Regarding digital filters ...
... yep, agreed to almost all.
Except for "Even low cost DSP is still way better than $500 worth
of inductors, caps and resistors." which i would like to relativize a bit,
at least in my view:
Using DSP for prototyping does not necessarily mean needing it in
the end product.
It is more a question of reason IMO. The more you need e.g. high Q
notches, bass lift, low XO frequencies etc. the more impractical and
insane a passive network gets, that is for sure.
On the other hand needing some of those "ingredients", points towards
the transducer needing refinement.
Now there is a distinction between acting as "speaker builder" or acting
(also) as "transducer maker", since the need for fancy eq is mostly tackled
at the transducer's idiosyncrasies in the best way.
A transducer needing complex eq is not a good one usually.
Also there is no reason to strictly separate both approaches:
Is it e.g. forbidden to compensate for VC inductances in an
active multiway speaker ? Could it be beneficial ?
Passive components can directly fullfill compensational tasks in
the electromechanical network (comprising drivers, box and
passive components) as a whole, making (mounted) drivers a more
resistive load, which is the intended kind of load usually.
Cheers
... yep, agreed to almost all.
Except for "Even low cost DSP is still way better than $500 worth
of inductors, caps and resistors." which i would like to relativize a bit,
at least in my view:
Using DSP for prototyping does not necessarily mean needing it in
the end product.
It is more a question of reason IMO. The more you need e.g. high Q
notches, bass lift, low XO frequencies etc. the more impractical and
insane a passive network gets, that is for sure.
On the other hand needing some of those "ingredients", points towards
the transducer needing refinement.
Now there is a distinction between acting as "speaker builder" or acting
(also) as "transducer maker", since the need for fancy eq is mostly tackled
at the transducer's idiosyncrasies in the best way.
A transducer needing complex eq is not a good one usually.
Also there is no reason to strictly separate both approaches:
Is it e.g. forbidden to compensate for VC inductances in an
active multiway speaker ? Could it be beneficial ?
Passive components can directly fullfill compensational tasks in
the electromechanical network (comprising drivers, box and
passive components) as a whole, making (mounted) drivers a more
resistive load, which is the intended kind of load usually.
Cheers
Last edited:
DSP in finished product
Hi Oliver,
Just reading your paper on Bending Wave drivers, very well thought out indeed, congratulations and thanks again!
I also agree with you that the more Eq a driver needs, it usually means that the driver needs, or at least would benefit from, further development.
But it all depends on bandwidth...ie a subwoofer only needs to cover 100Hz down so its a much easier task to get a good result with even low to medium cost.
Now try to engineer a full range driver on a low to medium budget...its going to need a lot of DSP help!
When you get to Manger levels, £500 per driver you do get very good results right out the box, no DSP required. But add some DSP and you can get way better results, this I have proved!
The £50.00 BMR ( it will soon be this or less) also gives really good results out the box, but again with some DSP you jump up a level or two.
Now the Manger requires state of the art DSP otherwise the DSP becomes the "bottle neck".
The BMR can be used with budget DSP, ie Mini DSP or PC / Mac software and give great results.
My point is the BMR 's will be sold WITH DSP so the customer can benefit from room Eq, custom crossovers and tonal / frequency balance fine tuning to his / her taste. Also some cool battery powered Ncore amps will complete the package.
Eventually, the ultimate incarnation of this will be the Zetex "one box to rule them all" where the DSP power amp module is built inside the custom tear drop shaped speaker housing...No more speaker cable!
All you need is a source with a digital out ( iPod, lap top, CD / DVD player or whatever) and you are up and running with a killer system. A good quailty A to D takes care of Vinyl/ FM / other analogue sources.
I have this on paper...all I need is the £50,000 to get it all tooled up for mass production...PM if you want to buy in!!
Cheers
Derek.
Hi Oliver,
Just reading your paper on Bending Wave drivers, very well thought out indeed, congratulations and thanks again!
I also agree with you that the more Eq a driver needs, it usually means that the driver needs, or at least would benefit from, further development.
