I have never done EXACTLY that measurement, but when I was first playing around with the tweeter (a single unit) I took this measurement with the tweeter mounted using only one screw holding it to the top of a baffle, so close to being in "free air". See attached. I didn't record the measurement angles but it mgiht be something like 0-11-22-33-44-55-66 degrees, or perhaps 0-22-44-66-90. Regardless, you get the idea.
What you can see is the waveguide not all that effective around 2kHz, but is increasingly so at higher frequencies. This means the driver alone is tending towards omni at the low end of its passband (where the traces bunch together).
When you compare this to the back-to-back (dipole) implementation SHOWN HERE, the on axis response shape remains about the same as the single driver, but around 1kHz the response is no longer quasi-omni but has taken on a dipole character. No sign of peaks or dips like you find for a two point-source model with the same separation, which I found very surprising. So my guess is that the front and rear tweeters interact to form a null at 90 deg around 2kHz, but above that the increasing directivity from the waveguide causes them to see each other less and less, and you get two independent sources, one frontward firing and one rearward firing. I can only guess that the waveguide makes this all happen by having the right directivity at just the right frequencies.
These tweeters are cheap (Dayton ND25FW-4 from PE at $17 each), so you can pick up a pair and try it out yourself. A pair can be used to make the least expensive dipole tweeter that I am aware of, and the distortion performance is really not too bad.
Attachments
Last edited:
. The Note used the OX20, in a small wave guide, from 6k up. The NaO II, which is what the first 2 figures are from, uses Seas Excel T25 front which is flush mounted on the 11" wide baffle and a 27TDFC rear tweeter mounted on a 5" diameter stand off, 2 1/4" tall.
Last edited:
Yes, I was correcting my original post, thought I had the ND25 at first. Sorry for the confusion.
Dave
Last edited:
I have not read all 45 pages but does this help?
Bohlender-Graebener RD75 - The Ultimate 2-Way System | Perry Marshall
Bohlender-Graebener RD75 - The Ultimate 2-Way System | Perry Marshall
I was curious. Comments at the old page at Parts-Express says it's open back, but the recommendation is to enclose the back, otherwise it will be a dipole.
Dave
Dave,
Your impressions about the BG RD75 are correct but I don't remember any application with an open back RD75 for a full range dipole speaker like Charlie has in mind. Perhaps other readers can update and straighten my forgetful thinking.
At any rate my point was that Perry Marshall's design was not intended to be a dipole as I stated in post #448.
Jim
Your impressions about the BG RD75 are correct but I don't remember any application with an open back RD75 for a full range dipole speaker like Charlie has in mind. Perhaps other readers can update and straighten my forgetful thinking.
At any rate my point was that Perry Marshall's design was not intended to be a dipole as I stated in post #448.
Jim
Last edited:
Unfortunately the RD75 is indeed not available anymore, as now I would have the money to buy a pair...
Still, Stigerik made a two way dipole, with the RD75 from 100hz upwards, and bass provided by multiple 21" woofers. There is a thread, but pictures are (mostly) gone.
Hi Charlie,
I was rereading the thread as I am looking to build another dipole speaker, and now I wondered: why did you use a 15" as midrange from 250 to 1600-1800Hz? I understand that it is good to get to the 250Hz, but doesn't it bundle too much at ~1700Hz? I know you measure and evaluate everything very carefully, so this is surely not a criticism, just curiosity. Or, with other words, wouldn't a 12" or 10" suffice to get to the same 250Hz, but behave much better at the top? (I see that the deltalite 15" has a very open construction, so it must have a nice dipole pattern)
Sorry for not responding here sooner. I missed that there were new posts. My notifications never work properly... You can always PM me with questions or to direct my attention to a post, etc.
So, for months now I have been tied up with a large landscaping project (ongoing since last fall) and not building anything. Over the winter it's too cold for speaker building for the most part where I live. I have been planning a couple of builds, however.
First, I want to add a 6" class nude midrange driver to my latest nude dipole system - the one that uses the 15" woofer and dipole back-to-back tweeters.
I then plan to build a slim floorstander plus sub all-dipole system using 8" woofers in a slim baffle, either another 6" or 8" nude driver, and a couple of those new GRS drivers that seem to be close copies of the B&G "NeoX" lineup. For that system I plan to use an onboard TDA3251 4-channel amp (from 3e audio) sourced from a computer+USB audio interface. Audio can then be streamed to the speakers using my Gstreamer based application.
As for the subwoofer driver, it's the Fi Audio IB318v2. It's perfect for my application in a deep H-frame. The driver is a total bargain and has 34mm Xmax, Fs 24Hz, Qts=0.79. What's not to like? I think this is the stacked magnet driver you mentioned?
