An externally hosted image should be here but it was not working when we last tested it.
Hello everyone
Facebook Page -> Polymate3D - Product/Service - London, United Kingdom - 50 Photos | Facebook
Over the past 2+ years, I have been designing speaker driver prototypes using my own built 3D printer. Printed Cabinets I have already done, as has everyone else, but no one has attempted to create a full range speaker driver that actually attempts to compete with drivers you can buy off the shelf. FD61 is being developed to change this!
Not only this, but I found it very frustrating when developing speakers to know how a speaker material will effect the sound short of developing lots and lots of speakers. With this in mind, FD61 is being developed to have a interchangeable drive train as well.
Except a steel tube, copper piece, wire, screws and glue, every single component off the design is 3D printed. These include:
Cone
Former (Bobbin)
Surround
Spider
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Over 20 prototypes have come and gone, and development is almost at a stage where I will be offering these out in a kit form to be assembled at home like a airfix kit. I have decided now is the time to create a thread and share some images and my facebook page that has been created around this. I am trying to get a feel of the scope this has, and what people would like to see...and what they would not.
To give you an idea of where I am at, here are some details:
60mm driver
76dB @ 1W
Qts of 1.2 (This is continuing to drop. Attempting to get below 1)
Mms of 4g
I have built multiple different speaker cabinets before in the past, and a avid fan of full range drivers. Especially my Mark Audio Alpair 7's, but Fountek, Dayton, Hivi, Peerless, and Visaton are all brands I have worked with and developed around.
I have a WT3 analyser and a flat response microphone which I have used in the past as well to get results from the finished models to get an idea of what is working and what is not.
If you got this far, thank you very much for your time as interest. I will keep an eye on this thread, but will routinely be putting updates on my page. Any questions and queries are more than welcome.
Paul
Just an update to my original post. Back at home so can get the exact results as analyzed from the WT3 for Prototype 24:
Re: 4.018
Fs: 83.44Hz
Qts: 1.102
Qes: 1.941
Qms: 2.551
Le: 0.412mH (Copper shorting ring coming)
Mms: 3.903g
Vas: 1.046L
SPL: 76.85dB @ 1W/1m
Also, here is a link to a short video showing excursion movement at this stage. The plan is to release as a kit on kickstarter once other issues and finished quality of parts like the surround are improved. Target pricing for a single driver in kit form is £35.
Polymate3D - 3D Printed Speaker Excursion Test - FD61 P24 | Facebook
Re: 4.018
Fs: 83.44Hz
Qts: 1.102
Qes: 1.941
Qms: 2.551
Le: 0.412mH (Copper shorting ring coming)
Mms: 3.903g
Vas: 1.046L
SPL: 76.85dB @ 1W/1m
Also, here is a link to a short video showing excursion movement at this stage. The plan is to release as a kit on kickstarter once other issues and finished quality of parts like the surround are improved. Target pricing for a single driver in kit form is £35.
Polymate3D - 3D Printed Speaker Excursion Test - FD61 P24 | Facebook
Hey lifting
Thanks for the message. I am using a mixture of materials and printing thicknesses depending on the part. This is both layer height and nozzle diameters.
The Cone is printed in a carbon fibre composite with a density of 1.37g/cm2. This has a high tensile modulus which helps with high frequency response. Another material I like the sound off is a wood like filament which has a much lower density of 0.9g/cm2. This helps improve SPL. Interesting tidbit - The speaker cone when printed consists of over 160 layers. Here is a picture of the wood like material in a earlier prototype:
The Former that the voice coil has been glued to has been an interesting part due to the distinct difference in sound depending on the material. For lower power testing I have used the wood like filament, but for higher power testing I use polycarbonate. This is due to a softening temperature of 120c which gives me enough scope for this form of driver.
Surround and spider are printed in a 85A Shore hardness flexible filament. Attempting to get the level of compliance I have wanted has proved tricky.
The basket and part that holds the magnets is currently being printed in PLA due to its rigidness and ease of printing, as are my speaker cabinet designs. I end driver will move to PETG to get away from the troublesome low melting temperature that PLA has.
@ Pinkmouse
Thanks for your interest. If you have facebook your welcome to like it and get any updates on there. I plan to update here anyway though if not.
Any other details anyone wants, or doubts/concerns on the design or development just let me know. I want to make it not only a 3D printed full range driver. I want to make it compete and open a new world of driver interchangeability and customization. Actual print finish on certain parts are planed to be improved as well.
Thanks for the message. I am using a mixture of materials and printing thicknesses depending on the part. This is both layer height and nozzle diameters.
The Cone is printed in a carbon fibre composite with a density of 1.37g/cm2. This has a high tensile modulus which helps with high frequency response. Another material I like the sound off is a wood like filament which has a much lower density of 0.9g/cm2. This helps improve SPL. Interesting tidbit - The speaker cone when printed consists of over 160 layers. Here is a picture of the wood like material in a earlier prototype:
The Former that the voice coil has been glued to has been an interesting part due to the distinct difference in sound depending on the material. For lower power testing I have used the wood like filament, but for higher power testing I use polycarbonate. This is due to a softening temperature of 120c which gives me enough scope for this form of driver.
Surround and spider are printed in a 85A Shore hardness flexible filament. Attempting to get the level of compliance I have wanted has proved tricky.
The basket and part that holds the magnets is currently being printed in PLA due to its rigidness and ease of printing, as are my speaker cabinet designs. I end driver will move to PETG to get away from the troublesome low melting temperature that PLA has.
@ Pinkmouse
Thanks for your interest. If you have facebook your welcome to like it and get any updates on there. I plan to update here anyway though if not.
Any other details anyone wants, or doubts/concerns on the design or development just let me know. I want to make it not only a 3D printed full range driver. I want to make it compete and open a new world of driver interchangeability and customization. Actual print finish on certain parts are planed to be improved as well.
You could try the Filaprint PLA+ rather than PetG. Prints really well, and I'm very impressed with the dimensional and temperature stability and strength once annealed. The semi-matt finish looks nice as well.
I have definitely had a good look at annealing HTPLA to improve the heat issue, but the Polycarbonate is currently good to 120c, and the driver is coping perfectly fine up to distortion on a TA2020 amplifier which is beyond what the driver needs to cope with. The wood like material however would be limited to around 6W. Okay for this driver, but its too easy to be damaged.
In the case of later on producing a bigger driver which need more power handling, I have materials lined up to cope up to around 190c.
Now matt finish is another story. Im not a fan of glossy finishes, so its definitely something I will consider. HIPS is also a great material for matt finishes if you can deal with the warping and potential layer cracks. Heated chamber is the way to go with that kind of thing.
The plan is a kit for the driver, but cabinets are open to be purchased or built yourself. Would making the cabinet on your 3D printer appeal, or would you go traditional with wood?
Paul
In the case of later on producing a bigger driver which need more power handling, I have materials lined up to cope up to around 190c.
Now matt finish is another story. Im not a fan of glossy finishes, so its definitely something I will consider. HIPS is also a great material for matt finishes if you can deal with the warping and potential layer cracks. Heated chamber is the way to go with that kind of thing.
The plan is a kit for the driver, but cabinets are open to be purchased or built yourself. Would making the cabinet on your 3D printer appeal, or would you go traditional with wood?
Paul
Been a while but P25 is underway.
I have been extensively reassessing each component in the aim to improve SPL, lower inductance, and extend the frequency range. P24 started to roll off early at 4Khz like a midrange driver, with -10dB @ 8Khz. Due to this I have been designing multiple speaker cone designs, then moving on some to production. These cone profiles have then been tried on 4 different materials, including 2 which push the limits of how high I can go in respect to Young's modulus (Stiffness) of the material.
I have also been working on different methods to improve the print quality of different components in different materials. Please see below the latest cone profile, former and surround, along with it next to P24 to show print quality differences.
Paul
I have been extensively reassessing each component in the aim to improve SPL, lower inductance, and extend the frequency range. P24 started to roll off early at 4Khz like a midrange driver, with -10dB @ 8Khz. Due to this I have been designing multiple speaker cone designs, then moving on some to production. These cone profiles have then been tried on 4 different materials, including 2 which push the limits of how high I can go in respect to Young's modulus (Stiffness) of the material.
I have also been working on different methods to improve the print quality of different components in different materials. Please see below the latest cone profile, former and surround, along with it next to P24 to show print quality differences.
Paul
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Hello TimbreAudio
It is something I have looked at for sure but I find the results harder to predict the result. Was debating maybe a flat surface with a spiraled support on the back to vary to distance. For now I will stick to the cone design, but I will certainly come back to BMR!
This next cone design if things go well should extend to 10khz. Possibly further. Hopefully will get this done this weekend. Results will be posted as soon as I have them.
Paul
It is something I have looked at for sure but I find the results harder to predict the result. Was debating maybe a flat surface with a spiraled support on the back to vary to distance. For now I will stick to the cone design, but I will certainly come back to BMR!
This next cone design if things go well should extend to 10khz. Possibly further. Hopefully will get this done this weekend. Results will be posted as soon as I have them.
Paul
P25 Results & 6 Way Shootout
Hello everyone
Big update post for you all. P25 is now up and running using a diamond dust PLA composite cone and former. SPL and Q values were worse, but this was expected due to using the same magnet motor but with a taller winding, resulting in a slightly lower inductance. results you see below are done on a Blue Snowball microphone so please use this only for comparison between one another. I attempted to use my old EMM-6 but the results were not right at all, reading nothing beyond 8khz.
Front view of the latest P25 prototype
Side view of P25 showing its mild steel galvanized pipe and the light permeating through the cone material. Possible back lit cone anyone?
So next up is the frequency response of P25, along with the results of P24 and 4 other drivers. The aim for P25 was to improve high frequency response.
Drivers:
FD61 - P25 = Diamond PLA composite prototype
FD61 - P24 = Carbon Fibre composite prototype
Dayton PS95 = My reference driver. Great full range driver
Tectonic BMR65 = Nice full range BMR
Bose Driver = Sounddock 2.55" drivers
Samsung HTK = Cheapo driver used in designs such as the J4550
So the plan was to have 2 of my own, 2 good well known drivers, and 2 OEM drivers so I can see where my driver fits in at this stage, and see what needs to be my next focus. Results were done at as close to the same volume as possible with the distance between driver and microphone being the same.
Click on to see in more detail
P25 = Dark Blue
P24 = Grey
PS95 = Dark Green
BMR65 = Light Green
Samsung = Orange
Bose = Light Blue
So what am I taking away from this?
1) P25 is definitely an improvement from P24 in high frequency extension
2) OEM drivers Lack in the lower end and fall between my high end frequency performance and better brands
3) There is still work to do on the magnet motor to lower the bass bump P24 and P25 both have, although P25 has also improved in this respect
4) Overall the driver is very capable when you consider the companies I am comparing it too.
So whats on the plans for P26?
First step will be fitting and testing P25 again with a different motor arrangement and with a dust cap fitted. Then extra tweaks to the cone profile to help improve rigidity further. The surround will be attempted with a thinner wall to improve compliance and lower Q, and the magnet motor is having a copper piece centrally fitted along with a larger steel tube to improve the magnetic gap strength.
Steel tubes. The left was used in earlier prototypes. The middle one is the current one for P24 and P25. The right are new tubes just come in. 3.2mm thick walls compared to previous 2.0mm.
As always any questions, feedback, input and criticisms are welcome. Hope you have enjoyed the update, and I will hopefully be back in a couple of weeks with P26
Hello everyone
Big update post for you all. P25 is now up and running using a diamond dust PLA composite cone and former. SPL and Q values were worse, but this was expected due to using the same magnet motor but with a taller winding, resulting in a slightly lower inductance. results you see below are done on a Blue Snowball microphone so please use this only for comparison between one another. I attempted to use my old EMM-6 but the results were not right at all, reading nothing beyond 8khz.
