No Capacitor
The in series capacitor for a ribbon is due to the very low resistance of a ribbon (as opposed to inductive impedance). From my readings on other BG products (the RD series), BG states that the load is comparable to 4/6 ohm resistor (ie not frequency related inductance) so no capacitor would be required.
The in series capacitor for a ribbon is due to the very low resistance of a ribbon (as opposed to inductive impedance). From my readings on other BG products (the RD series), BG states that the load is comparable to 4/6 ohm resistor (ie not frequency related inductance) so no capacitor would be required.
Hi,gp4Jesus said:
My experience so far is rather mixed. I've been using it in a dipole housing, which takes too much out of the range roundabout 1KHz and lower. The latest filter is a 3rd order filter (butterworth-like) at approximately 700Hz which results in an acoustical slope of a near textbook LR 4th order at 800Hz. A 3rd order acoustic slope at 800Hz would also work fine, but 4th happened to turn out better and less complicated in this setup. Maybe that will change when I go to a closed housing in the final design because of the low end boost compared to dipole.
I'm using the non-PDR version, which needs a rather heavy notchfilter at about 12KHz for the on-axis peak (which disappears rather fast when going off-axis). Since I'm using a digital crossover, terribly convenient during prototyping, I've done some more curve tweaking at 3-4KHz (slight dip) and above 18KHz to flatten the response. The Neo is currently behind a shallow (about 2") waveguide, which helps the dip in the response at and below 1KHz a bit.
The way it is set up now (as mid and tweeter in a dipole housing) is not working to my liking. I've heard ppl being extremely enthousiastic about the Neo8 used all the way up to 20KHz, but to my ears, that doesn't work very well... For one, you need to put your head in a vise in order to hear 'optimal performance' because the treble range is so damn directional concerning the big peak in the upper treble. Once you have your head in the mandatory vise, no matter how much you tweak the response, you're always left with a very large deviation between the on- and off axis response in the treble range which makes the speaker very picky about room acoustics and placement (this was the primary reason for me to try a few waveguides in order to mediate that problem). It's rather hard to find an 'optimum'. Getting is wrong is easy, resulting in either a dull overal treble or some sharp edges that find their way to your ear via the room boundaries.
After trying a fee weeks to get the Neo8 to behave, I asked Danny from GResearch, where I had purchased the units, what his experience with this driver was and he confirmed my findings, saying that's why he doesn't use it as a tweeter. In the meanwhile, I've also tried the Neo8 in a closed housing, which improves the low treble as opposed to an open baffle, but obiously does nothing for the uppper treble. I'll probably get a pair of Neo3 soon to put on top of the Neo8 as a proper tweeter. As far as its midrange goes, it's great, performs way, way beyond its price.
Hope that helps.
Hans.
Hans
keep me posted on what worked for you. If I could "acousticallly fill in" below about 1K that would be good. I'd like to XO them @ 600-700. I have a TDM; active, analogue, 4th order with infinite adjustability. The 12K peak is another matter; I can't hear above 16K, so that problem doesn't concern me.
thanks and good luck.
tony
keep me posted on what worked for you. If I could "acousticallly fill in" below about 1K that would be good. I'd like to XO them @ 600-700. I have a TDM; active, analogue, 4th order with infinite adjustability. The 12K peak is another matter; I can't hear above 16K, so that problem doesn't concern me.
thanks and good luck.
tony
Hans L said:
Hans.. you might try rotating the driver "outward" ("toe-out" as opposed to "toe-in") by 45 degrees from your listening axis and then adding in a tweeter on-axis to fill-in the remainder final ocatave and a half. This should give enhanced imaging out of the speakers "boundries". It will however enhance the "power response" at higer freq.s and like require a more extended lower freq. response to "balance" the sound.
ScottG said:
Hi Scott,
thanks for the suggestion. Funny you should mention something like that: in testing the Neo8 I've already tried something very similar. I 'ductaped' a nice 3/4" tweeter close to the Neo for the upper treble and turned the enclosure straight forward with the tweeter pointing towards the listener, simply to avoid (that is, let the room 'smear it') the changes in on-off axis behaviour as much as possible, despite the fact that that area falls past the passband of the Neo when filtered.
In general I find that most speaker are placed pointed in toe, so that they cross just slightly in front of the listener. That usually enhanced pinpoint imaging, but can also confine the stereo image between the speakers. I used to enjoy this pinpoint, razorsharp 'accuracy' a lot, but over the years I (and the speakers
) have turned towards a more relaxed and open imaging, by pointing the speakers straight forward. That is the way I design my speakers nowadays. Much depends on the actual recording, but it seems that 90% of them consist of potted mono sources mixed to form a stereo image. Especially in those cases, a less sharp pinpoint imaging adds to a more musical representation of the event imo.As for a suitable tweeter, it seems obvious to try out the Neo3 as a companion to the neo8. Not a novel choice, I've seen it done before (in a commercial design, not yet in diy afaik), but I expect it to work very well.
Hans.
Hans L said:
Hi Scott,
thanks for the suggestion. Funny you should mention something like that: in testing the Neo8 I've already tried something very similar. I 'ductaped' a nice 3/4" tweeter close to the Neo for the upper treble and turned the enclosure straight forward with the tweeter pointing towards the listener, simply to avoid (that is, let the room 'smear it') the changes in on-off axis behaviour as much as possible, despite the fact that that area falls past the passband of the Neo when filtered.
