While looking pretty and I'm not arguing the benefits of smooth dispersion, the waveguides were not panacea of all problems.
If it was all down to dispersion smoothness, there are plenty solution which are better than XT25+wg.
diyaudio.com/forums/multi-way/178187-great-waveguide-list.html
I don't know why waveguided-tweeters sound harsher than plain ones.
Dare I say it ... HOM ?
If it was all down to dispersion smoothness, there are plenty solution which are better than XT25+wg.
diyaudio.com/forums/multi-way/178187-great-waveguide-list.html
I don't know why waveguided-tweeters sound harsher than plain ones.
Dare I say it ... HOM ?
I don't know why waveguided-tweeters sound harsher than plain ones.
Dare I say it ... HOM ?
I definitely hear nothing degraded from it. I'm going to buy another set to convert my other speakers to WG's. I can send the XT25 attached to Dave's WGs to you for testing and a subjective evaluation. Just dont blow them up and forward them back when you're done.I like the BMS 4550 and wish I could hear a set before I commit $400 with WGs for them.
I do agree but then nothing is. Like all things in life, it's all a question of compromises.
I don't know why waveguided-tweeters sound harsher than plain ones.
Dare I say it ... HOM ?
Strangely, I'm getting the opposite
. My XT25TG30 sounds smoother with Dave's Elliptical waveguide compared to the standard faceplate.I say mine sound better too but they are more transient/dynamic. I guess that would be less smooth in a less relaxed way but not harsh in any way.
No way I will not use a WG on every speaker from now on. I do have a dipole I need to bolt up that has a Neo3 PDR dipole tweeter that wont have a WG though
No way I will not use a WG on every speaker from now on. I do have a dipole I need to bolt up that has a Neo3 PDR dipole tweeter that wont have a WG though
I'd recently done a two way with a 6" mid woofer and an elliptical 165mm x 15mm x 37mm deep. I originally had it crossed at 1.6 kHz. It sounded harsh to me. I moved the crossover up to 2.1 kHz and it just seemed to come to life. I wonder if there is a tendency to push the X over to low with a waveguide.
I'm not so sure about the HOM at least in these shallower guides. What I have seen, which is probably the same thing, is what's been referred to as bunching. It's usually around 5kHz and 9kHz where the on axis will dip down below the 30 degree axis or right at it. It seems to flatten out by 30 degrees. This is pretty common in the round guides. The elliptical guides don't seem to do it or if it does not nearly as bad. If you go to my website and click on the Picasa link you can see measurements of tweeters with elliptical and round guides and you can see the difference.
I'm not so sure about the HOM at least in these shallower guides. What I have seen, which is probably the same thing, is what's been referred to as bunching. It's usually around 5kHz and 9kHz where the on axis will dip down below the 30 degree axis or right at it. It seems to flatten out by 30 degrees. This is pretty common in the round guides. The elliptical guides don't seem to do it or if it does not nearly as bad. If you go to my website and click on the Picasa link you can see measurements of tweeters with elliptical and round guides and you can see the difference.
It's likely there's some form of distortion from the tweeter. I just finished a 6.5" Silver Flute with your waveguide. Crossed the xt25tg30 at 2.5kHz (12dB/oct). No issues. I'm quite confident it can go as low as 2.0kHz with 18dB/oct. Under 2.0kHz, I prefer to use compression drivers.
Below are the distortion plots of the xt25tg30 with your waveguide. Top plot is the distortion with the raw xt25. A 170uF was inserted for tweeter protection. Quite a horrendous amount of distortion.
The second plot is with a 2.5kHz, 12dB network. Distortion is completely suppressed. What's interesting is no LCR was needed.
Ive done a few quick sketches, Im guessing the guide "a" is what yours resemble spinmonster?
Funny, Id have thought that b or c would have given a better low end considering they encompass the outer ring for what I presume is a greater SD?
Ignore d, its an ongoing project..
A has a 25mm throat,
B has a 32mm throat
C has a 40mm throat
Funny, Id have thought that b or c would have given a better low end considering they encompass the outer ring for what I presume is a greater SD?
Ignore d, its an ongoing project..
A has a 25mm throat,
B has a 32mm throat
C has a 40mm throat
Attachments
Last edited:
Interestingly Tymphany has these 2 tweeters listed, spec sheets indicate they were within a month of each other.
The first, XT25SC90-04, is listed with .05mm excursion.
http://www.tymphany.com/files/XT25SC90-04 Rev1_0.pdf
The second, XT25SC50-04, is listed with .1mm excursion.
http://www.tymphany.com/files/XT25SC50-04 Rev1_0.pdf
All the other specs, including the FR graph look identical or within testing tolerances, I wonder why the change?
The first, XT25SC90-04, is listed with .05mm excursion.
http://www.tymphany.com/files/XT25SC90-04 Rev1_0.pdf
The second, XT25SC50-04, is listed with .1mm excursion.
http://www.tymphany.com/files/XT25SC50-04 Rev1_0.pdf
All the other specs, including the FR graph look identical or within testing tolerances, I wonder why the change?
Have tried WG with every tweeters I have had since several years ago. It is not yet a WG. Beside other approaches, I have bought many woofer cones and modified the outer end to see if I will have luck with any shape of them.
What I don't like with WG-ed tweeter is I cannot make the tweeter disappear (this is more subtle than saying it harsher).
A 165mm wave guide only has lift to 13560/6.5" or 2086hz before it starts to slope down so its too small a wave guide for crossing under 2.1k or so. The idea of lower distortion with a WG requires to lift to be exaggerated lower so that you use a cap starting at about 5K or so would have 9db or so attenuation at 2k to get the power response back to flat. The cap isnt starting at 1700 or 2100, its starting at 5-6k.
To do this you need an 8" WG. At 8" the lift is out to 13560/8 or 1695hz for a cross there. Since the F2 distortion at 1700 is based on what is happening at 850hz, the attenuation from the capacitor is >20db by then lowering the F2 at 2000hz by >20db. You are also lowering the resonance bell at the res frequency by about 22db before you even apply a crossover. Wave guides arent about lifting response everywhere. Its about getting low end lift to couter with a cap to get back to flat. I use EQ up top to shape the response after the capacitor lowers response and distortion.
Using a 2nd or 3rd order cross isnt going to have the attenuation I have by using a 5th order electrical cross (or 9th order if I choose since its active on my speaker) so yes, it would be more harsh. I have a TMM system and the lower woofer has a coil for early attenuation.
A few other comments, the XT25tg30 is inferior in every way to the XT25tg60. The addition of the second magnet increases efficiency as also lowers the noise floor resulting in a much lower distortion from 1200 to 2000hz and it measures cleaner. The 5 additional bucks is well worth the price for being able to lower the noise floor and increase SPL in relation to the distortion. Distortion is created by lower frequencies so attenuation of distortion at 1700 is dependent upon what you have going on at 850 for F2. If you need to reduce the tweeter's output to match a midbass's efficiency, you want to start with a 93db sensitive tweeter and use a resistor because lowering the sensitivity with the restor 3db os so is also lowering the distortion by that same amount. the more sensitive version is also that much more head room above the noise floor with a stronger BL.
It looks to me your set up:
1-used the worse 30 version
2-used too small a WG for the lower cross.
3-used too shallow a slope
4-didnt use a cap starting at 5k or so with EQ to counter the rest
Last edited:
No 2 wave guides sound.
The tweeter disappearing is exactly what I hear with this combo and I own other honky sounding horns. If you don't EQ, I cant see it working unless its a copied system that measures perfectly like Zaph's TMM. If you use a WG and dont have EQ, much of what brings attention to the highs would then be freq response irregularities. As hard as system design is, a WG is another complex variable and it isnt solved by using one and then not liking it. All my friends that heard mine cant believe its using the cheap drivers it does and 2 are copying it exactly as I built it with the same WG.
I've heard bad sounding WG's too.
Found this interesting tweeter today, only a very small guide so it doesnt load low at all.
What is interesting is the throat, seems to be on the top of the outer surround,similar to my "b" drawing from the last page. Seems intuitive to me that it would be there (not that it means anything)
With the VC in the valley of the dual surrounds it would make sense that they are both driven equally.
http://www.tymphany.com/files/XT19NC30-04 Rev1_0.pdf
What is interesting is the throat, seems to be on the top of the outer surround,similar to my "b" drawing from the last page. Seems intuitive to me that it would be there (not that it means anything)
With the VC in the valley of the dual surrounds it would make sense that they are both driven equally.
http://www.tymphany.com/files/XT19NC30-04 Rev1_0.pdf
Divide 13560 by the number of inches wide to get a shallow wave guides lowest frequency where it starts rolling off. Depth also lowers the frequency. This doesn't mean you cant cross under that freq. Its a ball park.
A 6.5" WG thus begins to roll off its lift at 13560/6.5 = 2000 to 2100. A 10" WG boosts to 1300 or so. Deeper versions will reach a little lower but may be uneven in freq response. EQ fixes the honk that may be from too deep a guide. Just like a subwoofer, a guide that wont 'disappear' is because the freq response isn't even and it needs EQ.
If you have a 9db boost at 2000hz, you use a cap at 6000hz to counter the rise and get it back to flat. When you do so a cap at 6000 has you flat at 2000 but by 500hz the cap has you down 21db. F3 distortion at 2000hz is caused by activating 500hz which is then suppressed. F2 distortion is caused by whats happening at 1000hz. All of the distortion is then suppressed by around 20db or its as if you hit the mute button just for distortion.
A 6.5" WG thus begins to roll off its lift at 13560/6.5 = 2000 to 2100. A 10" WG boosts to 1300 or so. Deeper versions will reach a little lower but may be uneven in freq response. EQ fixes the honk that may be from too deep a guide. Just like a subwoofer, a guide that wont 'disappear' is because the freq response isn't even and it needs EQ.
If you have a 9db boost at 2000hz, you use a cap at 6000hz to counter the rise and get it back to flat. When you do so a cap at 6000 has you flat at 2000 but by 500hz the cap has you down 21db. F3 distortion at 2000hz is caused by activating 500hz which is then suppressed. F2 distortion is caused by whats happening at 1000hz. All of the distortion is then suppressed by around 20db or its as if you hit the mute button just for distortion.
Last edited:
He tried a ton of variables. The various tweeters have multiple WG sizes and he posted graphs for all of them. He only produces what worked.
Click on the tweeter and see the graphs for the various WG sizes for that tweeter. You cant buy what isnt available so what Dave did was a gift to DIY but you are limited to what tweeters he tried that worked.
https://plus.google.com/photos/101632266659473725850/albums?banner=pwa
Click on the tweeter and see the graphs for the various WG sizes for that tweeter. You cant buy what isnt available so what Dave did was a gift to DIY but you are limited to what tweeters he tried that worked.
https://plus.google.com/photos/101632266659473725850/albums?banner=pwa
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.