gomper,
Very interesting stuff you got going there. Like your approach towards building the driver. Keep up the good work and look forward to more updates.
Regarding cabinets, I understand you want to keep it simple. In addition to BR boxes, MLTL type cabinets are also an option; they can be designed as such that they are not that difficult to build and can add some low frequency extension and weight to the overall sound. All depends on the driver parameters though.
Very interesting stuff you got going there. Like your approach towards building the driver. Keep up the good work and look forward to more updates.
Regarding cabinets, I understand you want to keep it simple. In addition to BR boxes, MLTL type cabinets are also an option; they can be designed as such that they are not that difficult to build and can add some low frequency extension and weight to the overall sound. All depends on the driver parameters though.
Last edited:
Measurements and Cabinets
Driver measurements! Here are the T/S params - these first two fully assembled units have almost identical readings:
Piston Diameter: 8.543 in
Mmd: 18.5g
SPL: 95.57 1W/1m
Re: 7.67 Ohms
Fs: 19.51Hz
Qts: .1817
Qes: .1902
Qms: 4.071
Le: .3429mH (10k)
Mms: 23.51g
Vas: 19.2 cu ft
The rim dimension is 11 inches in diameter, and each driver weighs about 33 pounds. I think I’m happy with these numbers except for Qts, which is very low, and when modeled as a 4th order vented cabinet it suggests somewhat diminished bass response. My goal is to bring that value up to .22 or so, and my reading so far suggests that it can be done pretty easily with a) more copper windings on the voice coil (which affects both weight and likely Xmax, or b) adding mass elsewhere to the moving assembly (which affects Fs) or c) by simply adding a series resistor to the positive lead running to the driver, perhaps 3 Ohms or so in my case. That seems like the right option, since it only appears to alter Qes and as a result, Qts. The thing is, it sounds too good to be true, to me - someone without an understanding of the behavior. Does the amp when seeing a higher Re output higher compensating power, or something like that? In any case, I hope it does work and appreciate feedback from you smarties - when I model the cabinet design with .22 Qts in WinSpeakerz it gives me a beautiful result:
Pretty flat to 50Hz with +/- 3dB at 40Hz. That's going to work for me.
Cabinet design! The first design looked like this, with a .75 inch thick low lead glass channel running through the middle, giving the effect of the port and driver bridging the santos mahogany together. Glass is of course a very dense, hard and dimensionally strong material, as is the santos mahogany (also known as Cabrueva) at 57 pounds per cubic foot, and a janka hardness of 2400.
The driver sits on a wooden arch inside. I love this design... I had the glass cut for it, and one day, I'm going to build it. I have access at a local community workshop to a great CNC router that can handle the cutting of the complicated angles and miters in the wood - but there were just so many other open issues to resolve... complicated assembly, wiring, access to the driver should it need to be serviced etc... I'll tackle it eventually, but for now I'm opting for something a bit easier to manage. These renders are poor quality and a little out of date - there's a channel routed on the back with a brass face leading from the driver to terminals near the base of the cabinet in the latest work, and I've opted for a black granite base. Same materials, but glass around the sides and top in this case, and some clean simplicity and functional benefits:
Norelco and Bao, thanks so much for the pricing questions, and everyone for the kind words - I just don't know what the cost will be yet, or if I want to sell just finished, hand-made loudspeakers as opposed to the raw drivers as well. Hey zman01, I'd love to do an MLTL cabinet eventually, and, an open baffle. I need the right visual design to come to mind. I have a fantastic idea for the open baffle, but haven't yet come up with an MLTL that exposes view of the driver and is aesthetically pleasing. When I do, I'd really love your guidance if you're willing to offer suggestions and links where I can learn the concepts. For the above renderings, I'm making good time. The wood is routed and sanded and ready for finishing, and the glass and brass all arrive on Monday. Here are a couple of pics of that in progress - I couldn't resist dropping a rim in the baffles:
By the way(!), these drivers just sitting there on my carpet (well, along with a sub picking up the low end as well) sound better than anything I've ever personally heard. The detail, imaging, transients and tone are just spectacular... after all of the hard work and care I've poured into them, it was a pretty powerful thing for me to finally hear what they sounded like.
Alright guys... it's late once again and I'm off to get some rest. Looking forward to hearing your thoughts. Thanks again for the time and encouragement, it really means a lot to me. More soon.
- Ken
Driver measurements! Here are the T/S params - these first two fully assembled units have almost identical readings:
Piston Diameter: 8.543 in
Mmd: 18.5g
SPL: 95.57 1W/1m
Re: 7.67 Ohms
Fs: 19.51Hz
Qts: .1817
Qes: .1902
Qms: 4.071
Le: .3429mH (10k)
Mms: 23.51g
Vas: 19.2 cu ft
The rim dimension is 11 inches in diameter, and each driver weighs about 33 pounds. I think I’m happy with these numbers except for Qts, which is very low, and when modeled as a 4th order vented cabinet it suggests somewhat diminished bass response. My goal is to bring that value up to .22 or so, and my reading so far suggests that it can be done pretty easily with a) more copper windings on the voice coil (which affects both weight and likely Xmax, or b) adding mass elsewhere to the moving assembly (which affects Fs) or c) by simply adding a series resistor to the positive lead running to the driver, perhaps 3 Ohms or so in my case. That seems like the right option, since it only appears to alter Qes and as a result, Qts. The thing is, it sounds too good to be true, to me - someone without an understanding of the behavior. Does the amp when seeing a higher Re output higher compensating power, or something like that? In any case, I hope it does work and appreciate feedback from you smarties - when I model the cabinet design with .22 Qts in WinSpeakerz it gives me a beautiful result:
Pretty flat to 50Hz with +/- 3dB at 40Hz. That's going to work for me.
Cabinet design! The first design looked like this, with a .75 inch thick low lead glass channel running through the middle, giving the effect of the port and driver bridging the santos mahogany together. Glass is of course a very dense, hard and dimensionally strong material, as is the santos mahogany (also known as Cabrueva) at 57 pounds per cubic foot, and a janka hardness of 2400.
The driver sits on a wooden arch inside. I love this design... I had the glass cut for it, and one day, I'm going to build it. I have access at a local community workshop to a great CNC router that can handle the cutting of the complicated angles and miters in the wood - but there were just so many other open issues to resolve... complicated assembly, wiring, access to the driver should it need to be serviced etc... I'll tackle it eventually, but for now I'm opting for something a bit easier to manage. These renders are poor quality and a little out of date - there's a channel routed on the back with a brass face leading from the driver to terminals near the base of the cabinet in the latest work, and I've opted for a black granite base. Same materials, but glass around the sides and top in this case, and some clean simplicity and functional benefits:
Norelco and Bao, thanks so much for the pricing questions, and everyone for the kind words - I just don't know what the cost will be yet, or if I want to sell just finished, hand-made loudspeakers as opposed to the raw drivers as well. Hey zman01, I'd love to do an MLTL cabinet eventually, and, an open baffle. I need the right visual design to come to mind. I have a fantastic idea for the open baffle, but haven't yet come up with an MLTL that exposes view of the driver and is aesthetically pleasing. When I do, I'd really love your guidance if you're willing to offer suggestions and links where I can learn the concepts. For the above renderings, I'm making good time. The wood is routed and sanded and ready for finishing, and the glass and brass all arrive on Monday. Here are a couple of pics of that in progress - I couldn't resist dropping a rim in the baffles:
By the way(!), these drivers just sitting there on my carpet (well, along with a sub picking up the low end as well) sound better than anything I've ever personally heard. The detail, imaging, transients and tone are just spectacular... after all of the hard work and care I've poured into them, it was a pretty powerful thing for me to finally hear what they sounded like.
Alright guys... it's late once again and I'm off to get some rest. Looking forward to hearing your thoughts. Thanks again for the time and encouragement, it really means a lot to me. More soon.
- Ken
Last edited:
Ken,
Images are visible. Few words on the cabinet design:
1. You can probably go for a more gradual bass roll-off in your speaker modelling. See what a slightly bigger cabinet and/or slightly lower tuning gives you.
2. Be careful with shallow cabinets; if cabinets are too shallow, then there is chance of reflections coming back through the thin cone of the driver.
3. How tall is your cabinet? From the images it looks like tall enough to get some quarter wave action going.
4. What about internal lining or stuffing?
Images are visible. Few words on the cabinet design:
1. You can probably go for a more gradual bass roll-off in your speaker modelling. See what a slightly bigger cabinet and/or slightly lower tuning gives you.
2. Be careful with shallow cabinets; if cabinets are too shallow, then there is chance of reflections coming back through the thin cone of the driver.
3. How tall is your cabinet? From the images it looks like tall enough to get some quarter wave action going.
4. What about internal lining or stuffing?
Thoughtful feedback, zman01 and Destroyer OS - thank you.
Destroyer OS, you're not being a downer at all - that's just the kind of suggestion I'm looking for. Forgive the ignorance, but BSC is used just to attenuate a specific frequency range, isn't it? I'm guessing you mean to suggest another use. I used BSC to take the vocal range shout out of a Tang Band 1772 driver, but that's the limit of my experience so far. I do feel like I understand the basic idea though - to account for edge reflections on the baffle from spherically radiating sound waves. Is it those baffle edges you're noticing? Please, feel free to set me straight.
I've been worried about allowing for that more gradual roll-off you mention, zman. Won't the result sound thin..? When I adjust the tuning / size I typically see a 1dB drop begin at around 150Hz that becomes steep at about 50 -60Hz.
Glad you mentioned reflections back through the cone. I was careful to account for that in the first design, and ignored it in the second. Am I correct to assume that a nonparallel surface on the opposing back panel will reduce reflections? I’m pretty attached to that sleek, shallow depth - but I know I have to come up with something, preferably a subtle solution.
The dimensions are roughly 22.75” x 52” x 8.25” in inches.
For now, I’m going without any linings or stuffing. I’d like to see how the cabinet colors the sound first, and then decide. My intuition is that the cabinet can contribute something to the overall listening experience - something slightly less accurate, but likewise more musical. I know that's a topic with strong opinions.
Destroyer OS, you're not being a downer at all - that's just the kind of suggestion I'm looking for. Forgive the ignorance, but BSC is used just to attenuate a specific frequency range, isn't it? I'm guessing you mean to suggest another use. I used BSC to take the vocal range shout out of a Tang Band 1772 driver, but that's the limit of my experience so far. I do feel like I understand the basic idea though - to account for edge reflections on the baffle from spherically radiating sound waves. Is it those baffle edges you're noticing? Please, feel free to set me straight.
I've been worried about allowing for that more gradual roll-off you mention, zman. Won't the result sound thin..? When I adjust the tuning / size I typically see a 1dB drop begin at around 150Hz that becomes steep at about 50 -60Hz.
Glad you mentioned reflections back through the cone. I was careful to account for that in the first design, and ignored it in the second. Am I correct to assume that a nonparallel surface on the opposing back panel will reduce reflections? I’m pretty attached to that sleek, shallow depth - but I know I have to come up with something, preferably a subtle solution.
The dimensions are roughly 22.75” x 52” x 8.25” in inches.
For now, I’m going without any linings or stuffing. I’d like to see how the cabinet colors the sound first, and then decide. My intuition is that the cabinet can contribute something to the overall listening experience - something slightly less accurate, but likewise more musical. I know that's a topic with strong opinions.
Last edited:
Finished, at last
It's been a very long time since my last post, and much has happened over the past several months.
The glass cabinet was an epic fail, but I learned some valuable lessons. They were looking gorgeous just before attaching the back panels:
...but the morning after that picture was taken I heard a strange, screechy noise coming from the dining room and discovered that the glass was slowly cracking.
Over about an hour, both cabinets developed long cracks all down their length. I'm betting that because the epoxy adhesive could not flex with the natural expanding and contracting of the wood, eventually too much internal stress built up and the glass simply gave out. It was crushing at the time, but for the best in the end. I removed the glass and decided to finish the cabinets as a prototype in all wood and an internal volume of about 3.75 cubic feet. That (with a temporary paper port installed) looked like this:
With no lining and minimal bracing, they didn't sound good at all. Lining and better bracing made a considerable improvement, but nonetheless... the shallow cabinet was surely not helping, the box volume was insufficient, and in the end my wife and I agreed that they looked like giant mobile phones. Back to the drawing board. I opted for a 5 cubic foot enclosure, and designed for something I already knew looked good, based on my previous work. They were finally completed yesterday, and I'm so very happy with how they look, and sound.
The bases are santos mahogany and the cabinets are cherry and brass, about 90% lined inside and likewise well-braced, tuned to 33Hz. The backs have a nice access panel to service the driver should that ever be necessary, XLR power supply connectors and gold plated binding posts. The driver is mounted with a carbon steel M10 bolt to a thick wood arch inside, very solid. Bass performance is noticeably better than the previous cabinets - with the sub crossed over at 45Hz they sound at their full best, but it isn't required. I may add a small series resistor for slightly better low end performance at the cost of a little sensitivity - I'll have to try it out and see what I prefer.
