Thanks for the text. One question, in your opinion, where is the transition from small to large room, or, when is dipole sub superior to a closed box? I thought I read once that the rooms lenght should be at least 1/4 of the wavelenght of the dipole axis ?
Thank you!
erik
Their only dipole offering, the AMTPRO-4, every free air measurements I’ve seen are meh with a rising response from 1k to 10k, and almost useless performances past 10k, both on and off axis, the vertical dispersion is even worst.
Nice measurements Lewinski, you're going to need a real sensitive 8" to fully take advantage of that.
I would find an 8" that can match it's spl and try a TMM personally. Parallel the mids and place an inductor between the upper and lower midrange. Bring in the bottom mid right where the roll off of the upper mid is 3db down and flatten it a bit with the increased sensitivity of the paralleled pair.
Use dsp to do the crossover between the mids and Beyma, and the low shelf dipole compensation and high pass crossover on the low end of the mids. An active/passive hybrid that kills a couple birds with one stone. Poor birds. You should be able to run that panel all the way down to around 100hz or so. The Fs of a high sensitivity mid may be the limiting factor there.
iamvalheru If done on all four sides like you say, it would widen (lower Q) the resonance peak and drop it's level a bit, the center frequency still being determined by the depth of the cavity. Not sure the complication of the build would be worth it. Keeping the resonance at one higher Q peak, compensating it and only using the H frame below the resonant frequency is what many did.
Angling the driver helps in a similar way, and is one of the reasons so many went with two 10" drivers in a V formation. It helps mitigate the resonance with non parallel surfaces and the varying depth of the cavity, there is some force cancellation, you get the increased sensitivity of a paralleled pair and a long stroke 10" in a V frame with a front to back wave separation of 10" is about as small as you can go and still get down into the ~30hz area. The V moves the resonance up in frequency around by around 100hz and makes it more benign, so you can use it up higher.
In the design I posted, using the single 15" in a baffle with only 2 shallow angled sides and a bottom saved a lot of money and let me use it even higher. There is no resonance anywhere near the frequency range I'm using it in, and although the baffle is shallow, the front to back separation is just as great as the V or H frame because of the width of the driver itself.
I would find an 8" that can match it's spl and try a TMM personally. Parallel the mids and place an inductor between the upper and lower midrange. Bring in the bottom mid right where the roll off of the upper mid is 3db down and flatten it a bit with the increased sensitivity of the paralleled pair.
Use dsp to do the crossover between the mids and Beyma, and the low shelf dipole compensation and high pass crossover on the low end of the mids. An active/passive hybrid that kills a couple birds with one stone. Poor birds. You should be able to run that panel all the way down to around 100hz or so. The Fs of a high sensitivity mid may be the limiting factor there.
iamvalheru If done on all four sides like you say, it would widen (lower Q) the resonance peak and drop it's level a bit, the center frequency still being determined by the depth of the cavity. Not sure the complication of the build would be worth it. Keeping the resonance at one higher Q peak, compensating it and only using the H frame below the resonant frequency is what many did.
Angling the driver helps in a similar way, and is one of the reasons so many went with two 10" drivers in a V formation. It helps mitigate the resonance with non parallel surfaces and the varying depth of the cavity, there is some force cancellation, you get the increased sensitivity of a paralleled pair and a long stroke 10" in a V frame with a front to back wave separation of 10" is about as small as you can go and still get down into the ~30hz area. The V moves the resonance up in frequency around by around 100hz and makes it more benign, so you can use it up higher.
In the design I posted, using the single 15" in a baffle with only 2 shallow angled sides and a bottom saved a lot of money and let me use it even higher. There is no resonance anywhere near the frequency range I'm using it in, and although the baffle is shallow, the front to back separation is just as great as the V or H frame because of the width of the driver itself.
Guess I'll make a few "horn" H-frame when my GW-1858 get here, which should be in about a month now since I missed the cut-off date for the order that's theoretically leaving today.
I'll try both straights and curved walls affixed to the center baffle via hinges and by filling the gaps where the panels would normally meet, this should be fun!
I'll try both straights and curved walls affixed to the center baffle via hinges and by filling the gaps where the panels would normally meet, this should be fun!
For the upper midrange, as anybody compared the SB Acoustic SB17CAC (Ceramic coating) and the SB17CRC (Carbon fiber sandwich) for use in open baffle or nude application, mainly their response (front and rear) and the sound signature?
The carbon fiber SB17's are only 10 $CDN dollars more, and as per the factory frequency response curves, it looks like they could be crossed over higher than the ceramic coated SB17's, which could mean greater flexibility in matching various dipole ribbon tweeter and/or skip the lower frequency distortion they tend to suffer from.
The carbon fiber SB17's are only 10 $CDN dollars more, and as per the factory frequency response curves, it looks like they could be crossed over higher than the ceramic coated SB17's, which could mean greater flexibility in matching various dipole ribbon tweeter and/or skip the lower frequency distortion they tend to suffer from.
IMHO the CAC (white "ceramic" cone) is superior, because the CRC version has a little hiccup in the response around 1kHz, which you can just see in the impedance as a small bump. This is some kind of resonance or something. I bought a pair of the CAC and measured them nude with pretty high frequency resolution and the responses on and off axis looked very clean, with no sign of problems that I have seen in other SB17 drivers I have used in the past, e.g. the MFC (poly cone) version.
The CAC is an *excellent* driver, as long as you do not use it below 200Hz, where its distortion performance and eventually its limited Xmax is less desirable than other offerings. As a dipole midrange, up to 2k or 2.5k, it's really stellar.
I have seen some arguments related to the size of the room and the longest wavelength that will fit, or some fraction of that wavelength. I can't recall what these are off the top of my head, exactly.