But it all depends on bandwidth...ie a subwoofer only needs to cover 100Hz down so its a much easier task to get a good result with even low to medium cost.
Now try to engineer a full range driver on a low to medium budget...its going to need a lot of DSP help!
When you get to Manger levels, £500 per driver you do get very good results right out the box, no DSP required. But add some DSP and you can get way better results, this I have proved!
The £50.00 BMR ( it will soon be this or less) also gives really good results out the box, but again with some DSP you jump up a level or two.
Now the Manger requires state of the art DSP otherwise the DSP becomes the "bottle neck".
The BMR can be used with budget DSP, ie Mini DSP or PC / Mac software and give great results.
My point is the BMR 's will be sold WITH DSP so the customer can benefit from room Eq, custom crossovers and tonal / frequency balance fine tuning to his / her taste. Also some cool battery powered Ncore amps will complete the package.
Eventually, the ultimate incarnation of this will be the Zetex "one box to rule them all" where the DSP power amp module is built inside the custom tear drop shaped speaker housing...No more speaker cable!
All you need is a source with a digital out ( iPod, lap top, CD / DVD player or whatever) and you are up and running with a killer system. A good quailty A to D takes care of Vinyl/ FM / other analogue sources.
I have this on paper...all I need is the £50,000 to get it all tooled up for mass production...PM if you want to buy in!!
Cheers
Derek.
Hi Oliver
I also looked at your website and I was quite intrigued by the spacing pattern of the single drivers in your line array dipole. There will be still interference between each driver but it will occur at different frequencies for each diver pair minimizing interference effects. I guess only a single driver has none.
Hi Derek
If 10 BMR cost the same as a single MSW, price is not really an argument for a line array. The only advantage will be the dynamic output. I guess it depends on the distortion behavior at high SPL -not the strength of MSW but i don't know the data for any BMR design.
I am also sceptical about digital speaker correction as speaker problems are mostly complex mechanical problems. But I am completely convinced that a digital active cross over is far superior to analog ones. I am using the cross over function of my Devialet
amp (even as this function needs more development).
I also looked at your website and I was quite intrigued by the spacing pattern of the single drivers in your line array dipole. There will be still interference between each driver but it will occur at different frequencies for each diver pair minimizing interference effects. I guess only a single driver has none.
Hi Derek
If 10 BMR cost the same as a single MSW, price is not really an argument for a line array. The only advantage will be the dynamic output. I guess it depends on the distortion behavior at high SPL -not the strength of MSW but i don't know the data for any BMR design.
I am also sceptical about digital speaker correction as speaker problems are mostly complex mechanical problems. But I am completely convinced that a digital active cross over is far superior to analog ones. I am using the cross over function of my Devialet
amp (even as this function needs more development).
Hi Monteverdi,
yes and 3 drivers e.g. make up 3 "pairs", which can be made
prime to each other in distance. That thinking in "pairs" may
help visualising the distribution of distances ... of course there are
always all drivers involved in the resulting radiation pattern.
In the 'Dipol 08' configuration only the 3 upper drivers are
involved at HF.
That could btw. also be a possibillity for "arraying mangers".
Even a single (perfect) pistonic driver shows interference, as there are
reflections from the circumference, due to the cone velocity falling to
zero beyond the suspension discountinouusly.
At low frequencies with the driver small compared to wavelength it
may play little role, but at higher frequencies even a single driver
is not free from interference.
And as the cone usually does not move purely pistonic and there may
be baffle edge diffraction also at HF, things usually get more complicated
in reality.
Kind Regards
Last edited:
Hi Derek,
it's not just a paper, my bending wave transducer is a product
already. Nevertheless it is not suited for the mass market ...
I truly believe this ...
Nice idea, i wish you good luck ...
(unfortunately i am more a developer than an investor ...)
Kind Regards
Last edited:
Hi Oliver,
The Manger is supposed to simulate an infinite plane but the dampening of the surround is not perfect. I guess this causes back reflections and subsequently the frequency drops at 850 and 1700 Hz. That equates to a wavelength of 0.4 m and 0.2 m in air but I don't have any idea what the speed of sound in the membrane/disc is. Also that the MSW is a concentric design may be on of its limitations. Manger pretends that the MSW is a mature (= perfect?) design and did not implement any improvements since at least 20 years. I still like its basic sound quality.
