical method to calculate this flare, but from my crude calculations it seems that a simple radius that matches the wall angle at one end, and the baffle at the other is as good as anything.
I have built a test waveguide driven by a 25mm. dome tweeter.
This is essentially a conical horn with a flared mouth. The mouth flare was as far as I know first mentioned by Keele in a AES preprint, and allows constant directivity down to a lower frequency with a given mouth diameter. The flare has the secondary effect of reducing finite apperture diffraction that occurs at the upper end of the frequency range.
Geddes has published a proposal for a mathematical method to calculate this flare, but from my crude calculations it seems that a simple radius that matches the wall angle at one end, and the baffle at the other is as good as anything.
So how does one determine the dimensions to use and what relationship they have to what you want to achieve? (ie point at which directivity is controlled down to ...)
I don't follow your description of the radius ...
Baffle shown yellow ...
On the right the flare is basically a radius, and on the left it is stretched in the x dimension and the line shows the starting tangent of the curve which is 20degrees from the other line drawn ...
Perhaps you can clarify ...
I don't follow your description of the radius ...
Baffle shown yellow ...
On the right the flare is basically a radius, and on the left it is stretched in the x dimension and the line shows the starting tangent of the curve which is 20degrees from the other line drawn ...
Perhaps you can clarify ...
Attachments
I did a design prior to all the recent interest. The raw measurements didn't look too good but the aberations were fixable by the transfer function of the crossover.
Code:
[IMGDEAD]http://lowfat.sytes.net/~mike/projects/ono/P3120354sm.jpg[/IMGDEAD] [IMGDEAD]http://lowfat.sytes.net/~mike/projects/ono/P3120360-sm.jpg[/IMGDEAD] [IMGDEAD]http://lowfat.sytes.net/~mike/projects/ono/P3120358-sm.jpg[/IMGDEAD]
Zaph, did you notice in your tests that the midrange sounds more substantial after cd horn loading of a dome tweeter in a system?
After the needed ironing of the inflated mids and the gains in simplicity of suitable crossover and distortion, I guess the best element is that the dispersion shows continuity with a cone at crosspoint. This eliminates the flare in off axis power response and sounds fuller to me.
After the needed ironing of the inflated mids and the gains in simplicity of suitable crossover and distortion, I guess the best element is that the dispersion shows continuity with a cone at crosspoint. This eliminates the flare in off axis power response and sounds fuller to me.
Mac! Where have you been? 
Looks like you (finally) made some progress with that thing. Good to see. I'm still no MTM fan, even TM really, but hey, sometimes thats what's feasible. Looks like you really recessed the mids - for time alignment perhaps? Or are those the smallest controlled directivity horns I've ever seen?
Bottom line, does it sound good, or am I in Audiophilic dreamland thinking it (WG XT19) sounds better than the Millennium? I've always claimed its more personal taste than anything. What do you think? I ceased development on mine a while back. Taking a different turn.
BTW, totally off topic, but I'm using the single best purchase I've made in audio up to date. I've got Beyond Jazz (ch 71) cranking off my XM skyfi receiver hooked to my computer speakers (dipoles - of course, JBL P422's with those little coaxial tweets in a tiny waveguide/horn) right now as I type. Don't surf without it!
Cheers,
AJ
current/old state of affairs
sign of things to come
Looks like you (finally) made some progress with that thing. Good to see. I'm still no MTM fan, even TM really, but hey, sometimes thats what's feasible. Looks like you really recessed the mids - for time alignment perhaps? Or are those the smallest controlled directivity horns I've ever seen?Bottom line, does it sound good, or am I in Audiophilic dreamland thinking it (WG XT19) sounds better than the Millennium? I've always claimed its more personal taste than anything. What do you think? I ceased development on mine a while back. Taking a different turn.
BTW, totally off topic, but I'm using the single best purchase I've made in audio up to date. I've got Beyond Jazz (ch 71) cranking off my XM skyfi receiver hooked to my computer speakers (dipoles - of course, JBL P422's with those little coaxial tweets in a tiny waveguide/horn) right now as I type. Don't surf without it!
Cheers,
AJ
current/old state of affairs
An externally hosted image should be here but it was not working when we last tested it.
sign of things to come
re waveguide
What I did Paul is to have a look at several papers, the Manta Ray horn one by Hendricksen and Ureda, ones by Geddes, one by Johansen.
From Keeles paper he gives the the frequency down to which a cd horn will have costant directivity as , (in metric units).
f3=25000/(a*w)
where a = included horn angle and w = width in meters,
for a 1500Hz. waveguide this is around 185mm. in diameter.
By putting a mouth with two flares of greater angles, this lower frequency becomes the simple average of the cut off frequencies of all of the flares, replacing these two sections with a continuous curve is, as shown by Johansen, the same as using two conical sections and there is no abrupt transition that cause diffraction.
From his data if you maintain a conical section to about .65-.7 of the final mouth diameter, and use two conical sections thereafter this gives the overall smoothest result.
In my design I used a radius to connect the conical section to the baffle, if it meets both at right angles the slope of the radius at that point is the same as the lines it joins and there is no discontinuity.
My final result has a mouth diameter of about 150mm. and on axis has a frequency responce that falls at 6db. per octave above the tweeters 1500Hz. resonant frequency.
Feeding from a first order high pass filter flattens the responce and gives in the region of 15-18db. attenuation at around 1500Hz., you can expect the power handling at 1500Hz. to be increased by at least this amount.
Off axis measurements show that it has constant directivity over a 90 degree angle over the range of 1500-15000Hz., within plus or minus 3db.
What I did Paul is to have a look at several papers, the Manta Ray horn one by Hendricksen and Ureda, ones by Geddes, one by Johansen.
From Keeles paper he gives the the frequency down to which a cd horn will have costant directivity as , (in metric units).
f3=25000/(a*w)
where a = included horn angle and w = width in meters,
for a 1500Hz. waveguide this is around 185mm. in diameter.
By putting a mouth with two flares of greater angles, this lower frequency becomes the simple average of the cut off frequencies of all of the flares, replacing these two sections with a continuous curve is, as shown by Johansen, the same as using two conical sections and there is no abrupt transition that cause diffraction.
From his data if you maintain a conical section to about .65-.7 of the final mouth diameter, and use two conical sections thereafter this gives the overall smoothest result.
In my design I used a radius to connect the conical section to the baffle, if it meets both at right angles the slope of the radius at that point is the same as the lines it joins and there is no discontinuity.
My final result has a mouth diameter of about 150mm. and on axis has a frequency responce that falls at 6db. per octave above the tweeters 1500Hz. resonant frequency.
Feeding from a first order high pass filter flattens the responce and gives in the region of 15-18db. attenuation at around 1500Hz., you can expect the power handling at 1500Hz. to be increased by at least this amount.
Off axis measurements show that it has constant directivity over a 90 degree angle over the range of 1500-15000Hz., within plus or minus 3db.
Mac, that looks VERY interesting ...
Could you show us some measurements, I'd be very interested to see, it has 2 things I'd like to try - dipole mids and waveguide (both I have tried, but not yet properly)
Are the mids attached via the metal device on the back?
Do you consider that the fairly large magnets are a problem regarding the rear sound wave?
Shown in the 2nd pic?
Are you saying they are better than your seas/vifa WG dipole?
Could you show us some measurements, I'd be very interested to see, it has 2 things I'd like to try - dipole mids and waveguide (both I have tried, but not yet properly)
Are the mids attached via the metal device on the back?
Do you consider that the fairly large magnets are a problem regarding the rear sound wave?
Shown in the 2nd pic?
Are you saying they are better than your seas/vifa WG dipole?
Hi Paul,
no, I'm listening to these JBL's:
The 2nd pic is a JBL 328C. Its a (8") coaxial driver, mounted in a waveguide. I'm working on something similar, but with a larger (4" coaxial mounted) tweeter WG within a 8" (Dynaudio) driver - that is itself mounted in a WG - like the JBL. I'm using a (home) ring radiator tweeter, not a pro-type compression driver. Controlled directivity of the entire (semi) spherical wavefront. I have no idea whether it will sound better than my current setup, its not close to being done, but I'm a DIYer remember? I can't just leave well enough alone .
Now, a quick note. My projects take months to years. They are excruciatingly slow due to work/home type/neighbors,etc.
I'll do the best I can to post stuff, but there will be no Zaph type website any time soon
I do have measurement capability, but taking the time to coherently put it all together and post it etc, I don't have currently. Maybe in the future.
Cheers,
AJ
p.s. something kinda like this - mounted in the JBL type guide, dipole of course, which already has CD because of the side nulls.
no, I'm listening to these JBL's:
The 2nd pic is a JBL 328C. Its a (8") coaxial driver, mounted in a waveguide. I'm working on something similar, but with a larger (4" coaxial mounted) tweeter WG within a 8" (Dynaudio) driver - that is itself mounted in a WG - like the JBL. I'm using a (home) ring radiator tweeter, not a pro-type compression driver. Controlled directivity of the entire (semi) spherical wavefront. I have no idea whether it will sound better than my current setup, its not close to being done, but I'm a DIYer remember? I can't just leave well enough alone
.Now, a quick note. My projects take months to years. They are excruciatingly slow due to work/home type/neighbors,etc.
