• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Planar magnetic dual membrane bass panel

Joppe Peelen

Active Member
Joined
Sep 21, 2020
Messages
135
Likes
125
Location
Den Haag , Netherlands
Posting planar magnetic contraptions here is always scary... since well, people here know way more about measuring then i do, and they will NOT ever beat a woofer in a box... (to a extend :) then again... there are people liking them, like me, maybe because its something i can make from scratch and not buy some components.

Anyhow, i made a contraption (current size including frame 60x40 cm) , based on an idea i had for some time. instead of using metal with magnets on one side... (single ended , driven... pretty non linear) or metal and magnets on both sides.(PP push pulll, although it should be actually PUSH PUSH...) and the idea i had. no metal, but suspend the magnets in the middle and use 2 membranes. so thats what i did. downside... resonance shoots up, because its not having the same air amount damping it. secondly but for me most interesting is, one foil moves away from the magnets at excursion.. while the other moves towards it. in theory the field should kind of be equal at all times if you would look at the whole thing as one driver, now this is never 100% true :) but at least the idea is somwhat true, in this video where i play a tune i also measure the LRS compared to the contraption i made. now for all single ended planars lows 2e order is by far the biggest distortion (usually because of non linear motor design) and as you can see this panel with 6 times the x max, had less 2e order. anyhow. i dont want to write it all out.

here is a video with tune recorded with binaural mics on the cough where there is still to much low end , i also added some medium range measurements from a LRS vs this panel (70 cm dB not calibrated. but i did not want to push the LRS more) just to see what happens in the distortion.

In video 1 i include a measurement of the first prototype of 1 membrane VS 2. where you can see based on 2e order distortion the motor did get more linear. especially for a single ended magnet structure (i still use the same amount of magnets a single ended speaker would use)

if anyone has a good idea of damping the Main resonance? adding some felt in between the foils is instantly to much damping, i even tried a small piece of silk screen with low hole count.. it damps to much and i also think it hurts the way they work in tandem/distortion, you will damp a part of the foil but i got the feeling it moves less regular over its surface when doing so


Video 1 some building and first try of the idea (this one had to high resonance)
Video 6 Done building a version 3 and played 2 tunes and show the measurement LRS vs one of these small panels. i dont have any other commercial planar magnetic here to compare so thats what it is.
Video 3 i try to explain my idea... :)
 
Last edited:
These seem like really nice midbass drivers. Have you considered seeing what the off-axis looks like, especially at a corner? DI takes for granted the circular shape of most drivers.
 
I was thinking the same thing re midbass and would be interested in those off axis measurements. I use ribbon tweeters and in my case they are difficult to match to a midrange driver before it starts beaming.

Great project, love your enthusiasm. I've seen worse measuring speakers here on ASR.
 
BTW - you remind me A LOT of a good friend of mine, I did a double take when your voice came on as well. I know your doppelgänger!
 
if anyone has a good idea of damping the Main resonance? adding some felt in between the foils is instantly to much damping, i even tried a small piece of silk screen with low hole count.. it damps to much and i also think it hurts the way they work in tandem/distortion, you will damp a part of the foil but i got the feeling it moves less regular over its surface when doing so
Do like Magnepan - use "tie-downs", where the foil is locked down. The tie-downs are spread out in order to spread the resonance of the foil over a number of frequencies.

Since you have LRS's, take a look and you know how its done.
 
These seem like really nice midbass drivers. Have you considered seeing what the off-axis looks like, especially at a corner? DI takes for granted the circular shape of most drivers.


2 people want to make mid/bass from them :( how low should that mid bass play ? Because we dont need 6mm xmax :) for midbass and we dont need dual foils. This tester was mainly meant to play lows up to midrange, or a tweeter that crosses really low +- 350hz something like the ones sitting on top.
 
Do like Magnepan - use "tie-downs", where the foil is locked down. The tie-downs are spread out in order to spread the resonance of the foil over a number of frequencies.

Since you have LRS's, take a look and you know how its done.

thanks !, i know that :) but its not enough to damp the main resonance, . it gets split , and lowered a bit, but its still to high, unless you use a plate with tiny holes on the back of the magnets and speaker cloth :) the metal plates damps the most. because it has rather limited airflow.
 
Is the main resonance a real resonance that will ring in the time domain, or just a frequency response peak? If the latter - EQ will fix that for you perfectly.
 