How to improve the Manger design?
Your solution is a large dipole bending wave design. One advantage of a large plane is that the ratio of surface area to circumference is better than in small planes. That type of design would not work in the room I want to design a speaker for (dipoles should not be stuck against the wall). What I envision is something smaller and monopole. The penalty would be limited low frequencies, the need of a woofer and requiring a cross over..
The Manger is supposed to simulate an infinite plane but the dampening of the surround is not perfect. I guess this causes back reflections and subsequently the frequency drops at 850 and 1700 Hz. That equates to a wavelength of 0.4 m and 0.2 m in air but I don't have any idea what the speed of sound in the membrane/disc is. Also that the MSW is a concentric design may be on of its limitations. Manger pretends that the MSW is a mature (= perfect?) design and did not implement any improvements since at least 20 years. I still like its basic sound quality.
How to improve the Manger design?
Your solution is a large dipole bending wave design. One advantage of a large plane is that the ratio of surface area to circumference is better than in small planes. That type of design would not work in the room I want to design a speaker for (dipoles should not be stuck against the wall). What I envision is something smaller and monopole. The penalty would be limited low frequencies, the need of a woofer and requiring a cross over..
Last edited:
Hi Monteverdi,
this shows the official measurements by the manufacturer:
http://www.klangwerk.ch/images/pdf/MSW_Manger_Schallwandler.pdf
Starting on page 131 there are impedance curves and frequency responses
made by others e.g. "figure 8.27":
http://portal.tugraz.at/portal/page...ige BA PA DA/DA_Hiebel_Biegewellenwandler.pdf
There is strong modal (resonant) behaviour up to say 1500 Hz, showing
up not in the impedance plot only but also in the frequency response,
phase response and CSD plots. Above 2Khz there is fast decay and
resonant behaviour diminishes.
My own design instead is a fullrange transducer with lower cutoff frequency
far below 30Hz (dependent from positioning and room of course), having
a smooth response down to the lower frequency limit.
So we would be comparing a real fullrange transducer with a "tweeter"
or "wide range unit" in fact.
There is of course no one hindering you to use e.g. my kind of panel design
in a flat enclosure and use it from say 200Hz upwards. You can also mount
it against a wall then if you like.
______________
The problem with most small units is the modal overlap being too low up
to the midrange and even "lower highs". Thus most bending wave units act
obviously modal, where they should act "statistical" instead.
There is an analogy to the modal behaviour of listening rooms:
Like you have a hard time reproducing a recording made in a concert hall
in an acoustic small room due to low frequency performance, you also have
a hard time to reproduce a state of the art violin using a transducer which
possibly has a modal overlap significantly lower than the instrument
to be "reproduced". Of course this would not qualify as "reproduction" when
applying stringent standards.
Manger to some degree circumvents that problem in having a membrane with
slower wave propagation than sound in air. That causes resonances not
to be radiated efficiently. The price to pay is the unit beaming much like
a usual pistonic driver (tweeter).
If you want to have significantly wider dispersion than a pistonic acting
membrane of same size would have, you need membranes with wave
propagation faster than sound in air.
Unfortunately that causes modes (resonances) to be radiated more
efficiently, and the demand for high modal overlap, which can be
achieved by high modal density and an appropriate amount of damping,
increases even more.
It is obvious that a serious musical instrument we would call "loudspeaker"
(in the sense of reproduction) has to have a modal overlap by far superior
to all instruments to be reproduced, not to lay his own frequency response
and dispersion characteristics over the instrument too obviously.
The more "fast" and/or "small" the membrane gets, the higher you have to
go in frequency to call that particular unit a quality transducer.
To my own standards e.g. Manger is a transducer preferably for frequencies
higher than approx. 2Khz.
It competes in my view with conventional pistonic wideband and tweeter
units, due to smoothness in frequency response, dispersion etc.
