Sherical Speakers project based on B&W drivers
I have built some spherical speakers and thought there might be some interest in the tale and have a couple queries about my implementation.
Relevant photos can be found here:
http://www.picturevillage.com/photos/main.php/v/cnocbui_/Eyeball+Speakers/
The project arose from my purchasing some B&W car audio drivers on eBay with the intention of putting them in my car. Back in the 80's B&W made a brief foray into car audio then quite quickly pulled out.
For various reasons, I subsequently decided not to use the drivers in the car. So I came up with the idea of making spherical enclosures for them with the overall aim of emulating the heads on B&W's 802s - the octagonal ones on the series 80 and series 3 (matrix) models.
The heads on the 802 have a shell of styrene lined with a fibreglass fibre reinforced concrete.
I decided to make my spheres from carbon fibre / water clear epoxy resin - both because it is stiff and because I like the visual aesthetics of the way CF reflects light. I also happened to have these materials to hand ;-)
I have tried concrete before and didn't want to use it again, so I decided to make each speaker housing from two spheres, one nested inside the other, with the cavity between the two filled with sand.
I decided It would be best to make a half mould of a smooth spherical object and lay up the CF/epoxy on the inside of the mould. So I found a suitably sized plastic ball and made a half mould from it in silicone rubber and made a custom frame to hold the mould in shape while it was in use.
I then made two half spheres from the mould to form the back half of each outer sphere. I then made two flat rings of CF on a flat surface to form mounting flanges for the drivers.
I then placed each finished flange in the bottom of the mould (very tricky!) and then proceeded to lay up CF so the flange was incorporated in the front partial half sphere.
I then found another smaller plastic ball of the right size to act as a form for the inner spheres but this time I coated half the ball with a release agent then laid up a relatively thin layer of CF compared to the outer shell. Again I made two complete half shells then I made two rings of CF - like an egg poaching ring, but bigger, about 12mm deep - then placed each ring on top of the ball/form and laid up CF from the equator up to the ring so it became part of the resulting partial half sphere. The ring forms a projecting throat on the internal sphere to couple it to the inside of the front flange and positions the internal sphere centrally within the outer sphere.
Once the parts of the inner spheres were complete, I joined each front and rear section to form the completed spheres. I then searched for any possible holes, plugged them then tested the integrity by filling with water. I didn't want any sand leaking!
I then attached each internal sphere to the back of the flange of each front outer shell via the coupling rings on the former. Then I was finally able to attach each rear half section of outer shell to the front section to form a completed outer sphere. I then wrapped the visible joint with CF and epoxy to make it invisible. Then there was a bit of sanding, application of more epoxy and several further iterations of these steps. Finally I applied a flow coat of epoxy to provide a glassy finish.
Shortly after I had acquired the B&W drivers, I bought some MB Quart 25mm tweeters - again on eBay - to pair with them. With the change in plan I reasoned I could still use them, so I fashioned some CF Shells to house the tweeters and mounted these on top of the main housings, doing my best to offset them so as to align the voice coils.
The B&W drivers came with two sets of B&W crossovers intended to match the drivers to tweeters in their MASS range of automotive speakers. I used one set of these crossovers - removed from their enclosures - inside the main housings. The characteristics of the tweeters I have, and those B&W intended the crossovers for, are slightly different. Would someone be able to tell me if they are likely to be reasonably suitable? There is a discrepancy of 2 ohms between the impedance of the actual and intended tweeters and also a difference in their efficiencies.
Midrange: B&W LM50: 120mm Kevlar cone with Kapton voice coil former with a total moving mass of 4.3 grams for both. Frequency Response (with equaliser) 45hz - 15khz: Impedance 4 ohms : Sensitivity 90db for 1 watt at 1m. Magnet 1250 grams
The equaliser is a black box that goes inline, with two input and two output connections. I presume it is a filter that prevents frequencies lower than 45hz reaching the driver? Is this called a high-pass filter?
Tweeter: MB Quart RTC 25: 25mm titanium diaphragm: Frequency Response 920hz - 40khz: Impedence 6 ohms: Sensitivity 86 db: DC R 5.3: Fs 1650: Qms 1.91: Qes 2.66: Qts 1.11 Prms 80
I admit ignorance as to the meaning of the specs that follow the sensitivity so would be grateful if someone could enlighten me. I suspect Fs might be the resonant frequency and Prms might be the constant power handling?