I can relate some subjective experience about the dipole speakers that I mentioned earlier in post 52. These use an M-frame woofer section loaded with two Eminence Alpha 15A for bass up to 150Hz. In a medium sized demo room on the first day of the event the bass was anemic. The next day in a larger conference room at a hotel, they sounded much better on the low end, and this was the small/medium to large room changes that I have previously experienced with other dipole systems.
Charlie, thank you for the info on the CAC vs CRC.
For the lower midrange, 200 - 600 Hz, I'm currently playing around in Edge/Basta with different baffle shapes and dimensions for use with a 10” or 12" woofer and I have a few questions:
1. Are there any advantages in using a 10” or 12" in a small baffle for the lower midrange versus a nude 15" woofer (e.g. Eminence DeltaLite II 2515)?
2. Would a high efficiency nude 12" woofer (e.g. Eminence DeltaLite II 2512) provide similar performances as a nude 15” woofer to bridge the gap between the 18” H-frame and the SB17?
Also, and since I kind of have a fixation on MTM configurations with a ribbon (damn you GoldenEar Triton’s), what are your thoughts on a system consisting of a GW-1858 in a flat or U-baffle, nude SB17CAC and AST2560 in a MTM configuration?
Is the jump from an 18” woofer to 6” midrange MTM too much? How would the dipole radiation be impacted?
D
For the lower midrange, 200 - 600 Hz, I'm currently playing around in Edge/Basta with different baffle shapes and dimensions for use with a 10” or 12" woofer and I have a few questions:
1. Are there any advantages in using a 10” or 12" in a small baffle for the lower midrange versus a nude 15" woofer (e.g. Eminence DeltaLite II 2515)?
2. Would a high efficiency nude 12" woofer (e.g. Eminence DeltaLite II 2512) provide similar performances as a nude 15” woofer to bridge the gap between the 18” H-frame and the SB17?
Also, and since I kind of have a fixation on MTM configurations with a ribbon (damn you GoldenEar Triton’s), what are your thoughts on a system consisting of a GW-1858 in a flat or U-baffle, nude SB17CAC and AST2560 in a MTM configuration?
Is the jump from an 18” woofer to 6” midrange MTM too much? How would the dipole radiation be impacted?
D
Last edited:
Does this mean that if my room has a natural node situated at 32 Hz, attempting to go lower than that frequency is just asking for trouble and that I should add a 24 dB crossover at 30 Hz?
I own, and measured, a pair of the Deltalite 2510 (10" version) and it actually has some defects that are not present in the 15" version (2515) response. I haven't owned or measured the 12" 2512 so I am not sure about it only to say if you look at the MFG curves the impedance has too many lumps and bumps below 1k to make me comfortable.
If you want to go the MTM route, I think you could use two SB17CAC's above the H-frame with the 18" Goldwood provided that you design the H-frame shallow and wide enough that the 1/4-wave resonance is at 300Hz or above, assuming you will cross over at 250Hz to the SB17s. This won't give you much bass augmentation, but you can't have both at the same time with an H-frame. You could explore the option of a stuffed U-frame. It might work better under the MTM, actually, and because the driver is on the front panel there is no resonance there. Also for comparable bass loading, a U-frame only needs to be half as deep as an H-frame. For more info on that see:
DIY-dipole-1
especially the last figure at the bottom of the page. Also this page has an example of a damped U-frame:
NaO U-frame
As far as the dipole response of the M's in the MTM, I don't think it is all that different from an MTM in a boxed loudspeaker. The M drivers will be "far" apart at the crossover point and this will introduce some lobing/interference on top of their dipole response. If this does not represent an issue of concern for you in a boxed loudspeaker I imagine it will perform fine for you when implemented as a dipole.
Last edited:
It's not the same as room modes. Try to read through this dense material on it (and part 2 via link at bottom):
roomgain
Another note about this:
One issue with H or U frames is the possibility of positioning a flat surface adjacent to the other drivers. For example, if you put the 18" driver in an H or U frame with the MTM above, the lower "M" will be close to the top surface of the 'frame. Since the U/H frame is typically built with a square cross-section, its top is a flat, reflecting surface that is close to the mid driver and might cause some problems.
One potential fix for this is to build the 'frame using an equilateral triangle for the cross section. A little more complicated to build, yes. But from the viewpoint of the lower "M" driver, there would no longer be any flat surfaces that are perpendicular to the driver. Instead they would slope away at 60 degrees. I haven't thought about building a system in this way until now, but it makes sense and I might just try it myself.
This is a great idea! I had not thought about this option.
I don't mean to derail this thread so will just paste the link to my build thread here
I'm using B&C 8PE21 mids, nude for now. Quite sensitive. Measurements on post #118.
Thank you for the great idea!
I hope my post re tweeter is useful for iamvalheru.
Is there a recommended minimum gap/distance between nude drivers and adjacent surfaces to prevent unwanted delayed reflections?
As an example, for X driver operating within X frequency range, a distance greater than X will generate a response/reflection sufficiently delayed to just melt into the background and not be perceived by the brain as a second instance of the original.
On the flip side, we could probably apply broad band dispersion or absorption to the flat surface of the H-frame, but to what effectiveness?!?
Hello Charlie.
In learning about dipoles I've been doing quite a bit of reading, and your posts/threads have been important to me. As I saw the above I recalled you preferred larger midranges, so went back to your "pursuit of 20-20k dipole" thread and confirmed that back in mid 2019 your preference for midrange (200-2k) was 10" and up. What drove the change in preference?
Regards
Stupid question, isn’t having the 15” woofer suspended above the M-frame causing problems?
D
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.