As far as the driver is concerned, I wanted to raise the QTS a little. That led me to increase the Re of the voice coil to 9.7 Ohms. The Le now comes in at about .49, not too bad. I haven't run tone sweeps through the finished speakers yet, but they certainly seem to have good treble performance. Better to my ears than the prototype did, and it was falling off at about 12kHz. Sensitivity is a little better too, 96.5 1W/1m.
The final carbon fiber / TPU spider design ended up like this:
It mates to the voice coil via a very light and thin carbon fiber collar:
I spent quite a bit of time (perhaps a month) experimenting with factory spiders in the interest of incorporating something field-proven over years of hard use, but I just couldn't get the driver to sound as good with one. Even with the softest spider I could find, further treated for softness with acetone and abrasive wool, it just couldn't match the upper range, detail, and transients with the custom version. It's definitely a part of why the driver sounds so revealing and wonderful.
Finally, a couple of videos - first, 'Life is Beautiful' by Keb' Mo', and then 'Morning' by Beck. This is recorded with no sub, just the loudspeakers alone. I hesitate to share them though... the recording on my iPhone just can't do it justice. Sitting in the room with them, these playbacks felt full down to 45Hz and were truly breathtaking.
http://kensonger.com/field_coil/keb_mo.mp4
http://kensonger.com/field_coil/beck.mp4
The next adventure is to somehow make this achievement into a small business, and never need to go back to writing software for a living. I hardly know where to begin, really - I don't even know how to price them yet. I'll code up a nice little website, build a social media presence, design the brand identity, and most importantly get them in front of people who will love them as much as I do... None of this so far would have been possible without the amazing generosity and knowledge of those around me - I'm counting on that again as I head into this next phase. Cheers, all.
- Ken
It's been a very long time since my last post, and much has happened over the past several months.
The glass cabinet was an epic fail, but I learned some valuable lessons. They were looking gorgeous just before attaching the back panels:
...but the morning after that picture was taken I heard a strange, screechy noise coming from the dining room and discovered that the glass was slowly cracking.
Over about an hour, both cabinets developed long cracks all down their length. I'm betting that because the epoxy adhesive could not flex with the natural expanding and contracting of the wood, eventually too much internal stress built up and the glass simply gave out. It was crushing at the time, but for the best in the end. I removed the glass and decided to finish the cabinets as a prototype in all wood and an internal volume of about 3.75 cubic feet. That (with a temporary paper port installed) looked like this:
With no lining and minimal bracing, they didn't sound good at all. Lining and better bracing made a considerable improvement, but nonetheless... the shallow cabinet was surely not helping, the box volume was insufficient, and in the end my wife and I agreed that they looked like giant mobile phones. Back to the drawing board. I opted for a 5 cubic foot enclosure, and designed for something I already knew looked good, based on my previous work. They were finally completed yesterday, and I'm so very happy with how they look, and sound.
The bases are santos mahogany and the cabinets are cherry and brass, about 90% lined inside and likewise well-braced, tuned to 33Hz. The backs have a nice access panel to service the driver should that ever be necessary, XLR power supply connectors and gold plated binding posts. The driver is mounted with a carbon steel M10 bolt to a thick wood arch inside, very solid. Bass performance is noticeably better than the previous cabinets - with the sub crossed over at 45Hz they sound at their full best, but it isn't required. I may add a small series resistor for slightly better low end performance at the cost of a little sensitivity - I'll have to try it out and see what I prefer.
As far as the driver is concerned, I wanted to raise the QTS a little. That led me to increase the Re of the voice coil to 9.7 Ohms. The Le now comes in at about .49, not too bad. I haven't run tone sweeps through the finished speakers yet, but they certainly seem to have good treble performance. Better to my ears than the prototype did, and it was falling off at about 12kHz. Sensitivity is a little better too, 96.5 1W/1m.
The final carbon fiber / TPU spider design ended up like this:
It mates to the voice coil via a very light and thin carbon fiber collar:
I spent quite a bit of time (perhaps a month) experimenting with factory spiders in the interest of incorporating something field-proven over years of hard use, but I just couldn't get the driver to sound as good with one. Even with the softest spider I could find, further treated for softness with acetone and abrasive wool, it just couldn't match the upper range, detail, and transients with the custom version. It's definitely a part of why the driver sounds so revealing and wonderful.