One has to decide for himself, what the result of that competition is from
case to case ...
Many bending wave based speakers having faster and/or smaller membranes
(than Manger e.g.) are - again to my personal standards - no quality
transducers at all within the audible band.
It would be like reproducing the spectral fingerprint and dispersion
(recorded at the microphone position) of a stradivarius using a unit having
idiosyncrasies by far dominating that of the instrument.
Kind Regards
this shows the official measurements by the manufacturer:
http://www.klangwerk.ch/images/pdf/MSW_Manger_Schallwandler.pdf
Starting on page 131 there are impedance curves and frequency responses
made by others e.g. "figure 8.27":
http://portal.tugraz.at/portal/page...ige BA PA DA/DA_Hiebel_Biegewellenwandler.pdf
There is strong modal (resonant) behaviour up to say 1500 Hz, showing
up not in the impedance plot only but also in the frequency response,
phase response and CSD plots. Above 2Khz there is fast decay and
resonant behaviour diminishes.
My own design instead is a fullrange transducer with lower cutoff frequency
far below 30Hz (dependent from positioning and room of course), having
a smooth response down to the lower frequency limit.
So we would be comparing a real fullrange transducer with a "tweeter"
or "wide range unit" in fact.
There is of course no one hindering you to use e.g. my kind of panel design
in a flat enclosure and use it from say 200Hz upwards. You can also mount
it against a wall then if you like.
______________
The problem with most small units is the modal overlap being too low up
to the midrange and even "lower highs". Thus most bending wave units act
obviously modal, where they should act "statistical" instead.
There is an analogy to the modal behaviour of listening rooms:
Like you have a hard time reproducing a recording made in a concert hall
in an acoustic small room due to low frequency performance, you also have
a hard time to reproduce a state of the art violin using a transducer which
possibly has a modal overlap significantly lower than the instrument
to be "reproduced". Of course this would not qualify as "reproduction" when
applying stringent standards.
Manger to some degree circumvents that problem in having a membrane with
slower wave propagation than sound in air. That causes resonances not
to be radiated efficiently. The price to pay is the unit beaming much like
a usual pistonic driver (tweeter).
If you want to have significantly wider dispersion than a pistonic acting
membrane of same size would have, you need membranes with wave
propagation faster than sound in air.
Unfortunately that causes modes (resonances) to be radiated more
efficiently, and the demand for high modal overlap, which can be
achieved by high modal density and an appropriate amount of damping,
increases even more.
It is obvious that a serious musical instrument we would call "loudspeaker"
(in the sense of reproduction) has to have a modal overlap by far superior
to all instruments to be reproduced, not to lay his own frequency response
and dispersion characteristics over the instrument too obviously.
The more "fast" and/or "small" the membrane gets, the higher you have to
go in frequency to call that particular unit a quality transducer.
To my own standards e.g. Manger is a transducer preferably for frequencies
higher than approx. 2Khz.
It competes in my view with conventional pistonic wideband and tweeter
units, due to smoothness in frequency response, dispersion etc.
One has to decide for himself, what the result of that competition is from
case to case ...
Many bending wave based speakers having faster and/or smaller membranes
(than Manger e.g.) are - again to my personal standards - no quality
transducers at all within the audible band.
It would be like reproducing the spectral fingerprint and dispersion
(recorded at the microphone position) of a stradivarius using a unit having
idiosyncrasies by far dominating that of the instrument.
Kind Regards
This is horizontal dispersion at 0,15,60 degrees at 95 cm distance:
Model2 Messungen
The diagram above that shows frequency response at different heights above ground
85,110,140cm at same distance to the panel.
This is 30Hz to 300Hz response at the listening position (stereophonic), using
"peak hold" when averaging (white noise) around the sitting listener's contour
using a moving microphone:
Model2 Messungen
Concerning "drop" in highs: This is a diffuse field equalisation. Remember
there is virtually no beaming >4Khz thus energy radiated in the top octave
is still comparable to common monitors used for mastering in studios.
Leaving that speaker "flat on axis" would sound way overbright.
From bandwidth and smoothness of dispersion in might get obvious, that this
unit is something different than common bending wave designs. It shares
utilizing bending wave propagation, but this is also where common properties
end.
Kind Regards
Model2 Messungen
The diagram above that shows frequency response at different heights above ground
85,110,140cm at same distance to the panel.
This is 30Hz to 300Hz response at the listening position (stereophonic), using
"peak hold" when averaging (white noise) around the sitting listener's contour
using a moving microphone:
Model2 Messungen
Concerning "drop" in highs: This is a diffuse field equalisation. Remember
there is virtually no beaming >4Khz thus energy radiated in the top octave
is still comparable to common monitors used for mastering in studios.
Leaving that speaker "flat on axis" would sound way overbright.
From bandwidth and smoothness of dispersion in might get obvious, that this
unit is something different than common bending wave designs. It shares
utilizing bending wave propagation, but this is also where common properties
end.
Kind Regards
Last edited:
This example shows vertical dispersion of the DDD Transducer, which is based on the
Walsh/Ohm design.
http://img.hifitest.de/lautsprecher_stereo_german_physiks_hrs_120_bild_1292322091.jpg
Text says FR is taken "on axis", 25cm, 50cm "above the box" - unfortunatly there is
no distance given ...
http://www.hifitest.de/test/lautsprecher_stereo/german_physiks-hrs_120_2971.php#
Anyhow, those plots are pretty similar to the BMR plots shown in the posts above.
I have some close listening experience with drivers using that base design e.g. .
They all sound (very) different to my ears.
(Mainly @Derek)
If rigid or semi rigid structures having faster than sound in air wave propagation
in the frequency range under question are excited in their modes of vibration,
a lobed dispersion is inavoidable usually, if the modal overlap is too low.
In other words: The pronounced lobing itself shows, that modal overlap is insufficient.
"Balancing modes" using e.g. counterweights on a membrane is a technique in bringing
the average velocity over the membrane's area close to zero (i.e. close to the remaining
pistonic component). That may help in making the overall radiated energy reasonably flat,
but you cannot have all 3 at the same time:
1) low modal density (only few modes utilized for radiation)
2) flat energy response
3) flat anechoic response under "all likely angles in the listening window"
Unfortunately 3) would be the priority for a driver to sound uncoloured.
So we kindly neglect the most crucial point. Stating that
"musical instruments do not radiate evenly over frequencies" is very true.
But the musical instrument's character is the one to be reproduced, thus
no existent musical instrument is suitable as a benchmark in direct comparison
to a loudspeaker: The loudspeaker is not allowed to behave in the same way as
a musical instrument does.
Assume a violin produces a peak in spectrum at the microphone's position at
say 1,2Khz at the recording venue. And the reproducing speaker does the same
(or even the contrary ...) due to a certain listening position relative to
the speaker in the listening room: Is this an acceptable reproduction ?
Do you mean this will be "self averaging" because the speaker has gross flat
"energy response" and the listening room's reflections will do the rest for you ?
Does anyone think that doesn't matter because we mostly listen to orchestral
music, where the instrument's individual dispersion(s) are averaged anyhow
in the recording (by concert hall acoustics and/or multitude of instruments) ?
That may hold for certain genres - sure - but there are opportunities when
unbalanced dispersion of a speaker shows up quickly. That is (e.g.) the
presentation of solo instruments or voices, recorded in a rather close manner.
But it also shows up in good and balanced orchestral recordings.
Only listening to programme which suits a certain speaker is quite boring, right ?
Try programme that unveils the deficiencies instead, that is more interesting.
To have sung vowel's formants and also sibilants believable and without coloration
in those situations, calls for a speaker not showing up perceivable
"Eigenklang" (own sonic character ?).
What is "perceivable own sonic character" regarding to bending wave loudspeaker
designs and their specific properties ?
Rather than stating what that is - or whether i personally know it -
i would suggest analysing that very carefully before making speakers.
Has it been done in this field ? How good is "good enough" ?
If you have a 200x200 pixel resolution picture of a dalmatian dog, you might
recognize there is a black spot right at the end of the dog's tail.
Reducing the picture to 100x100 might look as if the dog's tail is just shorter,
assuming a dark background. But it may still look like a dalmatian dog
(with tail a bit short of course ...).
It depends on the context, if this is sufficient.
Kind Regards
Walsh/Ohm design.
http://img.hifitest.de/lautsprecher_stereo_german_physiks_hrs_120_bild_1292322091.jpg
Text says FR is taken "on axis", 25cm, 50cm "above the box" - unfortunatly there is
no distance given ...
http://www.hifitest.de/test/lautsprecher_stereo/german_physiks-hrs_120_2971.php#
Anyhow, those plots are pretty similar to the BMR plots shown in the posts above.
I have some close listening experience with drivers using that base design e.g. .
They all sound (very) different to my ears.
(Mainly @Derek)
If rigid or semi rigid structures having faster than sound in air wave propagation
in the frequency range under question are excited in their modes of vibration,
a lobed dispersion is inavoidable usually, if the modal overlap is too low.
In other words: The pronounced lobing itself shows, that modal overlap is insufficient.
"Balancing modes" using e.g. counterweights on a membrane is a technique in bringing
the average velocity over the membrane's area close to zero (i.e. close to the remaining
pistonic component). That may help in making the overall radiated energy reasonably flat,
but you cannot have all 3 at the same time:
1) low modal density (only few modes utilized for radiation)
2) flat energy response
3) flat anechoic response under "all likely angles in the listening window"
Unfortunately 3) would be the priority for a driver to sound uncoloured.
So we kindly neglect the most crucial point. Stating that
"musical instruments do not radiate evenly over frequencies" is very true.
But the musical instrument's character is the one to be reproduced, thus
no existent musical instrument is suitable as a benchmark in direct comparison
to a loudspeaker: The loudspeaker is not allowed to behave in the same way as
a musical instrument does.
Assume a violin produces a peak in spectrum at the microphone's position at
say 1,2Khz at the recording venue. And the reproducing speaker does the same
(or even the contrary ...) due to a certain listening position relative to
the speaker in the listening room: Is this an acceptable reproduction ?
Do you mean this will be "self averaging" because the speaker has gross flat
"energy response" and the listening room's reflections will do the rest for you ?
Does anyone think that doesn't matter because we mostly listen to orchestral
music, where the instrument's individual dispersion(s) are averaged anyhow
in the recording (by concert hall acoustics and/or multitude of instruments) ?
That may hold for certain genres - sure - but there are opportunities when
unbalanced dispersion of a speaker shows up quickly. That is (e.g.) the
presentation of solo instruments or voices, recorded in a rather close manner.
But it also shows up in good and balanced orchestral recordings.
Only listening to programme which suits a certain speaker is quite boring, right ?
Try programme that unveils the deficiencies instead, that is more interesting.
To have sung vowel's formants and also sibilants believable and without coloration
in those situations, calls for a speaker not showing up perceivable
"Eigenklang" (own sonic character ?).
What is "perceivable own sonic character" regarding to bending wave loudspeaker
designs and their specific properties ?
Rather than stating what that is - or whether i personally know it -
i would suggest analysing that very carefully before making speakers.
Has it been done in this field ? How good is "good enough" ?
If you have a 200x200 pixel resolution picture of a dalmatian dog, you might
recognize there is a black spot right at the end of the dog's tail.
Reducing the picture to 100x100 might look as if the dog's tail is just shorter,
assuming a dark background. But it may still look like a dalmatian dog
(with tail a bit short of course ...).
It depends on the context, if this is sufficient.
Kind Regards
Last edited:
And two HARP (High Aspect Ratio) models -
HiWave HIBM130H10-8/LP 1" x 5" HARP BMR Speaker 8 Ohm 297-2159
HiWave HIBM130H10-6 130mm HARP BMR Speaker 6 Ohm 297-2158
HiWave HIBM130H10-8/LP 1" x 5" HARP BMR Speaker 8 Ohm 297-2159
HiWave HIBM130H10-6 130mm HARP BMR Speaker 6 Ohm 297-2158
I got one Hiwave HIBM85C20-4/DD and made some initial tests. Shown in a open baffle. Looks quite good for the price except the peak at 9 kHz (also high distortion). In a 3l box identical above 5khz but a -10dB notch at 3.5 kHz. I am wondering what is the optimal enclosure as some data are missing on the HiWave website.