Very interesting, well done.
I suppose to make it more Isobaric, and improve efficiency, one could use triple magnet setup! Possibly smaller magnets to keep the weight down.
With some Fazor-ish idea thrown in.
Just thinking out loud . . .
 
Very interesting, well done.
I suppose to make it more Isobaric, and improve efficiency, one could use triple magnet setup! Possibly smaller magnets to keep the weight down.
With some Fazor-ish idea thrown in.
Just thinking out loud . . .
yes you could , although it does not become more or less isobaric :) it would be more efficient and propably more linear, but when doing so its even more costly and more effort then a regular PP with magnets front and back and only one foil :) so it beats its purpose im afraid (or my purpose , cheap but slightly better result then single ended. by the looks of it)
 
Last edited:
2 people want to make mid/bass from them :( how low should that mid bass play ? Because we dont need 6mm xmax :) for midbass and we dont need dual foils. This tester was mainly meant to play lows up to midrange, or a tweeter that crosses really low +- 350hz something like the ones sitting on top.
It looks perfect to me from 80hz to 500hz. This is where you cross to a subwoofer anyways, no? This still means they put out substantial bass and can play lower. But it also means they can cross to a lower panel or sub and play very loud and clear, unless I don’t get something. Isn’t this a very good result for a panel that size?
 
it is :) but it already beats a much larger panel (that people used for low end) so they could be used as low end down to 50 at least. :) at least thats my idea. there are simpler ways to play 100-200 hz and up. i for fun and giggles crossed it at 40-45 Hz, distortion will lower a pretty decent amount. (using a very steep filter 48dB) trick is get the resonance where it goes berserk out of the picture and i think it could be rather good :) then again i still need to look for a proper damping method. since a DSP cannot fix everything, besides that it wil only damp the output not the honk associated with it completely, it lowers it since it is hitting it less hard. but its not a proper fix. maybe tomorow i laser some foam pieces and play around with them.
 
I always look forward to your videos :)

I think you may be interpreting my takeaways too harshly. Bass is super hard for flat drivers to accomplish without distorting, so this is all major accomplishments. Plenty of setups have that much distortion and no one complains. I bet the low pass filter helps a lot.
 
No harsh taken :) haha i am not at all sure if these are any good , they produce allot of lows. not yet sure if i like the huge resonance and high resonance fequency... besides that i am not sure i like making 2 foils haha. i might play with it some more, and seee if i can use it for some functions... so far i just have a listen and think about it some more :)
 
I had trouble with the video, may try later. Perhaps try adding some mass (in the form of little dots/stickers/whatever of heavy'ish material) at different spots on each panel (front and back)? Lower and spread the fundamental panel resonance? Magnepan uses asymmetric "dots" (tie points) on a pair of speakers so both panels do not exhibit the same panel modes (frequencies). I've always thought that not unreasonable but seems like there ought to be a better way. Staggering the drive strips with some phase shifting (LC component array) might work, I think you have tried that in the past on 'stats? Similar to what Quad and Sound Lab (IIRC) do?

Cool project... Magnepan's MG20 (and presumably MG30) uses magnets on both sides but all the others have the magnets on just one side so as you said distortion gets very high as excursion increases (e.g. in the bass).
 
well the tie downs will only increase the resonance :( witch is already really high when using 2 foils :( it would also hurt excursion in general. weighing down would lower res at the costs of efficiency. could do that that. but yesterday i removed some magnets in the middle and was greeted by an even higher peak in resonance. so i might be able to get the magnets closer to one and other limit airflow this way and see some changes in the resonance !
 
well the tie downs will only increase the resonance :( witch is already really high when using 2 foils :( it would also hurt excursion in general. weighing down would lower res at the costs of efficiency. could do that that. but yesterday i removed some magnets in the middle and was greeted by an even higher peak in resonance. so i might be able to get the magnets closer to one and other limit airflow this way and see some changes in the resonance !
Ah, I think I misunderstood, high resonance meaning high in frequency or in amplitude (or both)? I was not thinking of tie-downs, just adding mass to the diaphragm itself. My, or one, solution is to cross over to a different panel to avoid exciting the resonance (too much).

I suppose you could seal the panels in a vacuum, or nitrogen like some other designs have done, but that would cause a few other problems... ;)
 
hehe yeah that is a rather cool idea of using denser gas :) bit to much effort i think still cool. yes i meant High in res frequency, and it has a huge peak since its undamped (or half as much then regular by the air its pushing)
 
Back
Top Bottom