Finally, a couple of videos - first, 'Life is Beautiful' by Keb' Mo', and then 'Morning' by Beck. This is recorded with no sub, just the loudspeakers alone. I hesitate to share them though... the recording on my iPhone just can't do it justice. Sitting in the room with them, these playbacks felt full down to 45Hz and were truly breathtaking.
http://kensonger.com/field_coil/keb_mo.mp4
http://kensonger.com/field_coil/beck.mp4
The next adventure is to somehow make this achievement into a small business, and never need to go back to writing software for a living. I hardly know where to begin, really - I don't even know how to price them yet. I'll code up a nice little website, build a social media presence, design the brand identity, and most importantly get them in front of people who will love them as much as I do... None of this so far would have been possible without the amazing generosity and knowledge of those around me - I'm counting on that again as I head into this next phase. Cheers, all.
- Ken
Last edited:
Hi Ken,
Great too see progress on your project, congratulations.
Sorry to hear that the glass paneled enclosure did not work for you; however the new cabinet (is it made of plywood?) is elegant too.
As you said, difficult to judge the real quality of the sound through a smartphone mic recording, nonetheless thanks for sharing the recordings. Are you planning to take any measurements?
Best wishes in advance as you continue your journey on getting your dream transformed into an actual commercial product!
Great too see progress on your project, congratulations.
Sorry to hear that the glass paneled enclosure did not work for you; however the new cabinet (is it made of plywood?) is elegant too.
As you said, difficult to judge the real quality of the sound through a smartphone mic recording, nonetheless thanks for sharing the recordings. Are you planning to take any measurements?
Best wishes in advance as you continue your journey on getting your dream transformed into an actual commercial product!
Thanks so much, zman! The new cabinet is solid wood - an uncommon choice I know - but I've been careful to damp and brace it well, leaving only a bare area inside behind the port to draw a little warmth out of the sound. Cherry is a perfect tone wood for mids, and I'm kind of a sucker for that sound.
Yes, I'm definitely going to dust off the Omnimic and measure the frequency response sometime in the next several days. I'll post my results here, along with the T/S params on final driver. There was another improvement worth mentioning - I managed to trim off another 2 grams of weight, with a slightly lighter cone and a much better execution of the assembly. It's surprising how much extra adhesive weighs if you aren't careful about applying it.
Yes, I'm definitely going to dust off the Omnimic and measure the frequency response sometime in the next several days. I'll post my results here, along with the T/S params on final driver. There was another improvement worth mentioning - I managed to trim off another 2 grams of weight, with a slightly lighter cone and a much better execution of the assembly. It's surprising how much extra adhesive weighs if you aren't careful about applying it.
If anybody is still interested in the original thread you can follow it on my facebook page: Project Ryu - Home | Facebook
I do not visit here so often so I apologize for those who have sent a message and were waiting for reply.
Visit my facebook page as I am more active there.
I do not visit here so often so I apologize for those who have sent a message and were waiting for reply.
Visit my facebook page as I am more active there.
Hi Gomper,
my deepest respect for your work. Although measurements don´t tell the whole story, I would be interested in response and distortion measurements. Should you ever intend to make an 8-inch version, just keep us informed...
Looking forward to hear and see more of your project, and whishing you the best of success, I remain with best regards
Mattes
my deepest respect for your work. Although measurements don´t tell the whole story, I would be interested in response and distortion measurements. Should you ever intend to make an 8-inch version, just keep us informed...
Looking forward to hear and see more of your project, and whishing you the best of success, I remain with best regards
Mattes
Measurements!
Hi all,
Thank you so much for the kind words, Mattes.
It's worth mentioning that the room I'm taking these measurements in seems like it couldn't be much worse for the job - it's shallow and long with high ceilings, full of furniture and plants and things... The measurements for FR and distortion are from a single loudspeaker with a 6 Ohm series resistor, taken about 2 feet from the cone.
The field coils are being powered by 12V 50W linear power supplies that I picked up last week - I hope this PSU wasn't a poor choice, I've discarded the packaging.
12v Linear Power Supply 50W (sonicTransporter i5 or Roon Nucleus) – Small Green Computer
Omnimic documentation doesn't appear to make any recommendation regarding the volume of the sine sweep, and so I've assumed it's not much of a factor.
Driver measurements from DATS below, followed by frequency response and distortion. If the images aren't legible, feel free to use the links below each. Looking forward to hearing your thoughts, guys.
http://kensonger.com/field_coil/field_coil_dats_final.png
http://kensonger.com/field_coil/prod_cabinet_6Ohm_resist_fr.png
http://kensonger.com/field_coil/prod_cabinet_6Ohm_resist_distortion.png
Hi all,
Thank you so much for the kind words, Mattes.
