README.md 8.85 KB
Newer Older
Sam Calisch's avatar
Sam Calisch committed
1
# Plotting membrane speakers
Sam Calisch's avatar
Sam Calisch committed
2

Sam Calisch's avatar
Sam Calisch committed
3
4
## Overview

Sam Calisch's avatar
Sam Calisch committed
5
Project for wire plotting of membrane speakers.  for wildcard week in mas.836 2018
Sam Calisch's avatar
Sam Calisch committed
6

Sam Calisch's avatar
Sam Calisch committed
7
8
9
- Sam Calisch
- Elena Chong
- Zijun Wei
Sam Calisch's avatar
Sam Calisch committed
10
- Victoria Shen
Sam Calisch's avatar
Sam Calisch committed
11

Sam Calisch's avatar
Sam Calisch committed
12
We will use a custom magnet wire plotting tool for the Zund cutter to plot coils for audio speakers.
Sam Calisch's avatar
Sam Calisch committed
13
14
15
16
17
18

<img src="https://calischs.pages.cba.mit.edu/kiri-tww/img/flatcoil/flatcoil-40awg-small.mp4" width=600px>

<img src="https://calischs.pages.cba.mit.edu/kiri-tww/img/flatcoil/flatcoil-40awg-microscope.jpg" height=250px>
<img src="https://gitlab.cba.mit.edu/calischs/kiri-tww/raw/master/img/flatcoil/flatcoil-cut.jpg" height=250px>

Sam Calisch's avatar
Sam Calisch committed
19
20
## TODO

Sam Calisch's avatar
Sam Calisch committed
21
- ~~waterjet steel layers~~
Elena Chong Loo's avatar
Elena Chong Loo committed
22
- ~~laser cut shim layer (500um? 1mm?). (elena)~~
Sam Calisch's avatar
Sam Calisch committed
23
- test heat+vacuum for membrane layers (sam)
Elena Chong Loo's avatar
Elena Chong Loo committed
24
25
- ~~verify magnet orientations with gaussmeter.~~
- ~~modify coil cutout to extend beyond frame for easy stretching, run job (sam)~~
Elena Chong Loo's avatar
Elena Chong Loo committed
26
27
- ~~solder leads (mount RCA jack?) (victoria)~~
- ~~test with audio amp~~
Sam Calisch's avatar
Sam Calisch committed
28
29
30

optional
- do a version with 1/16" magnets and 1/16" phenolic layer
Elena Chong Loo's avatar
Elena Chong Loo committed
31
- do a couple more iteration and polish final product
Sam Calisch's avatar
Sam Calisch committed
32

Sam Calisch's avatar
Sam Calisch committed
33
34
35

## Simulation

Sam Calisch's avatar
Sam Calisch committed
36
37
Two approaches: Coil-magnet, and coil-coil.

Sam Calisch's avatar
Sam Calisch committed
38
39
### Coil-coil

Sam Calisch's avatar
Sam Calisch committed
40
41
The coil-coil topology relies on coils to generate all magnetic fields, not just the time varying one.

Sam Calisch's avatar
Sam Calisch committed
42

Sam Calisch's avatar
Sam Calisch committed
43
44
<a href='comsol/coil-coil-speaker.mph'>comsol file</a>

Sam Calisch's avatar
Sam Calisch committed
45
<img src="img/coil-coil-flux.png" height=200px>
Sam Calisch's avatar
Sam Calisch committed
46

Sam Calisch's avatar
Sam Calisch committed
47
<img src="img/coil-coil-force.png" height=200px>
Sam Calisch's avatar
Sam Calisch committed
48

Sam Calisch's avatar
Sam Calisch committed
49
### Magnet-coil
Sam Calisch's avatar
Sam Calisch committed
50

Sam Calisch's avatar
Sam Calisch committed
51
52
The coil-magnet topology (planar magnetic) is used for some high end headphones.  It can produce significantly more force than coil-coil, even with very thin magnets.

Sam Calisch's avatar
Sam Calisch committed
53
54
<a href='comsol/magnet-coil-speaker.mph'>comsol file</a>

Sam Calisch's avatar
Sam Calisch committed
55
56
<img src="img/magnet-coil-flux.png" height=300px>
<img src="img/magnet-coil-force.png" height=300px>
Sam Calisch's avatar
Sam Calisch committed
57

Sam Calisch's avatar
Sam Calisch committed
58
59
60
61
62
63
64
#### magnet grids

In our meeting this morning, we talked about how rod magnets are more widely available than long, thin bar magnets, and that making grids of rod magnets could make assembly really easy and offer design freedom.  Below are two sketches of square and hex grids with coils routed to capture perpendicular flux.  Magnets are blue circles, all closest pairs have opposite polarity.

<img src="img/square-grid.png" height=400px>
<img src="img/hex-grid.png" height=400px>

Sam Calisch's avatar
Sam Calisch committed
65
The design below has rotational symmetry, which may help flatten its frequency response.
Sam Calisch's avatar
Sam Calisch committed
66

Sam Calisch's avatar
Sam Calisch committed
67
68
<img src="img/hex-spiral-screenshot.png" height=400px>
<img src="img/hex-spiral-coil.jpg" height=400px>
Sam Calisch's avatar
Sam Calisch committed
69

Sam Calisch's avatar
Sam Calisch committed
70
71
72
73
Watching the coils being plotted is fun:

<img src="img/hex-spiral-plotting.mp4" height=400px>

Elena Chong Loo's avatar
Elena Chong Loo committed
74
75
## Design and Fabrication

Elena Chong Loo's avatar
Elena Chong Loo committed
76
77
78
### Long coil for testing

First, we decided to start off with making a rectangular frame for clamping firmly the membrane (wire plotting on 25 micron kapton tape) that Sam initially designed for another project of his.
Elena Chong Loo's avatar
Elena Chong Loo committed
79

Elena Chong Loo's avatar
Elena Chong Loo committed
80
<img src= "img/wireplotting1.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
81
82
83
84
85
86
87
88
89
90
91
92
<img src= "img/wireplotting2.jpg" height=400px>  

The initial frame that we designed contained pockets on each side for placing the "rubber" that will create friction between the frame plates. 

<img src= "img/wireplotting4.jpg" height=400px>  

We used the Epilog Lasercutter to lasercut the frames on 1/8" acrylic sheet.  

<img src= "img/wireplotting3.jpg" height=400px>  

For the pocket, we used 60% speed, 100% power, and 200ppi, for the outline of the frame we used 3% speed, 100% power, and 200ppi.  

Elena Chong Loo's avatar
Elena Chong Loo committed
93
<img src= "img/wireplotting5.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
94
95
96
97
98
99
100

One of the resulting frame looked like the picture below:

<img src= "img/wireplotting6.jpg" height=400px>  

We used Crazy Glue to glue the rubber in place.  

Elena Chong Loo's avatar
Elena Chong Loo committed
101
<img src= "img/wireplotting7.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
102
103
104

For the first attempt, we just placed the membrane between the two plates and used a few M4 screws to clamp it.

Elena Chong Loo's avatar
Elena Chong Loo committed
105
<img src= "img/wireplotting8.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
106
107
108

We noticed that the acrylic is not a good material for exerting pressure. It started bending, which defied its purpose.

Elena Chong Loo's avatar
Elena Chong Loo committed
109
<img src= "img/wireplotting9.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
110
111
112

A bird view of membrane clamped with the long rectangular frame.

Elena Chong Loo's avatar
Elena Chong Loo committed
113
<img src= "img/wireplotting10.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
114

Elena Chong Loo's avatar
Elena Chong Loo committed
115
116
### Magnet-Coil Hexagonal design

Elena Chong Loo's avatar
Elena Chong Loo committed
117
After brainstorming for the first design iteration, we decided on using magnet and coil with Sam's idea of a magnet grid. He succesfully got a few plotted out nicely on 25 micron kapton tape.  
Elena Chong Loo's avatar
Elena Chong Loo committed
118

Elena Chong Loo's avatar
Elena Chong Loo committed
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
<img src= "img/wireplotting11.0.jpg" height=400px>

<!--need to verify if it is 1/8" magnet and wood... was it plywood?-->
We then lasercutted the outer frame from 1/8" wood to clamp the coil membrane. 

<img src= "img/wireplotting11.jpg" height=400px>

We then lasercutted the magnet grid plate from the same kind of wood.

<img src= "img/wireplotting12.jpg" height=400px>

Then, 1/8" magnets were placed in the holes from the magnet-grid plate. Each grid has the opposite polarity.  

<img src= "img/wireplotting13.jpg" height=400px>

<!--Just for fun, we placed a MAGNETIC SHEET?????????????? FORGOT NAME OF THE PAPER AND ITS FUNCTION-->
<!--<img src= "img/wireplotting14.jpg" height=400px>-->

We used a Gaussmeter to check for the magnets polarity. 

<img src= "img/wireplotting17.jpg" height=400px>

Elena Chong Loo's avatar
Elena Chong Loo committed
141
Here is the mapped out magnet-grid. Green for positive and black for negative polarity. We later spray painted this plate with yellow paint to cover the drawing.
Elena Chong Loo's avatar
Elena Chong Loo committed
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187

<img src= "img/wireplotting18.jpg" height=400px>

Sam used the water-jet to cut out a steel plate for the ground plate (the base).

<img src= "img/wireplotting19.jpg" height=400px>

<img src= "img/wireplotting20.jpg" height=400px>

We tried to lasercut some of the materials that Sam provided us. We tried to cut the 1/8" and 1/16" phenolic sheets with the lasercutter using 0.7% speed, 100% power, 400ppi and were not successful, so we decided to just do it on the plastic sheet.

<img src= "img/wireplotting21.jpg" height=400px>

It did not even go through the sheet! Here is a picture of the burned phenolic sheet at different speed level (0.7-1.5%).

<img src= "img/wireplotting22.jpg" height=400px>

Sam also provided us with a very thin plastic sheet.

<img src= "img/wireplotting23.jpg" height=400px>

We tested the cutting settings and found that the best values were: 3% speed, 80-85% power, 400 ppi.

<img src= "img/wireplotting24.jpg" height=400px>

Here are some five shim/spacers we lasercutted from this thin plastic sheet.

<img src= "img/wireplotting25.jpg" height=400px>

During our design discussing, we did not get that far ahead into thinking about the testing process. We forgot to add clearance for an RCA plug which will be used for sending audio signals to the coil. Hence, we had to separate the magnet-grid plate from the steel plate (We snapped them together beforehand.)

<img src= "img/wireplotting26.jpg" height=400px>

We used the drill to create a bigger hole for the RCA plug. Had to do it for all layers: steel plate, magnet-grid plate, spacers/shims, frame plate.

<img src= "img/wireplotting27.jpg" height=400px>

Some fine soldering work came next. Had to peel off the insulation layer from the copper wire for soldering. The copper used for plotting was 40 gauge which makes it difficult to connect to the RCA plug. Thus, we stepped up the copper wire by soldering it to a thicker one.

<img src= "img/wireplotting28.jpg" height=400px>

The resulting planar speaker.

<img src= "img/wireplotting29.jpg" height=400px>

<img src= "img/wireplotting30.jpg" height=400px>
Elena Chong Loo's avatar
Elena Chong Loo committed
188

Elena Chong Loo's avatar
Elena Chong Loo committed
189
## Testing
Sam Calisch's avatar
Sam Calisch committed
190

Elena Chong Loo's avatar
Elena Chong Loo committed
191
192
193
194
It is testing time!

<img src= "img/wireplotting31.jpg" height=400px>

Elena Chong Loo's avatar
Elena Chong Loo committed
195
196
197
198
199
200
201
202
203
204
We tested our first iteration and here are the results:

No spacer/shim. Minimum separation between the coil membrane and the magnets plate. (First Reaction!)

<img src="img/out_noshim.mp4" height=400px>

With two spacers separation between the coil membrane and the magnets plate.

<img src="img/out_twospacers.mp4" height=400px>

Elena Chong Loo's avatar
Elena Chong Loo committed
205
Three spacers separation between the coil membrane and the magnets plate. This video also shows some of the layers of this planar speaker.
Elena Chong Loo's avatar
Elena Chong Loo committed
206
207
208
209
210

<img src="img/out_threespacers.mp4" height=400px>

First iteration was a success!

Elena Chong Loo's avatar
Elena Chong Loo committed
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
## Stereo Planar Speaker

We made some more speakers today.

<img src="img/wireplotting-101.jpg" height=400px>

All the parts of one planar speaker.

<img src="img/week13-02.jpg" height=400px>

Coil membrane of the speaker.

<img src="img/wireplotting-100.jpg" height=400px>

An assembled planar speaker.

<img src="img/week13-01.jpg" height=400px>

Test of the stereo speaker with coil membrane pasted on the plastic. 
<img src="img/output.mp4" height=400px>

If you compare this video with the two spacers one, you will notice that it seems to have better sound with a stiffer coil membrane.

Sam Calisch's avatar
Sam Calisch committed
234
## Links:
Sam Calisch's avatar
Sam Calisch committed
235
236
237
238

- [Planar magnetic headphones](https://www.innerfidelity.com/content/how-planar-magnetic-headphones-work)
- [Ben Katz's Planar magnetic headphones pt. 1](http://build-its-inprogress.blogspot.com/2017/08/planar-magnetic-headphones-part-1.html)
- [Ben Katz's Planar magnetic headphones pt. 2](http://build-its-inprogress.blogspot.com/2017/09/planar-magnetic-headphones-part-2.html)
Sam Calisch's avatar
Sam Calisch committed
239
- [Applied Science magnetics video](https://www.youtube.com/watch?v=4UFKl9fULkA)