The 15" driver I like to use is the Eminence Deltalite II 2515. When used nude it has an amazing dipole response, and yes, I did use it up to 1.6kHz. Response into cone breakup is relatively smooth and below that it's CD-like all the way thru the very broad dipole peak, and downwards. It's very easy to work with. Using it in a 2-way is really pushing the envelope, and this is why I want to add the 6" midrange so that I can cross over around 700Hz between these drivers and move the crossover point to the tweeter up a bit. The 15+tweeter combo was more of a "what if?" experiment designed to see if it could work, and it did for the most part. But the system will be much better with the addition of the 6" in the middle. Because the 15 is very efficient and I only use it down to 150Hz or so I can get very high SPL levels without having to use insane amounts of power. When I demoed the system last, it was using a 60W into 8R amp and that was plenty loud in a large room. I plan to use the Deltalite 2515 in future systems, since it has low distortion, a very open back/basket, and is not expensive.
In general l like to use relatively large drivers for a particular band, compared to what one would use for a boxed speaker. IMO, the dipole pattern at low frequencies is helping to even out the directivity compared to when the driver is used in a box. So you can get away with using a larger driver, as long as the driver's response allows it (e.g. vis a vis breakup, distortion, etc.). It's difficult to figure it out from the datasheet, so you have to buy and test/measure a driver to know. That is especially the case to the rear, where the basket/magnet can cause some issues, especially when the magnet/motor is large and basket windows small. The Deltalite 2515 is the total opposite of that - the motor is relatively small compared to the cone and the basket is very open. I think that is the reason why it makes such a great dipole driver.
Hi Charlie,
Many thanks for your extensive reply! In our apartment there is little landscaping, even though I planted some stuff on our balcony during the last spring/summer!
Interesting that the Deltalite can go up to the 1600Hz. I am thinking it could be a nice combination with a dipole ESS AMT. No personal experience, but apparently they can go even lower than 1.6kHz.
It was indeed the Fi Audio sub I saw in the other thread. It does indeed look very nice for the job, but shipping to Europe would be prohibitively expensive, I think
I had a system with double 12” (SB34NRX75-16), a NEO 10 planar and AMT dipole tweeter, but with three little ones running around barely the time to listen to music and constant fear of these speakers being damaged. So I hanged a woofer + back to back Fullrange drivers (Scanspeak 10F) from the ceiling, which gives sort of enough SPL for a well behaved family in Switzerland I am now working on a center table, which I want to use to hide all electronics + baffleless speakers when these are not being used. Will take me some 5 minutes to take speakers out and position them, which is ok as amplifiers need to warm up anyway for better sound. I have a bigger system in mind where the center table will, besides hiding electronics and speakers, be used an H-frame for a pair of 15”subs. I paused this project as I do not have the subs yet (thinking about the Ultimax 15 from Dayton), and want to test the idea with drivers I have on hand.
Finally, I always enjoy your posts: clear posts, backed up with measurements, using drivers one can pay without going bankrupt. Thank you!
Many thanks for your extensive reply! In our apartment there is little landscaping, even though I planted some stuff on our balcony during the last spring/summer!
Interesting that the Deltalite can go up to the 1600Hz. I am thinking it could be a nice combination with a dipole ESS AMT. No personal experience, but apparently they can go even lower than 1.6kHz.
It was indeed the Fi Audio sub I saw in the other thread. It does indeed look very nice for the job, but shipping to Europe would be prohibitively expensive, I think
I had a system with double 12” (SB34NRX75-16), a NEO 10 planar and AMT dipole tweeter, but with three little ones running around barely the time to listen to music and constant fear of these speakers being damaged. So I hanged a woofer + back to back Fullrange drivers (Scanspeak 10F) from the ceiling, which gives sort of enough SPL for a well behaved family in Switzerland
I am now working on a center table, which I want to use to hide all electronics + baffleless speakers when these are not being used. Will take me some 5 minutes to take speakers out and position them, which is ok as amplifiers need to warm up anyway for better sound. I have a bigger system in mind where the center table will, besides hiding electronics and speakers, be used an H-frame for a pair of 15”subs. I paused this project as I do not have the subs yet (thinking about the Ultimax 15 from Dayton), and want to test the idea with drivers I have on hand. Finally, I always enjoy your posts: clear posts, backed up with measurements, using drivers one can pay without going bankrupt. Thank you!
To come back to this; why not use the same drivers front and rear in a single enclosure?
Linkwitz has this rather personal journal article about his home system and the denouement is when he adds a rear-facing tweeter. Do I remember correctly? Link anybody?
Truly a classic good-read.
For sure, dipole sound is wonderful because of the back sound. Duh. But the driver(s) aren't living by the same rules of dispersion, flat FR, etc. but creating less-focused ambient sound. So traditionally, whatever comes out the back of a Rice-Kellogg driver is OK. Of course, nothing coming out the back of most tweeters today since they aren't cones often.
B.
Truly a classic good-read.
For sure, dipole sound is wonderful because of the back sound. Duh. But the driver(s) aren't living by the same rules of dispersion, flat FR, etc. but creating less-focused ambient sound. So traditionally, whatever comes out the back of a Rice-Kellogg driver is OK. Of course, nothing coming out the back of most tweeters today since they aren't cones often.
B.
Last edited: