TW034 - datasheet says it's perfectly wide radiating until at least 5kHz. It's not that much bigger as a 1".
Just print/build a blind plate to be sure? Or Quick and dirty - close the hole with a few layers of Gaffa 🤓
Using the D7608 with the factory flange helps with the mid not diffracting quite as bad. Its just a pain trying to get CTC spacing tighter with it. At lower xover points its not so critical since it doesn't require as tight of spacing with the longer wavelengths.
I've been looking for my D7608 measurment data without the flange. I used Clio to test and measure, but it doesn't save the data in the same way as other platforms do, which is a big problem with Clio.
I've been looking for my D7608 measurment data without the flange. I used Clio to test and measure, but it doesn't save the data in the same way as other platforms do, which is a big problem with Clio.
I have one extra D7608 which I'm using for testing before I commit to a baffle layout design. Part of me is torn between the wider soundstage of a narrow baffle with side mounted, outward radiating mids and MTM arranged mids also in the vertical plane.
There's also the Horbach Keele MTM arrangement which uses special filter slopes and specific CTC spacing in both the vertical and horizontal along with tilting the array by 22 degrees. I've built a few on wall speakers using this topology and it has some big advantages over other arrangements.
There's also the Horbach Keele MTM arrangement which uses special filter slopes and specific CTC spacing in both the vertical and horizontal along with tilting the array by 22 degrees. I've built a few on wall speakers using this topology and it has some big advantages over other arrangements.
Do you have a link to a build log or anything for those speakers? Sounds very similar to what I am planning for an in-wall design specifically utilizing the D7608s.
Currently busy with medical crap. Not fun.
Easiest way to remove the stock felt is saturate the felt with isopropyl alcohol with the magnet SIDE DOWN. Otherwise you'll ruin the dome from the back side. Once the felt is wet, it can be slid off with a bondo spreader or old credit card edge.
Easiest way to remove the stock felt is saturate the felt with isopropyl alcohol with the magnet SIDE DOWN. Otherwise you'll ruin the dome from the back side. Once the felt is wet, it can be slid off with a bondo spreader or old credit card edge.
Trying to catch up on this build now having a better baffle layout in mind. The circular array on the larger baffle is what I'm going for after some extensive contemplating. I'd rather have the more direct sound of a wide baffle with better directivity and more low mid gain for the lowest possible HP on the D7608s. Having 4 of them will cut excursion requirements for a given SPL on half and allow roughly 500 hz 2nd order HP. Sensitivity is predicted to be 94 - 95 dB / 2.8V.
So an update on some more D7608 research.
I've been testing the limits of the dome's output capabilities and power handling. Its done a bit on the crude side of testing methods, but nonetheless I've found its possible to get 103 dB with a 500 hz sine and still have tolerable HD levels along with remaining within the excursion limits. The trick here is using the appropriate back chamber. This has lately been the focus of my research.
This dome is extremely sensitive to back chamber volume and requires the correct chamber size to get the best from, as I've been saying throughout my testing and sharing of info. I've been finalizing a chamber design to have a CAD file (and STL file) designed for the D7608 with the help of a generous fellow member on here. This will permit the use of the original felt material already on the dome. Some very light stuffing with fiberglass, Rockwool or sheeps wool is all the dampening that is required.
The new findings have led me to the conclusion that its possible to use crossover points down to 650 hz 2nd order while still allowing for healthy output levels ie. peaks of 105 dB.
The power handling will be no issue either. I've tortured a pair of D7608s with over 30W of continuous input for close to an hour without any permanent parameter changes. Its actually a very stout driver with good dynamic capabilities and can tolerate a little abuse.
At limited SPLs, response down to 400 hz is possible with a 98 - 100 dB output limit. The dome resonance Fs with proper chamber is 350 hz after driver break in. Free air Fs ends up at about 200 hz without chamber. Driver break in is a must. The dome responds well to loosening up with a few hours of pink noise at around 85 dB. It just gets better as it further loosens up.
I'll post the appropriate info once I have more of it.
I've been testing the limits of the dome's output capabilities and power handling. Its done a bit on the crude side of testing methods, but nonetheless I've found its possible to get 103 dB with a 500 hz sine and still have tolerable HD levels along with remaining within the excursion limits. The trick here is using the appropriate back chamber. This has lately been the focus of my research.
This dome is extremely sensitive to back chamber volume and requires the correct chamber size to get the best from, as I've been saying throughout my testing and sharing of info. I've been finalizing a chamber design to have a CAD file (and STL file) designed for the D7608 with the help of a generous fellow member on here. This will permit the use of the original felt material already on the dome. Some very light stuffing with fiberglass, Rockwool or sheeps wool is all the dampening that is required.
The new findings have led me to the conclusion that its possible to use crossover points down to 650 hz 2nd order while still allowing for healthy output levels ie. peaks of 105 dB.
The power handling will be no issue either. I've tortured a pair of D7608s with over 30W of continuous input for close to an hour without any permanent parameter changes. Its actually a very stout driver with good dynamic capabilities and can tolerate a little abuse.
At limited SPLs, response down to 400 hz is possible with a 98 - 100 dB output limit. The dome resonance Fs with proper chamber is 350 hz after driver break in. Free air Fs ends up at about 200 hz without chamber. Driver break in is a must. The dome responds well to loosening up with a few hours of pink noise at around 85 dB. It just gets better as it further loosens up.
I'll post the appropriate info once I have more of it.
I also did a foam board mock up of an MTM using a pair of D7608s and a SB29RDNC. The vertical dispersion is very good which ironically is helped by the mid's early rolloff up top. The vertical spacing and crossover point is what I've been trying to dial in. Too tight crossed too low, it lobes badly. A vertically limited tweeter would be best for higher crossover points. I'd limit dome LP to 3200 and use the built in rolloff to your advantage. More on this soon along with measurements of the chambered mid.
FYI, yanks and canucks: the flangeless version appears available from https://solen.ca/en/products/d76089200
Great work, very interesting for reasonably priced but killer 3 way designs 💪 when do you expect the back chamber CAD file to be available? (Still have not glued a pair of speakers together using the D7608’s admire, there is still plenty of time to adjust back chambers 🙂
@Shertzy @motokok It will be likely a few weeks on my side and depending on the other gracious fellow's schedule. The chamber surprisingly ends up being much smaller than anticipated.
I've realized this is also the key to possibly horn loading it for more lower mid support and directivity. Up until now I've discarded the idea of applying this, being confused by the lower mid behavior, thinking it was the composition of the dome and its rigidity working against it.
As most of us know, the big disadvantage of a mid driver with such a wide radiation pattern down low becomes a challenge mating it with a larger LF driver with a narrow pattern where it crosses, having an abrupt shift in directivity. A decent 10" woofer would be no problem with a 600 hz HP. There are quite a few decent woofers which fit this application ie. from Faital, Seas and ScanSpeak.
I realize many people dismiss an MTM arrangement based on the popular misconception of lobing ruining the vertical radiation pattern. With correct spacing and filter point, it's no issue in midfield listening applications, considering how the drivers sum at that distance. Nearfield is potentially an issue, but nobody would logically consider using a large 3 way speaker in the nearfield.
The issue of excursion limited output potential with a 3" driver at lower crossover points is an issue that plagues most larger mid domes. This makes designing a crossover challenging and is the main reason why so many shy away from dome mids. The only way around this is using a larger motor to compensate so that more VC overhang can provide the low end output needed for a lower crossover point.
The D7608 is definitely not a driver for beginner speaker builders and requires a lot of forethought to end up with a successful design. Its initial limitations would keep most of us from considering its use for a higher end design. Only the savvy designer knows the potential of a good large mid dome and would consider it over the many excellent cone mids already out there. I do admit being one of those people who prefer the road less traveled approach, wanting something more unique and exclusive.
The problem is that out of the few people who would even consider a large mid dome, most would quickly dismiss the D7608 based on its factory data and FR curves. The only reason I didn't gloss over it is having heard it in other higher end speakers which left a lasting good impression on me.
Large dome mids can sound considerably more accurate than most cone mids. Once you've heard a good 3 way with a mid dome, you'll likely be hooked on the resolution, imaging and depth it can achieve. Many 3 ways with cone mids and even higher end 2 way systems don't have the precision, refinement and resolution of a good 3 way with larger mid dome. Unless you've experienced this, you won't know the true potential of a large mid dome, even if it isn't the most exclusive model of driver out there.
I've heard better midrange from systems using D7608s than quite a few planars and electrostats. The younger crowd won't know this because of their lack of exposure to designs using mid domes. They would think that AMTs are the greatest thing that came along since the IPhone. The old timers need to pass the mid dome torch to the gen X/Y/Zs so they can experience how good they actually can be. This may convince manufacturers like ScanSpeak and Bliesma to continue producing their current models in the long term future. I feel an obligation to spread the word about the D7608 and other mid domes. Even the lowly Hivi DMB-A is capable of better performance in its limited range than many other mids.
I've realized this is also the key to possibly horn loading it for more lower mid support and directivity. Up until now I've discarded the idea of applying this, being confused by the lower mid behavior, thinking it was the composition of the dome and its rigidity working against it.
As most of us know, the big disadvantage of a mid driver with such a wide radiation pattern down low becomes a challenge mating it with a larger LF driver with a narrow pattern where it crosses, having an abrupt shift in directivity. A decent 10" woofer would be no problem with a 600 hz HP. There are quite a few decent woofers which fit this application ie. from Faital, Seas and ScanSpeak.
