Yet another Planar Magnetic Line Source, the SMAPPP

Cartesian or CoreXY refers to the kinetics; the math behind the gantry's movement.
In a bedslinger, a Cartesian machine, the Y movements are the result of bed movements.
In ordinary CNC, also a Cartesian machine, the Y movement are the result of the gantry moving along a second axle.
In a CoreXY machine there's no X or Y stepmotors, they are often named A and B.

MWSnap914 2025-04-11, 08_30_23.jpg

Pros are for example a lighter gantry; no step motors for the Y direction.
Cons for example are that the belts are rather long, that's not suitable for larger machines.

In theory, a bedslinger machine can have CoreXY kinectics for the X direction and the bed slings for the Y direction.
 
Cartesian or CoreXY refers to the kinetics; the math behind the gantry's movement.
In a bedslinger, a Cartesian machine, the Y movements are the result of bed movements.
In ordinary CNC, also a Cartesian machine, the Y movement are the result of the gantry moving along a second axle.
In a CoreXY machine there's no X or Y stepmotors, they are often named A and B.

View attachment 1447208
Pros are for example a lighter gantry; no step motors for the Y direction.
Cons for example are that the belts are rather long, that's not suitable for larger machines.

In theory, a bedslinger machine can have CoreXY kinectics for the X direction and the bed slings for the Y direction.
thank ok the one i had in my head is the core XY then 🙂 they are interesting !

nice got a whole laminating line there ?
 
No, it's for the corregator. The two 40µm aluminium layers upholds the corrugation while baking.
ah well still a form of laminating haha nice !! be it temporarily. great idea by the way using a second layer for baking. i remember this baking step only from the amts and back then it never worked for me but ,i clearly did something wrong !! looking forward to the results
 
The full length membrane measured the same at the short ones but SPL and distortion wise.

Sub to 120Hz @96dB/octave,
Bass 80Hz to 300Hz @96dB/octave,
Mid 300Hz to 3000Hz @ 96dB/octave,
Tweeter from 3000Hz @ 96dB/octave.

I need to raise the gain of the tweeter amplifiers so that I can compensate for the slope above 10kHz.
If I instead EQ all below 10kHz, the overall SPL will be too low.
 
One thing I have forgotten to ask:

When measuring and comparing different membrane configurations like corrugations vs no corrugations, membrane thicknesses and materials and so on, did you test with a full length planar or a shorter one? And if it was shorter, how long was it?

I am thinking about how much going from 1/4 to full length impacts the low end distortion profile. Does it just output 4x the volume @ the same distortion or does it reduce the low end distortion even more?
 
They look ... well insane. (they look super cool) i do wonder about the tiny opening for the tweeter. did you use the same heftly weight for the tweeter (alu foil and backing) ? one last question you use the weird o window strip ? whats on he other side ? on the front plastic pieces. since i usually had problems with uneven compression.. so if you add the front. the membrane might be pushed further to the back (towards the magnets) instead of ending up in the middle. also by the squish it will make the foil go more limp since if something squishes it becomes or want to become wider. so it will change the vertical tension ?
 
Last edited:
another question, you got 2 foams between tweeter and mid, you could use just 1, if they use the same alu foil ? getting them even closer. if they where on the same mylar/kapton. dont get this is a critique they look damn great. im just wondering
 
The tweeter is 3M 74 film with 6µm aluminium foil. The mid is 3M 56 film with 10µm aluminium foil.
I'm not using foam as it has higher distortion that the EPDM P-list and also does naot compress equal when pushed together.
I don't want the two membranes to touch each other and also not influence each other mechanically. Hence separate P-lists in the middle.

I'm using FIR EQ and filters in Roon Convolution.