|
I've been doing the thing that you're not supposed to do: building my own condenser capsules. I thought I'd share my technique.
I wanted to experiment with the shape of the capsule, membrane tension, and pattern of backplate holes to get a feel for the difference it makes. However my metalwork skills suck, and I only have soldering equipment and a desktop CNC milling machine to hand. I also wanted each capsule to be cheap to make because the intention is to make a variety of shapes / hole patterns to see what effect this has.
So I've been building them from unetched sheets of copper-clad PCB (exploded view):
Cheap space blankets seem to make viable membranes for experimenting, being conductive on one side and about 20?m thick. Scotch tape is a great insulation spacer being only 60?m thick. The generic PCB you can get seems to have a very uniform surface though I've not measured it.
I'm using my desktop CNC milling machine to drill the backplate and faceplate holes and to cut the faceplate aperture. I'm then using a Dremel to polish the backplate and take off any little shards on the cut edges.
To tension the membrane, I've simply been stretching it over a 80mm large metal ring conducting side down, using PVC tape to tack it down so it's visibly flat. Then once I've applied the Scotch tape perimeter to the backplate with a very thin layer of glue, I place the backplate face-down on the membrane and position a weight on top. After about 15 minutes I can cut the membrane free from the ring. Then I position the faceplate, prick through the mounting holes with a needle and push through the nylon bolts and fasten with nylon nuts. I then solder the backplate and faceplate leads. The final step is to fold the Mylar edges over the faceplate and tape them down to make electrical contact.? I found that putting the faceplate copper-side down always caused shorts to the backplate, likely around the mouting holes and it would happen no matter how careful I was.
The capacitance measures 250-800pF depending on capsule size. Some of that is from the clamped section of the membrane.
So far the capsules are quite resonant in the mid frequencies, but disappointingly lacking in the highs, rolling off above 10kHz. The low end response seems to depend strongly on tension, the slacker leading to more prominent lows. The output levels are pretty low, needing about 3 o-clock on the gain to get a reasonable signal. I'm not sure if this is to do with the mass of the membrane or due to the membrane resistance (~3 ohms per 10mm).?
Does anyone know of source of low resistance 6?m aluminiumised Mylar? I'm not keen to go ripping apart electrolytic caps if I can get my hands on factory samples!
|
Awesome! And yes¡. You are ¡°supposed to do that¡± lol! Best Regards,
Jules Ryckebusch?
toggle quoted message
Show quoted text
On Apr 11, 2022, at 16:43, michaeljtbrooks@... wrote:
? I've been doing the thing that you're not supposed to do: building my own condenser capsules. I thought I'd share my technique.
I wanted to experiment with the shape of the capsule, membrane tension, and pattern of backplate holes to get a feel for the difference it makes. However my metalwork skills suck, and I only have soldering equipment and a desktop CNC milling machine to hand. I also wanted each capsule to be cheap to make because the intention is to make a variety of shapes / hole patterns to see what effect this has.
So I've been building them from unetched sheets of copper-clad PCB (exploded view):
Cheap space blankets seem to make viable membranes for experimenting, being conductive on one side and about 20?m thick. Scotch tape is a great insulation spacer being only 60?m thick. The generic PCB you can get seems to have a very uniform surface though I've not measured it.
I'm using my desktop CNC milling machine to drill the backplate and faceplate holes and to cut the faceplate aperture. I'm then using a Dremel to polish the backplate and take off any little shards on the cut edges.
To tension the membrane, I've simply been stretching it over a 80mm large metal ring conducting side down, using PVC tape to tack it down so it's visibly flat. Then once I've applied the Scotch tape perimeter to the backplate with a very thin layer of glue, I place the backplate face-down on the membrane and position a weight on top. After about 15 minutes I can cut the membrane free from the ring. Then I position the faceplate, prick through the mounting holes with a needle and push through the nylon bolts and fasten with nylon nuts. I then solder the backplate and faceplate leads. The final step is to fold the Mylar edges over the faceplate and tape them down to make electrical contact.? I found that putting the faceplate copper-side down always caused shorts to the backplate, likely around the mouting holes and it would happen no matter how careful I was.
The capacitance measures 250-800pF depending on capsule size. Some of that is from the clamped section of the membrane.
So far the capsules are quite resonant in the mid frequencies, but disappointingly lacking in the highs, rolling off above 10kHz. The low end response seems to depend strongly on tension, the slacker leading to more prominent lows. The output levels are pretty low, needing about 3 o-clock on the gain to get a reasonable signal. I'm not sure if this is to do with the mass of the membrane or due to the membrane resistance (~3 ohms per 10mm).?
Does anyone know of source of low resistance 6?m aluminiumised Mylar? I'm not keen to go ripping apart electrolytic caps if I can get my hands on factory samples!
|
Really amazing! Brilliant! All encouragement
Best wishes
David P
I've been doing the thing that you're not supposed to do: building my own condenser capsules. I thought I'd share my technique.
I wanted to experiment with the shape of the capsule, membrane tension, and pattern of backplate holes to get a feel for the difference it makes. However my metalwork skills suck, and I only have soldering equipment and a desktop CNC milling machine to hand. I also wanted each capsule to be cheap to make because the intention is to make a variety of shapes / hole patterns to see what effect this has.
So I've been building them from unetched sheets of copper-clad PCB (exploded view):
Cheap space blankets seem to make viable membranes for experimenting, being conductive on one side and about 20?m thick. Scotch tape is a great insulation spacer being only 60?m thick. The generic PCB you can get seems to have a very uniform surface though I've not measured it.
I'm using my desktop CNC milling machine to drill the backplate and faceplate holes and to cut the faceplate aperture. I'm then using a Dremel to polish the backplate and take off any little shards on the cut edges.
To tension the membrane, I've simply been stretching it over a 80mm large metal ring conducting side down, using PVC tape to tack it down so it's visibly flat. Then once I've applied the Scotch tape perimeter to the backplate with a very thin layer of glue, I place the backplate face-down on the membrane and position a weight on top. After about 15 minutes I can cut the membrane free from the ring. Then I position the faceplate, prick through the mounting holes with a needle and push through the nylon bolts and fasten with nylon nuts. I then solder the backplate and faceplate leads. The final step is to fold the Mylar edges over the faceplate and tape them down to make electrical contact.? I found that putting the faceplate copper-side down always caused shorts to the backplate, likely around the mouting holes and it would happen no matter how careful I was.
The capacitance measures 250-800pF depending on capsule size. Some of that is from the clamped section of the membrane.
So far the capsules are quite resonant in the mid frequencies, but disappointingly lacking in the highs, rolling off above 10kHz. The low end response seems to depend strongly on tension, the slacker leading to more prominent lows. The output levels are pretty low, needing about 3 o-clock on the gain to get a reasonable signal. I'm not sure if this is to do with the mass of the membrane or due to the membrane resistance (~3 ohms per 10mm).?
Does anyone know of source of low resistance 6?m aluminiumised Mylar? I'm not keen to go ripping apart electrolytic caps if I can get my hands on factory samples!
-- - - - - - - - - - - - - - - - - - - - - - - - -
David Pinnegar, B.Sc., A.R.C.S.
- - - - - - - - - - - - - - - - - - - - - - - -
+44 1342 850594
|
Le 11/04/2022 ¨¤ 23:43, michaeljtbrooks@... a ¨¦crit?:
The output levels are pretty low, needing about 3 o-clock on the gain to get a reasonable signal.
What impedance converter (head amp) are you using? What polarization voltage (bias) are you using? What preamp?
|
Interesting project.
I made a basic omni circular capsule many years ago (I had access to a friends lathe and pretty well equipped workshop at the time).? Diaphragm material was an issue (especially as I had very limited income), but I did get it working.? It's certainly possible.? Good luck.
Every Blessing
Tony
toggle quoted message
Show quoted text
On 12/04/2022 08:01, Jerry Lee Marcel wrote:
Le 11/04/2022 ¨¤ 23:43, michaeljtbrooks@... a ¨¦crit?:
The output levels are pretty low, needing about 3 o-clock on the gain to get a reasonable signal.
What impedance converter (head amp) are you using?
What polarization voltage (bias) are you using?
What preamp?
--
This email has been checked for viruses by AVG.
|
What impedance converter (head amp) are you using?
