Everyone has a different idea of what "balancing tradeoffs" means.. If someone is supremely happy with his results, even after attempting to "bend" or in some cases even "break" the laws of physics, then the design is an unqualified success..For him. By the same token, If someone respects and obeys all of the physical laws and ends up hating the results, the design is an unmitigated failure..For him. Ultimately, the success of a design is purely subjective. But I do believe that a design that attempts to weave gently through the compromises without bending, bruising, or denting any of them has a better chance of success than one that doesn't.
Seth
Seth
Last edited:
Latest Pictures and description.. Pictures by remlab - Photobucket
Computer based system consisting of;
Computer
Squeezebox touch
Behringer DEQ(For measurements only), Behringer SRC, and two Behringer DCX's(One per channel).
Signal stays in the digital domain until the DCX output..
Four subwoofer amps purchased from Jack Hidley(Internal crossovers bypassed)
Three Behringer A500's
Dayton cabs
Speaker bases(For main's) made from 10" Dayton subwoofer baffles.
Peerless XXLS's
Peerless HDS Exclusives
BG Neo 10's and 3's
BG Neo's attached with Dayton gasket putty only. No screws..
T-6 Aluminum channels covered with acoustic vinyl. (Prototypes)
Xover points(8th order); 75 150 450 1,800 (Constantly Evolving) Best configuration so far..
Seth
Computer based system consisting of;
Computer
Squeezebox touch
Behringer DEQ(For measurements only), Behringer SRC, and two Behringer DCX's(One per channel).
Signal stays in the digital domain until the DCX output..
Four subwoofer amps purchased from Jack Hidley(Internal crossovers bypassed)
Three Behringer A500's
Dayton cabs
Speaker bases(For main's) made from 10" Dayton subwoofer baffles.
Peerless XXLS's
Peerless HDS Exclusives
BG Neo 10's and 3's
BG Neo's attached with Dayton gasket putty only. No screws..
T-6 Aluminum channels covered with acoustic vinyl. (Prototypes)
Xover points(8th order); 75 150 450 1,800 (Constantly Evolving) Best configuration so far..
Seth
Last edited:
Greg
Good question.. An unequalized dipole Neo 3's response rises with frequency. http://www.k4revenge.com/Mat/new/bgneo3pdr.pdf
The rise compensates almost perfectly for the lower axis I'm listening at..
Seth
I found a store that sells Igor Levitsy's horn loaded SLS PRD 500 design. These are typically not available to the public..
Speaker Exchange | SLS Ribbon Tweeter PRD500
Speaker Exchange | SLS Ribbon Tweeter PRD500
I haven't read all this thread but I wanted to tell you about my experience with the NEO8. For the last three years I've been designing and testing the NEO8 in an open baffle. I compared it to other midrange, full range and tweeters and found them stunningly transparent. I found that the rising frequency response helped with the design and that the peak at 12K was diminished enough, especially off axis, to not be a problem, in fact it helped with the loss off axis due to the width of the driver.
I found that by tweaking the baffle width, and using a trapezoidal shape, I can position the baffle peak so it lifts the response on axis between about 500Hz and 2K. This has allowed me to crossover at 650Hz (since I don't play my stereo really loud) without audible distortion. Now comes an extra payoff, as I measure further off axis and the highs diminish, the baffle peak between 500 and 2K goes down too and the response stays quite flat for such a wide driver, with such a bumpy on axis response (the bumps also disappear. Also, the bass and midrange response also diminishes somewhat off axis resulting in a nice wide and smooth listening window. I actually like the highs of the NEO8 in this configuration better than the Hiquphon OW1 in my MBOW1 speakers, mostly because there is no crossover in the tweeter range. Of course, it isn't as smooth or as extended as the OW1, but the compromise is worth it.
James
I found that by tweaking the baffle width, and using a trapezoidal shape, I can position the baffle peak so it lifts the response on axis between about 500Hz and 2K. This has allowed me to crossover at 650Hz (since I don't play my stereo really loud) without audible distortion. Now comes an extra payoff, as I measure further off axis and the highs diminish, the baffle peak between 500 and 2K goes down too and the response stays quite flat for such a wide driver, with such a bumpy on axis response (the bumps also disappear. Also, the bass and midrange response also diminishes somewhat off axis resulting in a nice wide and smooth listening window. I actually like the highs of the NEO8 in this configuration better than the Hiquphon OW1 in my MBOW1 speakers, mostly because there is no crossover in the tweeter range. Of course, it isn't as smooth or as extended as the OW1, but the compromise is worth it.
James
Hi guys
Interesting new "discovery" about the Neo 10. I suspected that the 2nd order HD peak at 5khz, and the 3rd at 2.8khz (As seen on Zaph's website) are "sub-harmonic" artifacts of the 9khz frequency response peak. I spoke to Igor about this to get a verification and he agreed. In other words, if you filter out the 9 khz fr peak by a given amount , the sub-harmonic distortion is also lowered by the same amount at 5 and 2.8khz. Ultimately this is probably trivial info, but it could be useful and informative..
Seth
Interesting new "discovery" about the Neo 10. I suspected that the 2nd order HD peak at 5khz, and the 3rd at 2.8khz (As seen on Zaph's website) are "sub-harmonic" artifacts of the 9khz frequency response peak. I spoke to Igor about this to get a verification and he agreed. In other words, if you filter out the 9 khz fr peak by a given amount , the sub-harmonic distortion is also lowered by the same amount at 5 and 2.8khz. Ultimately this is probably trivial info, but it could be useful and informative..
Seth
Hey Greg
That's the first "subjective" comment I've seen from Danny concerning the Neo 10. Interesting..
There is a subharmonic at around 1 khz that is also an artifact of the 9khz peak.
These subharmonics totally disappear when low passing the neo 10 in a typical crossover situation.
Seth
I love to experiment. So in the interest of the Neo 10, I've gone back to using it with the help of dsp as a mid tweet in a sealed cabinet (The same thing can be done for open baffle). Right now, I have a ruler flat(And I mean ruler flat!) response from 300 hz to 18khz on axis at 1m. There are good and bad ways of getting a flat response using DSP, and I feel that I've found the best way. A way that puts zero stress on the driver. The sound is totally astonishing to say the least! Hypnotizing..Very Quad ESL like. Before anyone questions the practicality of what I am doing, if you compare the Neo 10 to a typical electrostatic mid tweet element, the radiating area is extremely small and acts pretty much like a point source. Very much like the central element of a Quad ESL, which is what I'm trying to emulate.
The neo 3's are going to be implemented in an open back, planar magnetic headphone experiment. It may work, or it may not work. My preliminary testing shows that it measures?? ruler flat from 700hz to almost 20hz. What happens above that is a total mystery. You need a $20,0000 dummy head to get accurate readings in that range, so my above 700hz judgments will be purely subjective. I will probably send them to tyll Hertsens for measurements when I am done. The thing that will probably kill this experiment is the unit to unit upper frequency variability of the Neo 3. Let's wait and see...
Seth
The neo 3's are going to be implemented in an open back, planar magnetic headphone experiment. It may work, or it may not work. My preliminary testing shows that it measures?? ruler flat from 700hz to almost 20hz. What happens above that is a total mystery. You need a $20,0000 dummy head to get accurate readings in that range, so my above 700hz judgments will be purely subjective. I will probably send them to tyll Hertsens for measurements when I am done. The thing that will probably kill this experiment is the unit to unit upper frequency variability of the Neo 3. Let's wait and see...
Seth
As an aside, does anyone feel that the Neo10 is lacking somewhat at very low levels? This was my initial, possibly unfair impression after some very rough and quick experimentation. Could have been the system it was tested in as well.