There are actually going to be 2 very different 3.5" drivers, the HIBM65C20F (in 4 and 8 ohm versions) and the HIBM85C20F (8 ohm only). When I inquired as to why HiWave told me "85C was designed a while back by an audio company licensed to design BMR using our IP and the 65C was recently designed by HiWave to fulfil some other design / market requirements not met by the 85C. Also the 85C is Neo and therefore costly (it was designed pre-Neo crisis) whereas the 65C is Ferrite and therefore in a different price bracket (designed during the Neo crisis to give us a cost competitive driver)".
The 85C looks virtually identical to the driver Cambridge Audio uses in the Minx speakers, so if I were a betting man I would say they are the "audio company" who designed it. That's speculation on my part though. Personally I like the 65C better, so I'm holding out for that. It might be a while longer because the driver "is still not shipped from the factory due to some delays with engineering mass production sign-off. We are getting there but it’s been slower than expected".
I can't attach the PDF's HiWave gave me though, because they exceed the size limit on this forum. If anyone wants a copy PM me with an email address and I'll send them that way instead.
The HiWave HIBM85C20-4/DD looks very similar to:
http://www.cotswoldsoundsystems.com/specifications/CSS_Specification_Sheet_BMR85DD_N4Y_r1.pdf
Jman are the ones you mentioned: HIBM65C20F (in 4 and 8 ohm versions) and the HIBM85C20F (8 ohm only) similar to any BMR46 on the CSS website? These are mostly Neo and seem to be also only 2.5" but these apparently have an better highend. As none of these are complete full range I would prefer better mid and high frequency behavior even if the lows would be more restricted (higher fs).
http://www.cotswoldsoundsystems.com/specifications/CSS_Specification_Sheet_BMR85DD_N4Y_r1.pdf
Jman are the ones you mentioned: HIBM65C20F (in 4 and 8 ohm versions) and the HIBM85C20F (8 ohm only) similar to any BMR46 on the CSS website? These are mostly Neo and seem to be also only 2.5" but these apparently have an better highend. As none of these are complete full range I would prefer better mid and high frequency behavior even if the lows would be more restricted (higher fs).
The 85 model was designed by Cotswold Sound Systems (you can see their models which map roughly to Hi-waves, on the "models" page - CSS). Graham Bank was the chief designer of BMR at Hi-wave, when it was NXT, and he has some kind of cross licence of the technology to use in his own company, CSS.
They are nothing to do with Cambridge, although CSS worked with Cambridge Audio to design the Minx (along with other BMR speaker systems I believe, e.g. Wilmslow audio).
You are right about Hi-wave designing the 3.5" version with ferrite to get over the neo costs, and you can see from this press release the large differential ($6.79/unit per 100 pcs) - HiWave: sensation innovation
I'm not sure exactly why some BMR models are on CSS but not Hi-wave (some kind of exclusivity, or marketing exclusivity?). I might ask James Lewis, the CEO, as I'm a shareholder (for transparency I thought I'd mention it too).
Last edited:
They do indeed look very similar, but the spec's aren't the same. Even the depth of the two drivers is different, so I suspect they may actually be different.
The 85C does look quite like the CSS drivers, most of which share a familial appearance. The 65C looks nothing like any of the CSS drivers though. The mounting flange is round not square, the frame is far more substantial (probably to handle the extra weight of the ferrite magnets) and the surround is much larger while the diaphragm is smaller. The different 65C versions -- 4 and 8 ohm -- aren't even identical in appearance; the mounting frame and flange are similar to each other, but not the same. Even the terminals and their location differs.
The PDF's I have are marked "preliminary", but they're the most current (I got them less than 2 weeks ago) so it stands to reason they should be pretty accurate to the current drivers.