I am using neo3 pdr open backed on my speakers. However, I would like to cross them lower. So I am thinking of trying neo8s. The only problem with neo8s as tweeter is the vertical dispersion. So I am thinking of giving them a PDR treatment on the vertical axis. If you look at the picture, neo8s is pretty much like three neo3s stacked on each other. My plan is to cover upper and lower 1/3 of the driver with felt (just like neo3 PDR does horizontally. This will hopefully increase vertical dispersion with the only negative of some reduced efficiency. And then I will try to cross at around 800-1000 to a 8" cone midrange.
Has anyone tried this or something similar?
Can it work?
Is 800-1000 crossover point achievable?
And how much equalization am I going to need on the lower end of neo8s (baffleless)?
Has anyone tried this or something similar?
Can it work?
Is 800-1000 crossover point achievable?
And how much equalization am I going to need on the lower end of neo8s (baffleless)?
Attachments
Well, there is no question; neo8s would play down to 800 naked.
My question is whether it would play down to 800 after 2/3 of its front and back surface is covered with 1-2mm thick wool felt. This is definitely going to reduce the overall efficiency and will probably damp down the rising response.
My question is whether it would play down to 800 after 2/3 of its front and back surface is covered with 1-2mm thick wool felt. This is definitely going to reduce the overall efficiency and will probably damp down the rising response.
If you mean using wedges like in the RAAL, I think you can buy the wedges separately and give it a whirl. It is supposed to increase the vertical dispersion, but I don't know the parasitic effects these will induce, except for lower sensitivity.
Later,
Wolf
Later,
Wolf
My gut feeling is that it'll cause more problems than it solves, but clearly the only sensible way to proceed is to prototype it and measure. I'm just getting to this stage with the Neo 8/ RAAL 70-10D combo I'm building for a friend. When time allows I'm going to skill up on REW some more and use my new USB mic. to measure it all. There is the possibility that the Neo 8 might sound better for you without the foam, even if the vertical off-axis response is scary. On axis at least! Try it out, that's the best thing. Have a go with different thicknesses and profiles of foam too. Although not as simple as just using the '8 all the way up, it has to be worth trying it out crossed over to the '3, maybe somewhere in the 3-4K range? Less isn't always more!
Last edited:
Yes, it is the same idea RAAL uses in its taller tweeters.
I am a bit sceptical about the wedge shape though.
It pretends to mimic a waveguide.
But the more important effect probably comes from covering of the radiating surface making the driver physically shorter.
So, I think a layer of pure wool felt might be a better solution.
And yes, one can try different thickness and different angles.
If placed at an angle away from centre, it would look like a waveguide anyway.
Thanks to miniDSP, I blew one of my neo3s.
But if all does not go well, I can buy a new neo3 and cross them around 3k.
I am a bit sceptical about the wedge shape though.
It pretends to mimic a waveguide.
But the more important effect probably comes from covering of the radiating surface making the driver physically shorter.
So, I think a layer of pure wool felt might be a better solution.
And yes, one can try different thickness and different angles.
If placed at an angle away from centre, it would look like a waveguide anyway.
Thanks to miniDSP, I blew one of my neo3s.
But if all does not go well, I can buy a new neo3 and cross them around 3k.
My neo8s arrived in the weekend.
Initial experiment with compressed wool felt is promising.
I covered upper and lower 1/3 with a thin sheet of compressed wool felt.
Same on both front and back surface.
Think of an MTM: upper and lower sections of the driver are low-pass filtered physically (rather than an electrical crossover).
This theoretically will have two beneficial effects:
1. filter some of the HF output thus reduce the peak around 12kHz
2. HF comes from a 2/3 reduced radiating surface thus an increased vertical dispersion
The felt I am using is very dense. Next, I will experiment with different profiles of open cell foam to tweak the filter effect.
A small 90 degree waveguide is also in the making.
In terms of music reproduction I am not missing anything at the higher end: cymbals sound like cymbals. Perhaps there is some loss of "airiness". Once the 3 way speaker is optimized, I may add a supertweeter to see if there is any major benefit with doing so.
Initial experiment with compressed wool felt is promising.
I covered upper and lower 1/3 with a thin sheet of compressed wool felt.
Same on both front and back surface.
Think of an MTM: upper and lower sections of the driver are low-pass filtered physically (rather than an electrical crossover).
This theoretically will have two beneficial effects:
1. filter some of the HF output thus reduce the peak around 12kHz
2. HF comes from a 2/3 reduced radiating surface thus an increased vertical dispersion
The felt I am using is very dense. Next, I will experiment with different profiles of open cell foam to tweak the filter effect.
A small 90 degree waveguide is also in the making.
In terms of music reproduction I am not missing anything at the higher end: cymbals sound like cymbals. Perhaps there is some loss of "airiness". Once the 3 way speaker is optimized, I may add a supertweeter to see if there is any major benefit with doing so.
I'd like to do something similiar but I only need to XO at 2000 or 2500. I have been told that is just a little bit low on the NEO 3. Can you post a photo? Thanks.
Not to go off topic, but how did the MiniDSP blow your neo? I had a similar issue recently.
Cheers
Dean
Cheers
Dean
Dean,
I have a nanodigi from miniDSP.
It is basically a digital-in, digital-out DSP.
The order of channels are 1234 in their users manual.
The real order turned out to be 4321.
I ended up feeding fortified bass signal to the poor neo3.
I have a nanodigi from miniDSP.
It is basically a digital-in, digital-out DSP.
The order of channels are 1234 in their users manual.
The real order turned out to be 4321.
I ended up feeding fortified bass signal to the poor neo3.
JB,
Right now I have felt fabric squares, hold in place by rubber bands.
I will post pictures when I finalize the design.
Right now I have felt fabric squares, hold in place by rubber bands.
I will post pictures when I finalize the design.
experiences with the neo8-s
@zmyrna:
I am planning also with the BG neo8-s in a dipole project. How would you comment the audio experiences with the neo8-s?
The neo8 needs an equalization at 10kHz. How do you manage this?
At which xo frequence(s) do you separate the neo8-s?
Ruediger
@zmyrna:
I am planning also with the BG neo8-s in a dipole project. How would you comment the audio experiences with the neo8-s?
The neo8 needs an equalization at 10kHz. How do you manage this?
At which xo frequence(s) do you separate the neo8-s?
Ruediger
Here a good review of neo8: http://www.diyaudio.com/forums/multi-way/6228-bohlender-graebener-neo-8-magnetostat-2.html#post67297
That sounds promising for me.
The new neo8-s shall be a bit more better than the older neo8.
Ruediger
That sounds promising for me.
The new neo8-s shall be a bit more better than the older neo8.
Ruediger
Ruediger,
I tested Neo8S briefly without baffle.
It does not sound more peaky than Neo3.
And with the felt tweak it did not sound peaky at all.
I am now building a design similar to Martin Logan Mosaic (pic).
Neo8S will be dipole and horizontally placed covering 800 to 3000-4000Hz.
HF will be taken care of by Fountek CD3.5H (this design dictates a horn tweeter).
And I will definitely try the foam wedges on Neo8S to match dispersion at higher crossover point.
I tested Neo8S briefly without baffle.
It does not sound more peaky than Neo3.
And with the felt tweak it did not sound peaky at all.
I am now building a design similar to Martin Logan Mosaic (pic).
Neo8S will be dipole and horizontally placed covering 800 to 3000-4000Hz.
HF will be taken care of by Fountek CD3.5H (this design dictates a horn tweeter).
And I will definitely try the foam wedges on Neo8S to match dispersion at higher crossover point.
Attachments
Last edited:
I have to translate your English into German and sometimes I check LEO what´s the meaning of a special word. English - German Dictionary - leo.org: Start page
Is peaky equivalent to sickish or weak? This would be a bad result for the neo8.
How do you equalize the frequency response of neo8?
Do you make frequency measurements in 0 and 30 angle degree? Horizontal and vertical? Do you know these measurements? Please scroll down to the link wth the pdf document. http://www.audax-speaker.de/index.p...articles][data][shop_articles_id]=105&lang=en
What is the distance of neo8 backward to the wall? One meter is good, 1.5 meters are better.
Please consider: Neo8 as a magnetostat beams its sound vertical stronger than horizontal. User tomahack has built a line array with a very good sound result. http://www.diyaudio.com/forums/multi-way/6228-bohlender-graebener-neo-8-magnetostat-2.html#post67829
http://www.diyaudio.com/forums/mult...graebener-neo-8-magnetostat-4.html#post131084
Ruediger
Is peaky equivalent to sickish or weak? This would be a bad result for the neo8.
How do you equalize the frequency response of neo8?
Do you make frequency measurements in 0 and 30 angle degree? Horizontal and vertical? Do you know these measurements? Please scroll down to the link wth the pdf document. http://www.audax-speaker.de/index.p...articles][data][shop_articles_id]=105&lang=en
What is the distance of neo8 backward to the wall? One meter is good, 1.5 meters are better.
Please consider: Neo8 as a magnetostat beams its sound vertical stronger than horizontal. User tomahack has built a line array with a very good sound result. http://www.diyaudio.com/forums/multi-way/6228-bohlender-graebener-neo-8-magnetostat-2.html#post67829
http://www.diyaudio.com/forums/mult...graebener-neo-8-magnetostat-4.html#post131084
Ruediger
Last edited:
That tomahack speakers must have wild horizontal off-axis response, but vertical response is supposedly fine. Unfortunately we humans have our ears placed horizontally.... and we walk, not fly
Last edited:
The assumption is right by the alignment of line array. But the audio reviews of listeners speak for the line array of tomahack.
Ruediger
Neo8 does indeed have good HF extension.
But the problem with using it as a tweeter is its vertical dispersion.
Mind you, it is not much taller than the tall ribbon tweeters from Raal or Aurum Cantus.
I am pretty sure Neo8 would perform reasonably good as a tweeter if tweaked with the foam wedge method and crossed no lower than 1500 (ideally a driver should not cover more than 3 octaves - multi driver linearrays are different story).
By peaky, I mean the rising HF response and the peak at 12k Hz.
Someone has shown that the rising response of B&G planar drivers is partly because of the perforated front and back plates: every hole synergize with the one next to its neighbor. In theory, foam wedge method should also help reduce the rising response.
And my experiments with felt sheets tell me it is indeed the case; I have no measurements to prove it though.
But the problem with using it as a tweeter is its vertical dispersion.
Mind you, it is not much taller than the tall ribbon tweeters from Raal or Aurum Cantus.
I am pretty sure Neo8 would perform reasonably good as a tweeter if tweaked with the foam wedge method and crossed no lower than 1500 (ideally a driver should not cover more than 3 octaves - multi driver linearrays are different story).
By peaky, I mean the rising HF response and the peak at 12k Hz.
Someone has shown that the rising response of B&G planar drivers is partly because of the perforated front and back plates: every hole synergize with the one next to its neighbor. In theory, foam wedge method should also help reduce the rising response.
And my experiments with felt sheets tell me it is indeed the case; I have no measurements to prove it though.
- Status
- Not open for further replies.