In general I find that most speaker are placed pointed in toe, so that they cross just slightly in front of the listener. That usually enhanced pinpoint imaging, but can also confine the stereo image between the speakers. I used to enjoy this pinpoint, razorsharp 'accuracy' a lot, but over the years I (and the speakers
As for a suitable tweeter, it seems obvious to try out the Neo3 as a companion to the neo8. Not a novel choice, I've seen it done before (in a commercial design, not yet in diy afaik), but I expect it to work very well.
Hans.
How would you like to "have your cake and eat it too"?
Consider the listener axis as 0 degree..
Neo8 at 60 degrees off of the 0 degree axis, specifically "toe-out".
Aurum Cantus G2si at about 45-50 degrees off of the 0 degree axis, specifically "toe-in".
1st order filter about 6-7 kHz with some additional resistance to drop the sp-level down to the the Neo8's average level. G2si inverted (180 degrees) on baffle so that the filter's vertical polar "tilt" of 15 degrees is firing "upward". Neo8 at ear level. G2si below Neo8 line.
This will give a horizontal polar plot that looks rather like a "U" with the base of the "U" as the listener axis. IF you move off-center imaging will "shift" considerably less then if it was all directed with "toe-out". Additionally, you'll still retain the more expansive sound that "toe-out" can provide (..recording dependent). You have to take care with freq. balance though (there will be more energy in-room now and it will require either some shelving or an extended low freq. response).
Always interestedScottG said:
Sounds like an interesting and novel idea. It does trigger a few doubts though. Hope you don't mind
The G2Si has a pretty wacky on-axis freq plot as well up high. I haven't seen off-axis plots of it yet, though. The Neo3 PDR seems much better behaved, in fact, it's unusually smooth on axis compared to planar and ribbon types of tweeters, with a very nice, wide dispersion.A first order acoustical slope could be a challenge for the G2si, at least over a wide enough range in order to count on the 15% tilt. And as for the Neo8 LP, it wouldn't do much to avoid the harsh treble, but that's of course why you suggest to aim it far off axis. In any case, I'm quite intruiged by the idea of this kind of horizontal distribution of treble, eventhough it doesn't sound 'right' intuitively: you're stilling crossing over at very high freq, which can be problematic without regards to the toe-in/toe-out and of course the left and right angle per speaker will become much more freq dependent. I can theorize and visualize the horizontal and vertical graphs all day, but in all honesty, I'd have to hear it first so see wether this would work as intended.
Have you tried this before?
Hans.
Hans L said:Always interested
Sounds like an interesting and novel idea. It does trigger a few doubts though. Hope you don't mind
A first order acoustical slope could be a challenge for the G2si, at least over a wide enough range in order to count on the 15% tilt. And as for the Neo8 LP, it wouldn't do much to avoid the harsh treble, but that's of course why you suggest to aim it far off axis. In any case, I'm quite intruiged by the idea of this kind of horizontal distribution of treble, eventhough it doesn't sound 'right' intuitively: you're stilling crossing over at very high freq, which can be problematic without regards to the toe-in/toe-out and of course the left and right angle per speaker will become much more freq dependent. I can theorize and visualize the horizontal and vertical graphs all day, but in all honesty, I'd have to hear it first so see wether this would work as intended.
Have you tried this before?
Hans.
I haven't tried it - but I have been thinking about it for quite some time now. (..right now I'm testing out lower midranges for a radial configuration - after that I'll likely purchase some neo8's for the upper mid/lower treble.)
The vertical polar tilt is endemic to odd-order filters and spans the driver's entire bandwidth - you'll get 15 degrees with them even if you don't want it. The only question with them is: " do you want the tilt up or down?" Most don't even realize that you can tilt them up - and so you'll mostly see sloped baffle designs for them like Vandersteen and Theil (..though baffle diffraction and midrange position vs. response is also a serious deciding factor here). Another nice thing with the physically inverted 1st order is that time/phase alignment from tweeter offset is mostly taken care of because the tweeter is further away due to it being below the midrange driver. Also, its better for "stand-up/sit-down" time/phase integration (..thus allowing a listener considerably more freedom to move around while listening and still maintaining a good response).
While a 1st order on the G2si is a bit of strain - I really don't think its enough to cause it serious difficulties unless you like listening at spl's greater than 105db. Others have commented favorably on the driver crossed 1st order around 7 kHz - and I haven't seen anything about mechanical failure in such situations. (..and of course mechanical failure doesn't mean driver replacment in such a design - only a $15 ribbon replacement). Still, it is speculation on my part.
The G2si does have an irregular freq. response - but it is higher in freq. than any fundamental and effectivly contributes to "air", "space", "harmonic "sheen", and freq. balance. It also happens to be a "dip" similarly situated to the neo8's "peak" when considering in-room power response.
Now lets look at the off-axis responses of both drivers:
Neo8:
http://www.diyspeakers.net/Projects/BG/justus.php
G2si:
http://www.markk.claub.net/Testing/G2si/arum_cantus_g2si_test_results.htm
Look closely and think about what I've said.. do you see "it"?
Now part of the reason I spec.ed the G2si is that it is a true ribbon - and if given the chance I'll take the sound of a true ribbon over a planar (at higher freq.s) any day of the week. This isn't to say that the neo3 wouldn't also be a good choice - but IMO not nearly as good subjectivly. Additionally, it doesn't have a rising response on-axis that the G2si has to enable the "U" pattern.
..anyway - still just speculation on my part.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.