I have also measured the Sounderlink Neo8. Not quite as good as a BG Neo8S:
I had hoped it would perfrom similar to the Neo8S so I could have a column of them and play them full range, but with that performace they pretty much have to be crossed at ~ 1000 hz rather than at 400 hz where the Neo8S can go.
I have polars of it too but there it measures same as the BG Neo8S.
I had hoped it would perfrom similar to the Neo8S so I could have a column of them and play them full range, but with that performace they pretty much have to be crossed at ~ 1000 hz rather than at 400 hz where the Neo8S can go.
I have polars of it too but there it measures same as the BG Neo8S.
Attachments
Last edited:
I have also ripped out the stuffing from one of my Sounderlink Neo3s and measured it:
Sadly they are not comparable to the Neo3W which has 5 rows of magnets, the Sounderlink has 3 sets of magnets which I believe is the same amount as the Neo3PDR.
The edge wool damping is where I have inserted some wool yarn on the outer holes like the ripped out stuffing but not as compact.
Sadly they are not comparable to the Neo3W which has 5 rows of magnets, the Sounderlink has 3 sets of magnets which I believe is the same amount as the Neo3PDR.
The edge wool damping is where I have inserted some wool yarn on the outer holes like the ripped out stuffing but not as compact.
Attachments
The sounderlink Neo3 measures the same as the Neo3PDR, and seems to be constructed in the same way. Both have 3 rows of magnets in the middle and felt on the outer holes.
The Neo3W is a bit different. It has no felt and is made of 5 rows of magnets. The GR neo is also constructed in the same way (although in my tests it distorts a lot more than the BG version).
These differences make a difference in the polars, the Sounderlink Neo3PDR and the BG Neo3PDR have "too good" off axis response which causes a mini null on axis at 6-8 khz.
I don't have saved measurements of the PDR naked 0-90 but it performs same as the Sounderlink. This is why Juhazi in his Aino gradient speaker runs the Neo3PDR flipped 90 degrees. I tested that but didn't find the result satisfactory as horizontal off-axis response suffers a lot.
I included a vertical measurement, note it is 0-60. On the Neo3W in standard config it measures same on 0 and 15 degrees, but the vertical PDR already differs a lot between 0 and 15 degrees above 6 khz. And 30 degrees is already -12 dB @ 10 khz.
Attachments
Nice and uniform off-axis results.
(..well, except for 11 kHz - but not exactly a problem at that frequency.)While not a Neo3-derived driver it is a planar:
Two Dayton PTMini-6 mounted back-to-back and wired as a dipole. Measures pretty good, as long as you can cross it high as distortion @ 4 khz and below is a bit high.
I tried here with a 6 khz crossover as I'm building a CBT array but the result was not great. I get massive lobing at the crossover frequency due to the large ctc distance to the PTMini and the wideband driver below, and ofc the response is not consistent on left vs right sides either.
EDIT:
The combination is using a ~ 6 khz LR 24 dB / oct filter. I guess a steeper filter could solve that issue although I haven't tried. Most of the threads discussing steep filters talk about the problems they have with ringing so I've kept using more gentle slopes.
Two Dayton PTMini-6 mounted back-to-back and wired as a dipole. Measures pretty good, as long as you can cross it high as distortion @ 4 khz and below is a bit high.
I tried here with a 6 khz crossover as I'm building a CBT array but the result was not great. I get massive lobing at the crossover frequency due to the large ctc distance to the PTMini and the wideband driver below, and ofc the response is not consistent on left vs right sides either.
EDIT:
The combination is using a ~ 6 khz LR 24 dB / oct filter. I guess a steeper filter could solve that issue although I haven't tried. Most of the threads discussing steep filters talk about the problems they have with ringing so I've kept using more gentle slopes.
Attachments
Last edited:
You're not the first to try it but probably first to report it here. Another member I know is working with these, in a line along side a full ceiling-to-floor wideband array. He loves the sound, says it's great >8kHz in similar back-to-back dipole arrangement.
Hi Ollboll,
thank you for the nice mesurements of the radian lt2 (Is the spl db exact like you have mesured it? ) . Hoy do it perform in relationship to the bg neo 3 or the sounderlink?
Do you have mesurements of the radiation pattern of the radian lt2?
Thank you for your help! I want to build a dipol with one of the 3 driver in this topic as tweeter.
Greetings, Tomas
thank you for the nice mesurements of the radian lt2 (Is the spl db exact like you have mesured it? ) . Hoy do it perform in relationship to the bg neo 3 or the sounderlink?
Do you have mesurements of the radiation pattern of the radian lt2?
Thank you for your help! I want to build a dipol with one of the 3 driver in this topic as tweeter.
Greetings, Tomas
As far as I can remember I did not calibrate the SPL. The distortion measurements in the first page comparing the BG vs Sounderlink vs GR Neo3 are all calibrated and comparable to each other.
And on radiation pattern the measurements I have are in this post.
I should have time next weekend to do some calibrated SPL and distortion measurements and compare it to say a BG Neo3W if you want.
My guess is that it will perform kind of like the Neo3W but shifted up an octave or so. So it would prefer a crossover in the region of 3-4 khz @ LR4.
any news here?
on my open baffles i use a Mundorf AMT 17D2.2 which i like
then a Scanspeak 10F takes over from 3khz
a Seas Excel W22 takes over from around 800hz
and below i have a B&C 18SW115
all filtered and corrected with FIR on Minidsp DA8 OpenDRC and 8 channels of Hypex UCD amps; only the big woofer gets a PEQ boost on 20hz.
i keep on wondering whether a AMT instead of the Scanspeak 10F would be an improvement. But quality issues make me hesitate.
on my open baffles i use a Mundorf AMT 17D2.2 which i like
then a Scanspeak 10F takes over from 3khz
a Seas Excel W22 takes over from around 800hz
and below i have a B&C 18SW115
all filtered and corrected with FIR on Minidsp DA8 OpenDRC and 8 channels of Hypex UCD amps; only the big woofer gets a PEQ boost on 20hz.
i keep on wondering whether a AMT instead of the Scanspeak 10F would be an improvement. But quality issues make me hesitate.
Last edited:
Here is a comparison of impulse response of a friend egg and a Foster E110 ribbon-like tweeter (good above 3500 Hz) - may have been the top end of the famous Infinity Servo-Static.
Mic was about a foot away. No XO. The quality of the pulse and the unwanted activity up to say, 1 milli-second reflects on quality, I think. On that basis, the E110 (right image) is better.
Ribbon THD distortion can't be beat as well. My testing was at a low loudness with ambient noise, but even so, the ribbon came in OK around .3% and the dome about .9%.
Seems to me that's a good way to look at 'em. I can't attest to the history of these old drivers, but I am just illustrating that impulse is revealing.
Mic was about a foot away. No XO. The quality of the pulse and the unwanted activity up to say, 1 milli-second reflects on quality, I think. On that basis, the E110 (right image) is better.
Ribbon THD distortion can't be beat as well. My testing was at a low loudness with ambient noise, but even so, the ribbon came in OK around .3% and the dome about .9%.
Seems to me that's a good way to look at 'em. I can't attest to the history of these old drivers, but I am just illustrating that impulse is revealing.
Attachments
Last edited: