Taking the Linkwitz spreadsheet for excursion related max. SPL and the datasheet of the D7608 you get the following SPLs at +- 0.4mm (linear) excursion and at +- 1.5mm (maxlimitbeforedestruction) excursion:
400Hz: lin 93dB / max 105dB
500Hz: lin 97dB / max 109dB
650Hz: lin 102dB / max 113dB
For a decent low distortion design, I never would design to exceed the lin excursion range. Instead, in practice I often designed my speakers to 2/3 of theirs drivers lin excursion limit. Some max excursion reserve is certainly nice for robustness and safety issues of a driver. But not for educated listening.
So I understand your numbers as whole-system values crossed at the given frequencies, including the lower driver's +6dB contribution:
400Hz: lin 93dB + 6dB = 99dB
500Hz: lin 97dB + 6dB = 103dB
650Hz: lin 102dB + 6dB = 108dB
I indeed would very much be interested to see the raw distortion figures of the single, isolated D7608 crossed at these three frequencies and driven to the limits of lin excursion (+-0.4mm), e.g. to 93dB @ 400Hz, 97dB @ 500Hz and 102dB @ 650Hz.
@Daihedz Thats what I've been researching and trying to quantify as soon as I have some measurements. The tuning of the back chamber has alot to do with the maximum output of the D7608. If Qtc is too low, output is left on the table which is why running it with a larger chamber or none at all results in poor low end performance.
Your posted figures are close. I did some quick measurements and have seen a max of about 2% THD @ 500 hz / 103dB with H2 dominating. This was with a Qtc of 0.8, so at minimal overshoot in response, gaining about +1 dB at rolloff. I find these limits tolerable and still better sounding than with similar sized smaller cones crossed around the same frequency.
There is a significant benefit to running 2 x D7608s in shaded TMM configuration. An additional +3 dB outout is obtained using a second driver only operating under 1k. This is what I recommend and it will also improve the upper usable limit slightly if a tapered falling response is desired (an approximated Harman curve).
The motor on the D7608 isn't the strongest being an internal ferrite magnet design. It could benefit from a neo magnet upgrade, but so far I haven't been able to completely disassemble one of these mids to access the inner motor.
I still would like to test an array of 4 x D7608s in a tapered arrangement, surrounding a tweeter capable of lower HP point. Even augmenting some other dome mids of similar design. This basically mimicks a larger diameter driver behaving as a coax and is a good option if more output at a lower HP is desired. There are quite a few optional arrangements possible if your mind is open to some less orthodox solutions.
Your posted figures are close. I did some quick measurements and have seen a max of about 2% THD @ 500 hz / 103dB with H2 dominating. This was with a Qtc of 0.8, so at minimal overshoot in response, gaining about +1 dB at rolloff. I find these limits tolerable and still better sounding than with similar sized smaller cones crossed around the same frequency.
There is a significant benefit to running 2 x D7608s in shaded TMM configuration. An additional +3 dB outout is obtained using a second driver only operating under 1k. This is what I recommend and it will also improve the upper usable limit slightly if a tapered falling response is desired (an approximated Harman curve).
The motor on the D7608 isn't the strongest being an internal ferrite magnet design. It could benefit from a neo magnet upgrade, but so far I haven't been able to completely disassemble one of these mids to access the inner motor.
I still would like to test an array of 4 x D7608s in a tapered arrangement, surrounding a tweeter capable of lower HP point. Even augmenting some other dome mids of similar design. This basically mimicks a larger diameter driver behaving as a coax and is a good option if more output at a lower HP is desired. There are quite a few optional arrangements possible if your mind is open to some less orthodox solutions.
Last edited:
I understand that D7608 requires treatment to get the best out of it.
Makes me think about MDM55: neodymium magnet; higher Xmax; complete with enclosure. However halve the Sd.
An MTM with MDM55 and MDT12, crossover around 2k5, high pass some 800Hz.....
Makes me think about MDM55: neodymium magnet; higher Xmax; complete with enclosure. However halve the Sd.
An MTM with MDM55 and MDT12, crossover around 2k5, high pass some 800Hz.....
When you are happy with the sound of the MDT12 ...
With 2 MDM55 you can cross lower as 800Hz for a "normal" HiFi Speaker. It's not a super low THD driver but nice and stable.
The MDM55 is a very good 2" dome. As said before, not the lowest word in THD, but its distortion profile isn't offending to the ears. Thats what matters most. The upper end is slso very smooth and extended past 6k, so you can cross it above 4k. Its also a solid 89 dB/2.8V, which is more than plenty of sensitivity for 3 ways with larger drivers.
Also don't overlook the Hivi DMB-A. For $25 its a steal. Not as clean as the Morel, but close.
Many of these mid domes have a dominant 3rd order peak but its not off putting. The motors usually suffer from a little magnetic asymmetry, mainly caused by the inboard magnet setup. The short VC winding height to upper gap ratio further compounds this.
I've used the MDM55 in alot of monitor builds. They require an 800 hz 2nd order min HP which can be lowered to 650 hz using a pair shaded in one cabinet. ⁵ By the time you buy a pair of these, you're out of oocketalmost $250. You can get quite a better mid for that.
Also don't overlook the Hivi DMB-A. For $25 its a steal. Not as clean as the Morel, but close.
Many of these mid domes have a dominant 3rd order peak but its not off putting. The motors usually suffer from a little magnetic asymmetry, mainly caused by the inboard magnet setup. The short VC winding height to upper gap ratio further compounds this.
I've used the MDM55 in alot of monitor builds. They require an 800 hz 2nd order min HP which can be lowered to 650 hz using a pair shaded in one cabinet. ⁵ By the time you buy a pair of these, you're out of oocketalmost $250. You can get quite a better mid for that.
Were you keen enought to get this far while disassembling?
It seems that this perforated dome right behid the membrane provides some radiating impedance modification. What is that good for? Does it prevent the membrane to get too much dyscentric while moving? Immediately I thought of slightly widening a set of these holes as a modding experiment.
Scan-Speak does not publish magnet field strength inside the gap, so it is not obvious wheter the magnet's front iron plate might be magnetically saturated or not. If it already was fully saturated, then probably neo magnets would not change anything. And if not, it would indeed be another and attractive modding option. "Neoing" the magnet assembly until full saturation of the iron around the gap might not only improve the sensitivity, but also the linearity of the drive.
I'm pretty sure the upper plate is already saturated. Its thin and (aside from further investigation of the yokes) the magnet strength is already as high as the front pole plate will pull through. I tested several bucking magnet schemes and they all worsened the specs.
The front internal screen is likely there for dynamic loading of the dome surface for even drive anf to avoid other rocking modes. They claim it's mainly for protection of the dome, but thats an unnecessarily expensive way to do such a thing. It has to be for the sake of some.sort of an acoustical benefit.
The back felt is critical to the flow restriction damping of the dome, so the Qts is manageable with such a small motor. The dome is highly dependent on the chamber to modify the lower rolloff and it has to be a certain range of volume for any decent results. I can't imagine how many people have dismissed this driver after trying to run it without a chamber, open back. It performs poorly in that configuration.
That being said, its definitely not a fragile driver, that's for sure. It can take over 30W continuous for quite a while and put out some serious acoustic levels.
The front internal screen is likely there for dynamic loading of the dome surface for even drive anf to avoid other rocking modes. They claim it's mainly for protection of the dome, but thats an unnecessarily expensive way to do such a thing. It has to be for the sake of some.sort of an acoustical benefit.
The back felt is critical to the flow restriction damping of the dome, so the Qts is manageable with such a small motor. The dome is highly dependent on the chamber to modify the lower rolloff and it has to be a certain range of volume for any decent results. I can't imagine how many people have dismissed this driver after trying to run it without a chamber, open back. It performs poorly in that configuration.
That being said, its definitely not a fragile driver, that's for sure. It can take over 30W continuous for quite a while and put out some serious acoustic levels.
My understanding of ferromagnetic magnets was that if a ferromagnetic driver motor structure gets saturated, then you cannot further increase the strength of the magnetic field generated in the gap by using an even stronger magnet? If so, then stronger NEO's would be useless. Maybe I am wrong?
Yes, it seems to be a robust beast. Datasheet states power handling for 100h noise test (IEC 18.4) is 80W with 2nd order HP Butterworth @ 500Hz. That's a lot of thermal energy dissipated ... For a more daily use, this figure may point to an only modest dynamic thermal compression behavior, and thus being a part of the praised dynamic vertues of this driver.
Your thoughts on the magnet strength and motor saturation are likely correct. Hence my findings regarding the extra magnet theory.
80W continuous crossed at 500 hz BW2 is roughly 30W continuous without a crossover. Thats according to the noise signal they use to test with and the power spectrum that is passed through the crossover point stated.
80W continuous crossed at 500 hz BW2 is roughly 30W continuous without a crossover. Thats according to the noise signal they use to test with and the power spectrum that is passed through the crossover point stated.
I'm done with prototyping a design using some cone shapes made from foam board and epoxy. A terminated TL is IMO the best shape for dampening without over doing it, losing too much precious lower mid output. The rear volume is much less than originally speced out. Its enough for a 600 hz 2nd order HP with plenty of output to achieve a 100 dB+ at reasonably low distortion numbers.
Hi Profiguy,
I’m officially even more confused between the cone shapes and closed transmission line 🥲
Do you have an image that could help visualise the concept? What is the internal volume and how are you setting up the TL, is it a simple internal insert or more complex?
It all sounds quite exciting, will plans be an available to 3D print?
Jules
I’m officially even more confused between the cone shapes and closed transmission line 🥲
Do you have an image that could help visualise the concept? What is the internal volume and how are you setting up the TL, is it a simple internal insert or more complex?
It all sounds quite exciting, will plans be an available to 3D print?
Jules
The rear chamber shape is similar to those used on some B&W tweeters. The SB MD60N dome mid also has a type of 1/4 wave TL chamber. Its a bit more tricky on a larger dome without making it too large for most enclosure depths.
Yes. I'm anticipating the availability of an STL file for people to print their own. Depends on several other factors.
Yes. I'm anticipating the availability of an STL file for people to print their own. Depends on several other factors.
Aha nice, that stimulates a slightly more focused mental image. I used to own the B&W Nautilus 804’s with the tear drop tweeters, great speakers for a medium room.
Consider me signed up for the STL, when do you expect it to be available?
Have a great weekend!
Jules
Consider me signed up for the STL, when do you expect it to be available?
Have a great weekend!
Jules
Yes, sorry. Its been a crazy month. I'm going to send my design parameters to someone and hopefully they're not too busy. I wanted to keep the design on the simple side so it will print easy and fit most enclosure depths. Of course you'll be welcome to experiment with dampening as it substantially affects lower mid characteristics. Its specifically limiting excursion which is necessary to get desired low end cutoff with decent output.
I just posted this in the dome midrange thread -
I got the D7608 rear chamber CAD files back from the generous fellow on here who graciously volunteered his time to make this happen. Out of respect for his privacy and sanity that he doesn't get swamped with tons of PMs here, I'll let him decide whether he wants to be mentioned or is able to answer any questions regarding printing them.
I'll be posting the STL file once the design is tested with absorption media in place and it fits the D7608 as intended. The chamber design is for personal use only. Any commercial use should be cleared with me first.
Initially I'll be printing a pair of test chambers from PLA to verify the overall fitment and perform some measurements. I took inspiration from the newer Seas T35C002 design, as it requires minimal mechanical dampening to achieve decent decay spectrum and avoid most common standing wave modes.
Here's a teaser -
I got the D7608 rear chamber CAD files back from the generous fellow on here who graciously volunteered his time to make this happen. Out of respect for his privacy and sanity that he doesn't get swamped with tons of PMs here, I'll let him decide whether he wants to be mentioned or is able to answer any questions regarding printing them.
I'll be posting the STL file once the design is tested with absorption media in place and it fits the D7608 as intended. The chamber design is for personal use only. Any commercial use should be cleared with me first.
Initially I'll be printing a pair of test chambers from PLA to verify the overall fitment and perform some measurements. I took inspiration from the newer Seas T35C002 design, as it requires minimal mechanical dampening to achieve decent decay spectrum and avoid most common standing wave modes.
Here's a teaser -
Looks great, it is quite cute what is the volume? My current chamber design for the D7608 is 1.1L with triangular diffusers at the back, I think it was supposed to provide a Qts around 0.6. I have a very economic local 3d printing service nearby so I will likely give these a shot.
Awesome work and knowledge sharing (as usual) @profiguy.
Awesome work and knowledge sharing (as usual) @profiguy.
Last edited:
Volume is about 350 ml net. The Qtc will land around .65 - .70 and highly depends on dampening media quantity. Going lower will only throw away sensitivity and precious output potential. I'd recommend ABS filament for these as it glues very well with contact adhesive or cyanoacrylic resin.