I'll do the best I can to post stuff, but there will be no Zaph type website any time soon
I do have measurement capability, but taking the time to coherently put it all together and post it etc, I don't have currently. Maybe in the future.
Cheers,
AJ
p.s. something kinda like this - mounted in the JBL type guide, dipole of course, which already has CD because of the side nulls.
An externally hosted image should be here but it was not working when we last tested it.
Here's a link to the JBL site (56kwarning - 6MB file)
http://www.jblpro.com/Control300/PDF/CCBRO2005_CRA_trim.pdf
I don't like the idea of using the woofers cone as a WG (ala Tannoy,etc.), since it moves in relation to the tweeter. But there is something about the coherent, symmetric type wavefront produced by a more point source like driver, like the JBLs - that I like. I think (use of) DSP will be more effective with a coincident driver arrangement such as this. Hence no more development of my T(WG)M Orion main panel. Cheaper and better DSP in in our near future. Just thought I'd be prepared .
Cheers,
AJ
http://www.jblpro.com/Control300/PDF/CCBRO2005_CRA_trim.pdf
I don't like the idea of using the woofers cone as a WG (ala Tannoy,etc.), since it moves in relation to the tweeter. But there is something about the coherent, symmetric type wavefront produced by a more point source like driver, like the JBLs - that I like. I think (use of) DSP will be more effective with a coincident driver arrangement such as this. Hence no more development of my T(WG)M Orion main panel. Cheaper and better DSP in in our near future. Just thought I'd be prepared
.Cheers,
AJ
Ralph, Paul, AJ - hello.
Some of the inspiration came from AJ's Orion clones. I used the waveguide that AJ and others have used from MCM Electronics. In a nutshell, you have to mill down the throat of the guide, fabricate an adapter (of MDF) and then possibly make a new faceplate for the tweeter (of 1/8" aluminum). The tweeter I used was the Vifa XT19.
Here's a link to the raw measurements. This blurry photo shows how I milled the guide using a router jig.
Paul - the RS180's aren't ideal for dipole use due to their somewhat closed back design. However, they do sound good and I'm pleased with them. They are pressure mounted by the steel spline from the rear. Linkwitz used a similar mounting scheme for his Phoenix, which greatly reduces baffle resonances.
Cheers, mac.
Some of the inspiration came from AJ's Orion clones. I used the waveguide that AJ and others have used from MCM Electronics. In a nutshell, you have to mill down the throat of the guide, fabricate an adapter (of MDF) and then possibly make a new faceplate for the tweeter (of 1/8" aluminum). The tweeter I used was the Vifa XT19.
Here's a link to the raw measurements. This blurry photo shows how I milled the guide using a router jig.
Paul - the RS180's aren't ideal for dipole use due to their somewhat closed back design. However, they do sound good and I'm pleased with them. They are pressure mounted by the steel spline from the rear. Linkwitz used a similar mounting scheme for his Phoenix, which greatly reduces baffle resonances.
Cheers, mac.
There's a line of pro speaker I just noticed that uses a 1" dome with a waveguide. Only $6,995 for the MTM.
http://www.4sptech.com/products.php
http://www.4sptech.com/products.php
re coaxial waveguides
My currant speaker system uses coaxial wave guides.
This uses a 5 inch polycone driver with a center mounted Neodimium tweeter sold by Jaycar.
I made a small rectangular waveguide for this, and then mounted the whole thing within a larger rectangular wave guide.
I reasoned that the tweeter horn would represent a small obsticle if the crossover was in the 2.5-3kHz. region, and measurements confirm this.
The idea is to get 40 degree vertical directivity over 1000Hz. plus, and 90 degree horizontal over 300Hz. plus.
People who hear the system remark upon the image, and upon the dynamic range, it goes very loud with no signes of distress, this is due to the midrange driver having around 15db. of attenuation at 300Hz. plus an overall efficiency improvement by4db.
Drawbacks? it is big.
My currant speaker system uses coaxial wave guides.
This uses a 5 inch polycone driver with a center mounted Neodimium tweeter sold by Jaycar.
I made a small rectangular waveguide for this, and then mounted the whole thing within a larger rectangular wave guide.
I reasoned that the tweeter horn would represent a small obsticle if the crossover was in the 2.5-3kHz. region, and measurements confirm this.
The idea is to get 40 degree vertical directivity over 1000Hz. plus, and 90 degree horizontal over 300Hz. plus.
People who hear the system remark upon the image, and upon the dynamic range, it goes very loud with no signes of distress, this is due to the midrange driver having around 15db. of attenuation at 300Hz. plus an overall efficiency improvement by4db.
Drawbacks? it is big.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.