The tweeter the crossover was intended for: B&W LT30 based on the TZ26: Frequency response to 22khz: Impedence 8 ohms: sensitivity 90db for 1 watt at 1m.
Crossover: B&W LX 30: 3rd order Butterworth: Crossover frequency 3khz: high level adjustment switch in 1.5db steps (3 positions -1.5, 0, +1.5)
The Crossovers also include what B&W refer to as their APOC (audio powered overload circuit) which is sort of an auto resetting circuit breaker that protects both drivers from overload.
Sphericals
I have built some spherical speakers and thought there might be some interest in the tale and have a couple queries about my implementation.
Relevant photos can be found here:
http://www.picturevillage.com/photos/main.php/v/cnocbui_/Eyeball+Speakers/
The project arose from my purchasing some B&W car audio drivers on eBay with the intention of putting them in my car. Back in the 80's B&W made a brief foray into car audio then quite quickly pulled out.
For various reasons, I subsequently decided not to use the drivers in the car. So I came up with the idea of making spherical enclosures for them with the overall aim of emulating the heads on B&W's 802s - the octagonal ones on the series 80 and series 3 (matrix) models.
The heads on the 802 have a shell of styrene lined with a fibreglass fibre reinforced concrete.
I decided to make my spheres from carbon fibre / water clear epoxy resin - both because it is stiff and because I like the visual aesthetics of the way CF reflects light. I also happened to have these materials to hand ;-)
I have tried concrete before and didn't want to use it again, so I decided to make each speaker housing from two spheres, one nested inside the other, with the cavity between the two filled with sand.
I decided It would be best to make a half mould of a smooth spherical object and lay up the CF/epoxy on the inside of the mould. So I found a suitably sized plastic ball and made a half mould from it in silicone rubber and made a custom frame to hold the mould in shape while it was in use.
I then made two half spheres from the mould to form the back half of each outer sphere. I then made two flat rings of CF on a flat surface to form mounting flanges for the drivers.
I then placed each finished flange in the bottom of the mould (very tricky!) and then proceeded to lay up CF so the flange was incorporated in the front partial half sphere.
I then found another smaller plastic ball of the right size to act as a form for the inner spheres but this time I coated half the ball with a release agent then laid up a relatively thin layer of CF compared to the outer shell. Again I made two complete half shells then I made two rings of CF - like an egg poaching ring, but bigger, about 12mm deep - then placed each ring on top of the ball/form and laid up CF from the equator up to the ring so it became part of the resulting partial half sphere. The ring forms a projecting throat on the internal sphere to couple it to the inside of the front flange and positions the internal sphere centrally within the outer sphere.
Once the parts of the inner spheres were complete, I joined each front and rear section to form the completed spheres. I then searched for any possible holes, plugged them then tested the integrity by filling with water. I didn't want any sand leaking!
I then attached each internal sphere to the back of the flange of each front outer shell via the coupling rings on the former. Then I was finally able to attach each rear half section of outer shell to the front section to form a completed outer sphere. I then wrapped the visible joint with CF and epoxy to make it invisible. Then there was a bit of sanding, application of more epoxy and several further iterations of these steps. Finally I applied a flow coat of epoxy to provide a glassy finish.
Shortly after I had acquired the B&W drivers, I bought some MB Quart 25mm tweeters - again on eBay - to pair with them. With the change in plan I reasoned I could still use them, so I fashioned some CF Shells to house the tweeters and mounted these on top of the main housings, doing my best to offset them so as to align the voice coils.
The B&W drivers came with two sets of B&W crossovers intended to match the drivers to tweeters in their MASS range of automotive speakers. I used one set of these crossovers - removed from their enclosures - inside the main housings. The characteristics of the tweeters I have, and those B&W intended the crossovers for, are slightly different. Would someone be able to tell me if they are likely to be reasonably suitable? There is a discrepancy of 2 ohms between the impedance of the actual and intended tweeters and also a difference in their efficiencies.
Midrange: B&W LM50: 120mm Kevlar cone with Kapton voice coil former with a total moving mass of 4.3 grams for both. Frequency Response (with equaliser) 45hz - 15khz: Impedance 4 ohms : Sensitivity 90db for 1 watt at 1m. Magnet 1250 grams
The equaliser is a black box that goes inline, with two input and two output connections. I presume it is a filter that prevents frequencies lower than 45hz reaching the driver? Is this called a high-pass filter?
Tweeter: MB Quart RTC 25: 25mm titanium diaphragm: Frequency Response 920hz - 40khz: Impedence 6 ohms: Sensitivity 86 db: DC R 5.3: Fs 1650: Qms 1.91: Qes 2.66: Qts 1.11 Prms 80
I admit ignorance as to the meaning of the specs that follow the sensitivity so would be grateful if someone could enlighten me. I suspect Fs might be the resonant frequency and Prms might be the constant power handling?
The tweeter the crossover was intended for: B&W LT30 based on the TZ26: Frequency response to 22khz: Impedence 8 ohms: sensitivity 90db for 1 watt at 1m.
Crossover: B&W LX 30: 3rd order Butterworth: Crossover frequency 3khz: high level adjustment switch in 1.5db steps (3 positions -1.5, 0, +1.5)
The Crossovers also include what B&W refer to as their APOC (audio powered overload circuit) which is sort of an auto resetting circuit breaker that protects both drivers from overload.
Sphericals
sreten said:Hi,
You need to research crossover design and baffle step compensation.
/sreten.
Does this mean you don't know the answers to my questions but 'think' the answer might lie within theses subject areas?
Or does it mean you could answer the questions because you 'know' the answers and are therefore 'certain' these subject areas contain the answers?
spherical speakers
That's a "busy" crossover for this setup. probably built with all sorts of frequency tailoring to use for its specific intended purpose.
I think sreten is correct in that you need to do more research to determine whether this can be used for the intended purpose.
BTW: you don't mention how they sound in your post.
That's a "busy" crossover for this setup. probably built with all sorts of frequency tailoring to use for its specific intended purpose.
I think sreten is correct in that you need to do more research to determine whether this can be used for the intended purpose.
BTW: you don't mention how they sound in your post.
Re: spherical speakers
They sound better than I had hoped. I would say they are getting on for as good as the B&W 802 series 80 'heads' I was trying to emulate. It is hard to say for certain as I sold them so can not direct compare any more. They sound better than any speaker of comparable size I have encountered - which is probably only a handful so it may not be saying much.
I have a pair of B&W LM1 (1980's vintage) speakers that presently serve as the output for the TV. The spheres are significantly superior sonically in every regard. I plan to make stands for the Spheres and replace the LM1s with them.
I have played around briefly with some software to do a frequency response measurement of them but I doubt the results are hugely accurate as I don't have a proper microphone and just used my headphones. I am probably measuring the headphones and room as much as the speakers.
I asked my wife about the problem and she says the tweeter I am using is likely about 5db down on the originally intended tweeter and if you add the discrepancy in efficiency it would be more like 9db down - ouch!
Guess I might have a look for a better matched crossover. Odd though, because they don't sound like they are lacking ithat much in the high end.
auplater said:
They sound better than I had hoped. I would say they are getting on for as good as the B&W 802 series 80 'heads' I was trying to emulate. It is hard to say for certain as I sold them so can not direct compare any more. They sound better than any speaker of comparable size I have encountered - which is probably only a handful so it may not be saying much.
I have a pair of B&W LM1 (1980's vintage) speakers that presently serve as the output for the TV. The spheres are significantly superior sonically in every regard. I plan to make stands for the Spheres and replace the LM1s with them.
I have played around briefly with some software to do a frequency response measurement of them but I doubt the results are hugely accurate as I don't have a proper microphone and just used my headphones. I am probably measuring the headphones and room as much as the speakers.
I asked my wife about the problem and she says the tweeter I am using is likely about 5db down on the originally intended tweeter and if you add the discrepancy in efficiency it would be more like 9db down - ouch!
Guess I might have a look for a better matched crossover. Odd though, because they don't sound like they are lacking ithat much in the high end.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.