I certainly found it a very impressive driver otherwise.
I certainly found it a very impressive driver otherwise.
First of all, I have to reiterate that DSP is the only practical way of achieving this goal.
As for the box, I'm using the smallest Dayton cabinet. It's a perfect fit for the Neo 10. Since I'm also reproducing the high frequencies with this driver, It's extremely important to kill the back wave as much as possible. Completely filling the interior with stacked bonded Logic panels is by far the best sounding way of doing this. I tried acoustic foam, but it sounds terrible in comparison.
Now for the DSP. Your going to need a DCX for primary signal processing and a DEQ to swab the deck afterwards. First, set your mic at 1m on axis. Then set DCX on parametric and center the frequency at 9.06 khz. Set the level at -9.5. The Next part will be to tailor the Q to your specific situation. My setting is .7 . Fine tune the level and Q to get it as flat as possible before resorting to the DEQ to make the final adjustments. If you want, you can add more parametric settings and try fine tune the response that way instead of using a DEQ, but the interactions between the different parametric points probably makes that route to complex and unpredictable to get really accurate results. You will have to raise the 16k and 20k levels up pretty high after the parametric adjustments to get the extreme highs back, but that is optional. A Quad ESL central panel(Of which I am humbly trying to emulate) has little if any extreme highs, and it's considered one of the, if not the best sounding loudspeaker ever made.
Seth
As for the box, I'm using the smallest Dayton cabinet. It's a perfect fit for the Neo 10. Since I'm also reproducing the high frequencies with this driver, It's extremely important to kill the back wave as much as possible. Completely filling the interior with stacked bonded Logic panels is by far the best sounding way of doing this. I tried acoustic foam, but it sounds terrible in comparison.
Now for the DSP. Your going to need a DCX for primary signal processing and a DEQ to swab the deck afterwards. First, set your mic at 1m on axis. Then set DCX on parametric and center the frequency at 9.06 khz. Set the level at -9.5. The Next part will be to tailor the Q to your specific situation. My setting is .7 . Fine tune the level and Q to get it as flat as possible before resorting to the DEQ to make the final adjustments. If you want, you can add more parametric settings and try fine tune the response that way instead of using a DEQ, but the interactions between the different parametric points probably makes that route to complex and unpredictable to get really accurate results. You will have to raise the 16k and 20k levels up pretty high after the parametric adjustments to get the extreme highs back, but that is optional. A Quad ESL central panel(Of which I am humbly trying to emulate) has little if any extreme highs, and it's considered one of the, if not the best sounding loudspeaker ever made.
Seth
Last edited:
Proraum Vertriebs-GmbH | Startseite
If you have not already seen this, you can go to "Pro20D" in the download section on this site to get more raw data on the Neo 10.
Seth
If you have not already seen this, you can go to "Pro20D" in the download section on this site to get more raw data on the Neo 10.
Seth
Last edited:
Thanks Remlab. I downloaded all that stuff a while back and based my experimental baffle on their 'head' unit for the Neo 10 and Neo 3. Very useful since it shows the raw driver responses on their baffle and also the responses with their crossover. I did ask them if they'd supply the crossover alone, but- not unreasonably- it was the whole kit or nothing. Much as I like the Neo 3, I do prefer the RAAL 70-10D overall, although there is a very appealing transparency the Neo 3 has which can make the RAAL sound a bit over emphatic at times. On the other hand the '3 can sound a bit too relaxed... It's not really an apples to apples comparison though, since the small RAAL needs to be crossed over higher than the Neo 3, which compromises the power response of the '10, particularly in the vertical plane. For the '10 I'm probably going to try something along the lines of Planet 10's 'mid TL' as used in the Tysen design.
Last edited:
I spoke to Igor about those measurements. He told me that in the first part of the measurements section regarding the Neo 10, they used some type of sealed box for the measurements and then in the last part, they used the actual baffle from the kit..
Seth
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.