An externally hosted image should be here but it was not working when we last tested it.
Front view of the latest P25 prototype
An externally hosted image should be here but it was not working when we last tested it.
Side view of P25 showing its mild steel galvanized pipe and the light permeating through the cone material. Possible back lit cone anyone?
So next up is the frequency response of P25, along with the results of P24 and 4 other drivers. The aim for P25 was to improve high frequency response.
An externally hosted image should be here but it was not working when we last tested it.
Drivers:
FD61 - P25 = Diamond PLA composite prototype
FD61 - P24 = Carbon Fibre composite prototype
Dayton PS95 = My reference driver. Great full range driver
Tectonic BMR65 = Nice full range BMR
Bose Driver = Sounddock 2.55" drivers
Samsung HTK = Cheapo driver used in designs such as the J4550
So the plan was to have 2 of my own, 2 good well known drivers, and 2 OEM drivers so I can see where my driver fits in at this stage, and see what needs to be my next focus. Results were done at as close to the same volume as possible with the distance between driver and microphone being the same.
Click on to see in more detail
P25 = Dark Blue
P24 = Grey
PS95 = Dark Green
BMR65 = Light Green
Samsung = Orange
Bose = Light Blue
So what am I taking away from this?
1) P25 is definitely an improvement from P24 in high frequency extension
2) OEM drivers Lack in the lower end and fall between my high end frequency performance and better brands
3) There is still work to do on the magnet motor to lower the bass bump P24 and P25 both have, although P25 has also improved in this respect
4) Overall the driver is very capable when you consider the companies I am comparing it too.
So whats on the plans for P26?
First step will be fitting and testing P25 again with a different motor arrangement and with a dust cap fitted. Then extra tweaks to the cone profile to help improve rigidity further. The surround will be attempted with a thinner wall to improve compliance and lower Q, and the magnet motor is having a copper piece centrally fitted along with a larger steel tube to improve the magnetic gap strength.
An externally hosted image should be here but it was not working when we last tested it.
Steel tubes. The left was used in earlier prototypes. The middle one is the current one for P24 and P25. The right are new tubes just come in. 3.2mm thick walls compared to previous 2.0mm.
As always any questions, feedback, input and criticisms are welcome. Hope you have enjoyed the update, and I will hopefully be back in a couple of weeks with P26
Small update in relation to the magnet motor improvements.
I fitted the new thicker tube and run tests against the thinner version and results came out almost identical. I increased the winding length using a thicker wire and results came out almost identical again, meaning that saturation is not the issue in improving performance.
Up until this point, I have run a very large magnetic gap and tried to compensate for this, but my next step will have to be to reduce this gap. This should be too much of a issue as my alignment system works fairly well already, and there are improvements to this as well.
The next stage is solely going to be improving my magnet motor whilst maintaining or reducing moving mass.
Stay tuned to see what comes next!
Paul
I fitted the new thicker tube and run tests against the thinner version and results came out almost identical. I increased the winding length using a thicker wire and results came out almost identical again, meaning that saturation is not the issue in improving performance.
Up until this point, I have run a very large magnetic gap and tried to compensate for this, but my next step will have to be to reduce this gap. This should be too much of a issue as my alignment system works fairly well already, and there are improvements to this as well.
The next stage is solely going to be improving my magnet motor whilst maintaining or reducing moving mass.
Stay tuned to see what comes next!
Paul
Hello. 6 weeks and a LOT has happened in my personal life, but also to FD61. If you not interested in the details and want to see it in action, here is a link
YouTube
The larger steel tube did not help improve performance, but did allow extra space in the magnetic gap. With this in mind, I created a new frame which would improve my cone/former alignment further, ignored the cone profile, and got to work on improving the magnet motor power.
After a few iterations I have now got to this stage:
This new design has allowed me to reduce the magnetic gap width which has helped improve power output. This design also extends out through the center, making this a phase plug based design. This change also reduces the former length reducing mass. This version was 3.75g, and there is potential to reduce this.
Next step was to hook it up to the WT3 analyser and check the results:
With a SPL of 78.7dB @ 1W, this is the most powerful model so far. I am using the Aura NS3 as my benchmark in respect to SPL, as I feel this means I am matching a driver built by a manufacturer which is intended to the full range area, and not the 76dB Tang Band have on a very heavy 3" subwoofer design.
To finish off the youtube video and sound is off an older smartphone, so take it with a grain of salt, but hopefully it will give an insight to where the driver is at.
As always, all and any questions / concerns are welcome. Hope you have enjoyed the update
Paul
YouTube
The larger steel tube did not help improve performance, but did allow extra space in the magnetic gap. With this in mind, I created a new frame which would improve my cone/former alignment further, ignored the cone profile, and got to work on improving the magnet motor power.
After a few iterations I have now got to this stage:
An externally hosted image should be here but it was not working when we last tested it.
This new design has allowed me to reduce the magnetic gap width which has helped improve power output. This design also extends out through the center, making this a phase plug based design. This change also reduces the former length reducing mass. This version was 3.75g, and there is potential to reduce this.
Next step was to hook it up to the WT3 analyser and check the results:
An externally hosted image should be here but it was not working when we last tested it.
With a SPL of 78.7dB @ 1W, this is the most powerful model so far. I am using the Aura NS3 as my benchmark in respect to SPL, as I feel this means I am matching a driver built by a manufacturer which is intended to the full range area, and not the 76dB Tang Band have on a very heavy 3" subwoofer design.
To finish off the youtube video and sound is off an older smartphone, so take it with a grain of salt, but hopefully it will give an insight to where the driver is at.
As always, all and any questions / concerns are welcome. Hope you have enjoyed the update
Paul
80dB is finally within reach!!!
Just a quick update as this driver sets and gets some hours under its belt. Done analysis with my WT3 analyser and we have achieved 80dB in the magnet motor test driver. Alignment method is working very well with a hot swap of the magnet motor working perfectly. Inductance is high but there are some ideas around that.
Hope this update is a nice insight. If you are reading these and have any questions please don't hesitate to ask. Im open to all and any questions or general discussion.
Paul
Just a quick update as this driver sets and gets some hours under its belt. Done analysis with my WT3 analyser and we have achieved 80dB in the magnet motor test driver. Alignment method is working very well with a hot swap of the magnet motor working perfectly. Inductance is high but there are some ideas around that.
An externally hosted image should be here but it was not working when we last tested it.
Hope this update is a nice insight. If you are reading these and have any questions please don't hesitate to ask. Im open to all and any questions or general discussion.
Paul
Hello 3eepoint
Thank you very much . I do use FEMM 4.2 and recently adjusted the settings to a axisymmetric layout rather than planar. This is helping a lot, but there is differences in the magnets I am using and the ones FEMM has set out as standard. I have also been lucky enough to have access to play around with FINECONE in the past which helps my understanding of cone profiles and material effects before I actually try to implement them.
I will keep updating as improvements happen. The plan is still to make it available to buy in a kit form. Likely around the £35 mark and make sure there are options for different material cone and profiles. I would like to effectively make a learners ecosystem where you can experiment with the driver, the enclosure types and materials on a smaller scale. Then use what you learn to better effect with a bigger system.
Paul
Thank you very much
. I do use FEMM 4.2 and recently adjusted the settings to a axisymmetric layout rather than planar. This is helping a lot, but there is differences in the magnets I am using and the ones FEMM has set out as standard. I have also been lucky enough to have access to play around with FINECONE in the past which helps my understanding of cone profiles and material effects before I actually try to implement them.I will keep updating as improvements happen. The plan is still to make it available to buy in a kit form. Likely around the £35 mark and make sure there are options for different material cone and profiles. I would like to effectively make a learners ecosystem where you can experiment with the driver, the enclosure types and materials on a smaller scale. Then use what you learn to better effect with a bigger system.
Paul
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.