I realize many people dismiss an MTM arrangement based on the popular misconception of lobing ruining the vertical radiation pattern. With correct spacing and filter point, it's no issue in midfield listening applications, considering how the drivers sum at that distance. Nearfield is potentially an issue, but nobody would logically consider using a large 3 way speaker in the nearfield.
The issue of excursion limited output potential with a 3" driver at lower crossover points is an issue that plagues most larger mid domes. This makes designing a crossover challenging and is the main reason why so many shy away from dome mids. The only way around this is using a larger motor to compensate so that more VC overhang can provide the low end output needed for a lower crossover point.
The D7608 is definitely not a driver for beginner speaker builders and requires a lot of forethought to end up with a successful design. Its initial limitations would keep most of us from considering its use for a higher end design. Only the savvy designer knows the potential of a good large mid dome and would consider it over the many excellent cone mids already out there. I do admit being one of those people who prefer the road less traveled approach, wanting something more unique and exclusive.
The problem is that out of the few people who would even consider a large mid dome, most would quickly dismiss the D7608 based on its factory data and FR curves. The only reason I didn't gloss over it is having heard it in other higher end speakers which left a lasting good impression on me.
Large dome mids can sound considerably more accurate than most cone mids. Once you've heard a good 3 way with a mid dome, you'll likely be hooked on the resolution, imaging and depth it can achieve. Many 3 ways with cone mids and even higher end 2 way systems don't have the precision, refinement and resolution of a good 3 way with larger mid dome. Unless you've experienced this, you won't know the true potential of a large mid dome, even if it isn't the most exclusive model of driver out there.
I've heard better midrange from systems using D7608s than quite a few planars and electrostats. The younger crowd won't know this because of their lack of exposure to designs using mid domes. They would think that AMTs are the greatest thing that came along since the IPhone. The old timers need to pass the mid dome torch to the gen X/Y/Zs so they can experience how good they actually can be. This may convince manufacturers like ScanSpeak and Bliesma to continue producing their current models in the long term future. I feel an obligation to spread the word about the D7608 and other mid domes. Even the lowly Hivi DMB-A is capable of better performance in its limited range than many other mids.
Ähm ... 😏
I have a big 3 way in the front wall of my small studio room and that's a great solution. Doesn't take extra space and provides enough dynamic down to 20Hz for mixing and recording. But 4 very good 20cm drivers are not loud enough (for recording, good for mixing) so it will be 4x 12" soon.
I tested the MTM configuration and did a listening comparison - for closer distances (about 2m) I definitely prefer one mid dome cause of clear phantom sound source. Maybe you can test this with an active configuration - with 2 domes I got a vertical "dimension" in the sound, with one dome it get's a pinpoint between tweeter and mid dome.
I really wanted to go with 2 midranges ... but not any more.
@IamJF I understand your situation. It resonates with my findings using MTM arrangements of various drivers, usually struggling to achieve the vertical lobing (or lack thereof) in an attempt to produce a stable height across the entire midrange. If you do get stuck with a lobe large enough across the vertical plane, it has to favor the chosen listening angle, whether standing, sitting or both. I often find myself tilting speakers to find the best plane where imaging is most convincing, not just holographic and yet vaque. A successful MTM arrangement can mitigate alot of fussing around with speaker placement and aiming, but it requires very small footprint drivers to get the CTC spacing where it needs to be along with a well executed and implemented crossover. This usually means active DSP filtering is needed.
Phantom Image height is strongly dependent on ceiling, floor and wall reflections as a combined factor. You must treat all boundaries equally if you want reasonably convincing and accurate localization of instruments and the acoustic space they're presented in. The reason that most engineers prefer speaker over headphone monitoring is the realization of a convincing stereo image (at least in acoustical music). There are also quite a large number of well recorded and mixed rock/pop recordings which attempt to recreate a somewhat believable virtual soundscape that adheres to some set of acoustic rules.
Unfortunately there are a large number of modern recordings which are recorded/mixed/mastered in small bedroom studio setups using a laptop, headphones and possibly a cheap little pair of plastic monitors. These types of productions usually just sound weird to the trained ear and don't make much sense from an expert POV. They can be equally entertaining to listen to, mainly on headphones, which makes them sound trippy and euphonic, especially when lots of reverb and delay were used. Your typical bedroom production will almost always sound like this if there was no provision for accurate monitoring through a decent set of speakers. Even a basic little set of personal monitors would provide a stable phantom image sufficient enough to reference and mix with. Unless you're attempting to produce binaural recordings, headphones don't cut it. Equally so, if the monitoring isn't capable of revealing the acoustic space in all three planes, it won't be sufficient for mixing to a higher standard.
Even more important than vertical coverage is a flat relative phase response in the midrange from 500 - 8k at the mixing location. Otherwise you can kiss any decent results goodbye when trying to master a direct 2 channel recording for a given acoustic space ie. live performances with solo performers / singers and especially so with full orchestras and choirs.
Phantom Image height is strongly dependent on ceiling, floor and wall reflections as a combined factor. You must treat all boundaries equally if you want reasonably convincing and accurate localization of instruments and the acoustic space they're presented in. The reason that most engineers prefer speaker over headphone monitoring is the realization of a convincing stereo image (at least in acoustical music). There are also quite a large number of well recorded and mixed rock/pop recordings which attempt to recreate a somewhat believable virtual soundscape that adheres to some set of acoustic rules.
Unfortunately there are a large number of modern recordings which are recorded/mixed/mastered in small bedroom studio setups using a laptop, headphones and possibly a cheap little pair of plastic monitors. These types of productions usually just sound weird to the trained ear and don't make much sense from an expert POV. They can be equally entertaining to listen to, mainly on headphones, which makes them sound trippy and euphonic, especially when lots of reverb and delay were used. Your typical bedroom production will almost always sound like this if there was no provision for accurate monitoring through a decent set of speakers. Even a basic little set of personal monitors would provide a stable phantom image sufficient enough to reference and mix with. Unless you're attempting to produce binaural recordings, headphones don't cut it. Equally so, if the monitoring isn't capable of revealing the acoustic space in all three planes, it won't be sufficient for mixing to a higher standard.
Even more important than vertical coverage is a flat relative phase response in the midrange from 500 - 8k at the mixing location. Otherwise you can kiss any decent results goodbye when trying to master a direct 2 channel recording for a given acoustic space ie. live performances with solo performers / singers and especially so with full orchestras and choirs.
Last edited:
The center of the mid domes is 16,5cm apart, crossover is of course digital and in the 1,8-2kHz area. All walls and ceiling are highly dampened, the are no uneven reflections. It's an unusual case and room.
But still you hear a vertical dimension with 2 midrange drivers and not a pinpoint source. The sound comes from an area and not a single point.
That's probably not to hear in 3m distance in a normal room ... but hey, we don't want it easy, we want it perfect. 🤓
I can see why even the most thought out symmetrically driven MTM (both mids playing same FR) would still have some small lobes past 45 deg off axis. That would certainly throw off vertical power response.
Since a perfect CTC of less than 1/4 WL is almost always not possible, choosing the best compromise CTC which won't produce significantly fractured lobes across the vertical plane is very difficult.
I often find its a better compromise to cross higher so you work in the mid's narrower dispersion angle for less combing. That is, provided the mid doesn't have any significant off axis dips of its own. Then, when picking a suitable tweeter for this scheme, its best to use one which already has narrowing dispersion characteristics right around the mids narrowing dispersion point. This may sound impractical and against most teachings which advise the opposite - crossing both drivers where the widest dispersion is observed from both drivers in their overlapping FR at rolloff.
The benefit of doing an MTM the first mentioned way produces minimal lobing (given the drivers have similar narrowing dispersion points at chosen xover). The drawback is a much more narrow horizontal sweet spot, but there are other ways to mitigate this ie. with multiple drivers splayed out in the horizontal plane. This can become complicated and also has its obvious pluses and minuses depending on your goals and priorities. It can however fix other issues if executed correctly and improve lower mid sensitivity along with better dynamic range, having more emissive surface area down low. All I can say is dont knock it till you try it. I've built some successful designs using splayed multiple drivers, crossed in their more narrow dispersion range to the HF driver. Its not easy to do though and often more practical solutions are available.
Since a perfect CTC of less than 1/4 WL is almost always not possible, choosing the best compromise CTC which won't produce significantly fractured lobes across the vertical plane is very difficult.
I often find its a better compromise to cross higher so you work in the mid's narrower dispersion angle for less combing. That is, provided the mid doesn't have any significant off axis dips of its own. Then, when picking a suitable tweeter for this scheme, its best to use one which already has narrowing dispersion characteristics right around the mids narrowing dispersion point. This may sound impractical and against most teachings which advise the opposite - crossing both drivers where the widest dispersion is observed from both drivers in their overlapping FR at rolloff.
The benefit of doing an MTM the first mentioned way produces minimal lobing (given the drivers have similar narrowing dispersion points at chosen xover). The drawback is a much more narrow horizontal sweet spot, but there are other ways to mitigate this ie. with multiple drivers splayed out in the horizontal plane. This can become complicated and also has its obvious pluses and minuses depending on your goals and priorities. It can however fix other issues if executed correctly and improve lower mid sensitivity along with better dynamic range, having more emissive surface area down low. All I can say is dont knock it till you try it. I've built some successful designs using splayed multiple drivers, crossed in their more narrow dispersion range to the HF driver. Its not easy to do though and often more practical solutions are available.