It's worth mentioning that the room I'm taking these measurements in seems like it couldn't be much worse for the job - it's shallow and long with high ceilings, full of furniture and plants and things... The measurements for FR and distortion are from a single loudspeaker with a 6 Ohm series resistor, taken about 2 feet from the cone.
The field coils are being powered by 12V 50W linear power supplies that I picked up last week - I hope this PSU wasn't a poor choice, I've discarded the packaging.
12v Linear Power Supply 50W (sonicTransporter i5 or Roon Nucleus) – Small Green Computer
Omnimic documentation doesn't appear to make any recommendation regarding the volume of the sine sweep, and so I've assumed it's not much of a factor.
Driver measurements from DATS below, followed by frequency response and distortion. If the images aren't legible, feel free to use the links below each. Looking forward to hearing your thoughts, guys.
http://kensonger.com/field_coil/field_coil_dats_final.png
http://kensonger.com/field_coil/prod_cabinet_6Ohm_resist_fr.png
http://kensonger.com/field_coil/prod_cabinet_6Ohm_resist_distortion.png
Last edited:
Looks like your noise floor is too high for a swept distortion measurement. The REW beta has become quite capable, it can now do stepped sine distortion measurements, you should try using that. There is a stepped sine button in the RTA.
Also the spectrogram in REW for swept frequency response measurements is useful for identifying reflections. You can move the measurement window to exclude reflections.
Also the spectrogram in REW for swept frequency response measurements is useful for identifying reflections. You can move the measurement window to exclude reflections.
Last edited:
Hi Gomper,
thanks for posting your results. I´ll have to admit that my knowledge with regard to measurements is small, and maybe my assumptions will be corrected by others. However, I think that measurements can show many things, and for finetuning my speakers, I prefer my ears.
The most interesting measurement in my opinion are your parameters obtained with DATS, these show a lot, for example every small spike in the curve is a small resonance somewhere in your driver (in reality most of them are probably not very harmful, a CSD will show more). It also shows that you have a relatively high Le, which makes your impedance curve rising in the highs and makes your response curve falling after 10 or 11 Khz.
I would suggest to have a copper sleeve on the pole piece and in the airgap, there´s plenty of ideas and literature to be found for how to reduce the impedance rise.
All in all, your response curve looks quite good, and the general behaviour with a small peak in the treble (easy to correct, if needed) and a later decline is quite common for a larger fullrange driver.
Your PSU question is a bit difficult to answer. Your choice is probably not bad, depending on how much amperes your motors need. For my FC drivers, I started with 2 quite similar PSUs (one for each driver, don´t know why, seemed better for me...) but changed later to lithium-ion batteries. I can listen for app. 10 hours on batteries, then I have to charge. While charging, the motors are kept warm by another cheap PSU, all switched by relays with a single switch for all...
Keep up the good work! By the way, wouldn´t that all be better in an own thread?
However, I´m following with great interest...
All the best
Mattes
thanks for posting your results. I´ll have to admit that my knowledge with regard to measurements is small, and maybe my assumptions will be corrected by others. However, I think that measurements can show many things, and for finetuning my speakers, I prefer my ears.
The most interesting measurement in my opinion are your parameters obtained with DATS, these show a lot, for example every small spike in the curve is a small resonance somewhere in your driver (in reality most of them are probably not very harmful, a CSD will show more). It also shows that you have a relatively high Le, which makes your impedance curve rising in the highs and makes your response curve falling after 10 or 11 Khz.
I would suggest to have a copper sleeve on the pole piece and in the airgap, there´s plenty of ideas and literature to be found for how to reduce the impedance rise.
All in all, your response curve looks quite good, and the general behaviour with a small peak in the treble (easy to correct, if needed) and a later decline is quite common for a larger fullrange driver.
Your PSU question is a bit difficult to answer. Your choice is probably not bad, depending on how much amperes your motors need. For my FC drivers, I started with 2 quite similar PSUs (one for each driver, don´t know why, seemed better for me...) but changed later to lithium-ion batteries. I can listen for app. 10 hours on batteries, then I have to charge. While charging, the motors are kept warm by another cheap PSU, all switched by relays with a single switch for all...
Keep up the good work! By the way, wouldn´t that all be better in an own thread?
However, I´m following with great interest...
All the best
Mattes