Planar magnetic tweeter/mid with huge dispersion?

I made some weird planars with a 180 degree arc. and here are 2 tunes. where i show the dispersion, but besides that... its weird how it interacts with the room, when gated i get a flat response (relative) but then in real live its waay to mid/high heavy.
So i might need to tone that dispersion down so it does not interact so much with the room ? since it only adds up from 500hz to 10 khz or something.... (see measurements)
well who knows.. here it is

here are 2 measurements in the midle of the room one gated... and one not gated. and you can see there is quite some excesive mid/top added when not gated 5-9dB, while the top top end above 11khz does not add. witch makes the perception of me listening in the room pretty mid forward.... at least thats what i think. its sounds incorrect, while i love the stereo image.
i know there shoudl be a difference between the 2 but it looks excesive compared to flat panels. (i guess way more reflections because of the dispersion?)
 

Attachments

  • lengte kamer gated.jpg
    lengte kamer gated.jpg
    86.3 KB · Views: 77
  • lengte kamer niet gated.jpg
    lengte kamer niet gated.jpg
    87.2 KB · Views: 82
Last edited:
  • Like
Reactions: 3 users
Interesting tweeter. My own experiments with omnis led me to conclude that some forward directivity was good. I assume room treatments have a large impact on how much is needed.

Is it possible all you need is more output in the bass-mids to match efficiency?
 
Last edited:
Dispersion anxiety is wrong. If you have one seat, your only issue is how the sound reaches your one seat. Dispersion can be terrible, so long as you are happy at your seat. Fallacy of taking "birds' eye view" instead of "worm's eye view".

But you may ask, what about Toole's Spinorama that demonstrates the best sound at your seat arises (in his test room and including multiple seats) from speakers which meet certain standards for dispersion, total power distribution, and similar parameters?

Getting good sound to your seat includes the way sound across the full band is managed in your room. Most people do a kind of piece-meal micromanagement with reflections and absorbents and really just guesswork.

But if you go dipole, you can homogenize the sound and the result can often be wonderful and dipole fans (all ESL users) swear by it.

B.
 
  • Like
Reactions: 1 user
well yeah the reason for this dispersion experiement was in deed toole stuff :) and i am not sure whats nice. i think its somehwere in the middle. this is way to much and most flat panels beam to much...
so a 90 degree rotated version, might even work better curved in the vertical, and using a not to wide planar with low res in the horizontal plane. makes construction far more easy to
 
  • Like
Reactions: 1 users
well yeah the reason for this dispersion experiement was in deed toole stuff :) and i am not sure whats nice. i think its somehwere in the middle. this is way to much and most flat panels beam to much...
so a 90 degree rotated version, might even work better curved in the vertical, and using a not to wide planar with low res in the horizontal plane. makes construction far more easy to
Some good ideas there. About rotating the speakers, might be just changing one pattern for another.

Basically, inscrutable how to avoid setting up a system without REW - even using your laptop mic. No kidding. Once, using DSP or other hacks, you have a nice Toole freq response curve (a bit subdued treble, a bit boosted bass on your mic, THEN you trim to taste by ear (however deficient your hearing might be unknown to yourself). Simple as that.

Remember that Toole is giving good advice about visiting stores selling speakers (or how Harman InterGalactic Corp can make better speakers that suit more people). But for the DIY crowd, you are aiming to massage the sound in your room for your seat, not buying the "best guess" for $800.
 
You are trying to mate an omni directional radiator (the tweeter, or at least close to that) to a directional one (the woofer, operating at the point where it crosses over to the tweeter, will be strongly directional). If you match the on axis response, the power responses will be very different, with the omni radiating more acoustic power into the room. This would explain the feeling that your high frequencies are too hot. To balance the power responses at the crossover point is tricky because all you can do is vary the input power via input signal. If the power levels were balanced the on axis response would dip down at the crossover point going from low to high frequency. It's a difficult problem to overcome in some circumstances and is why it is a good idea to match the directivity of any two drivers around their crossover point.
 
  • Like
Reactions: 2 users
You are trying to mate an omni directional radiator (the tweeter, or at least close to that) to a directional one (the woofer, operating at the point where it crosses over to the tweeter, will be strongly directional). If you match the on axis response, the power responses will be very different, with the omni radiating more acoustic power into the room. This would explain the feeling that your high frequencies are too hot. To balance the power responses at the crossover point is tricky because all you can do is vary the input power via input signal. If the power levels were balanced the on axis response would dip down at the crossover point going from low to high frequency. It's a difficult problem to overcome in some circumstances and is why it is a good idea to match the directivity of any two drivers around their crossover point.
WElll the tweeter is crossed so low the woofer is acting like an omni as well. since i cross them at 450 hz 18dB. so i do not think thats the problem. i think the dispersion of the mid/tweet itself is the problem. adding to much in reflection in combination with the fact that flat panels often have a rising response but mainly because more surface area playing it reaches you ear. in this case thats not happening. (that might explain the need for adding top end :( )
 
Last edited:
  • Like
Reactions: 1 user
Basically, inscrutable how to avoid setting up a system without REW - even using your laptop mic. No kidding. Once, using DSP or other hacks, you have a nice Toole freq response curve (a bit subdued treble, a bit boosted bass on your mic, THEN you trim to taste by ear (however deficient your hearing might be unknown to yourself). Simple as that.

Remember that Toole is giving good advice about visiting stores selling speakers (or how Harman InterGalactic Corp can make better speakers that suit more people). But for the DIY crowd, you are aiming to massage the sound in your room for your seat, not buying the "best guess" for $800.
I think i need to get the room response and gated measurement a bit closer together if thats even possible. and i think it can. either by making it slightly less high dispersion or make the acoustics perfect.... witch will never happen.}
a bit of both would be nice :)
 
I just orderd some laser cut steel plates, to be used with 50mm long magnets so i can make a panel flat by adding a bunch. or curve it if i want to, and just add plates when i feel it has to be longer.. the steel is 1.25 mm thick and 2 mm longer then the magnets are. just to be save
i ordered some cutters for the CNC to make 1.25-1.4 mm wide pockets , to put the sides of these plates into a piece of MDF to make up the curved panel. this way i can use normal N35 magnets if i make it vertical.

1694397493185.png
 
Well done with this tweeter!
In the past, maybe still available, Elac made 360 degrees ribbontweeter, called it pi-tweeter. The magazin A&T tested it and it was supposed to give very good results unless the speaker had a hard reflecting (glass) close at the back of the speaker. In other words positioning such speaker can be critical and my guess is to start putting such speakers far away from boundaries and see/listen how things turn out when putting them closer.
 
  • Like
Reactions: 2 users
Any upper range driver mounted facing the ceiling is kind of 360-degrees, esp with a reflector facing down over it.

There is a somewhat odd AES publication from Linkwitz about listening to music in his own house. He says his experience got a whole lot better when he mounted a tweeter facing backwards. I don't recall if he matched his subjective assessment with measurements too.
 
Thanks much for tutorials. Sounds quite good, although I am hearing it 6 steps removed from the sound in your room. Good test track (I use "classical" percussion music, such as Paul Lansky's "Threads".... once a person is used to percussion on ESLs or film drivers, hard to listen on voice-coil drivers).

B.
 
so last tune before diy event.. (after playing with the DSP for ages.... they are not at all without problems). still not perfect and i just discovered i should tilt them forward instead of backwards... but could not make noise anymore... because of neighbours. so i guess we try that on the event it self. i mean its an experiment after all.