What polarization voltage (bias) are you using?
What preamp?
It is more of a relative observation. A Chinese K87-style capsule with exactly the same circuit gives a hotter output, requiring 11 o-clock on the gain.
Headphone amplifier: Behringer DX2000 mixer
Polarization voltage: 72V from a Jules' DC-DC Hex multiplier
Preamp: OPA Alice (OPA1642) phantom powered.
I'm pretty suspicious of the membrane resistance which is a few ohms per cm2 as the culprit. So any suggestions for more conductive aluminiumised Mylar would be greatly appreciated!
|
I'm pretty suspicious of the membrane resistance which is a few
ohms per cm2 as the culprit. So any suggestions for more
conductive aluminiumised Mylar would be greatly appreciated!
Whatever few ohms there don't matter compared to the Gigaohm input
resistance of the Alice circuit.
Look elsewhere. Surely 20 micron thickness does not help with
sensitivity, but all the parasitic capacitances are probably the
main culprit.
|
Hey, Jerry Lee, my namens Ralf and this is the first reply ?in this forum. I¡® m writing from Germany so please ? ? excuse my bad language. About 10 years ago I made a very similar experiment wich I stopped as I got me a lath to work with. I found out the same things that you posted here and came to the following conclusions: 1- It is important to get the membrahn streched the same degree to all sides? 2- the backplate cannot be stiff enough. That is one reason for partial cancelation of higher frequencies. This is because if it is not it will probably cancel out frequencies by slightly moving by soundpressure in the opposit direction of the membrahne. 3- the backplate had to be fairly thick so that you can make some sackholes in it for damping ?reasons? 4- to get rid of most of the parasitic capacitance, you simply should take of the copper of your backplate piece in the airia where the frame that holds your membrahn will be positioned?
I hope that this will bring you in a good direction.
Good luck?
I'm pretty suspicious of the membrane resistance which is a few
ohms per cm2 as the culprit. So any suggestions for more
conductive aluminiumised Mylar would be greatly appreciated!
Whatever few ohms there don't matter compared to the Gigaohm input
resistance of the Alice circuit.
Look elsewhere. Surely 20 micron thickness does not help with
sensitivity, but all the parasitic capacitances are probably the
main culprit.
|
Hi Ralf,
It's not me that's building a capsule, it's michaeljtbrooks.
However, I can make some comments:
"2- the backplate cannot be stiff enough. That is one reason
for partial cancelation of higher frequencies. This is because
if it is not it will probably cancel out frequencies by slightly
moving by soundpressure in the opposit direction of the
membrahne."
I'm not convinced. The backplate would need to be very flexible to
react in such a way. Most of the backplates are made of a piece of
metal (often brass) that is millions of times heavier and more
rigid than the diaphragm.
"3- the backplate had to be fairly thick so that you can make
some sackholes in it for damping ?reasons "
Indeed, blind holes (as well as through-holes) are necessary for
providing damping of the diaphragm
Le 12/04/2022 ¨¤ 12:38, Ralf Falk a
¨¦crit?:
toggle quoted message
Show quoted text
Hey, Jerry Lee,
my namens Ralf and this is the first reply ?in
this forum. I¡® m writing from Germany so please ? ? excuse my
bad language.
About 10 years ago I made a very similar
experiment wich I stopped as I got me a lath to work with. I
found out the same things that you posted here and came to the
following conclusions:
1- It is important to get the membrahn streched
the same degree to all sides?
2- the backplate cannot be stiff enough. That is
one reason for partial cancelation of higher frequencies. This
is because if it is not it will probably cancel out frequencies
by slightly moving by soundpressure in the opposit direction of
the membrahne.
3- the backplate had to be fairly thick so that
you can make some sackholes in it for damping ?reasons?
4- to get rid of most of the parasitic
capacitance, you simply should take of the copper of your
backplate piece in the airia where the frame that holds your
membrahn will be positioned?
I hope that this will bring you in a good
direction.
Good luck?
I'm pretty suspicious of the
membrane resistance which is a few ohms per cm2 as the
culprit. So any suggestions for more conductive
aluminiumised Mylar would be greatly appreciated!
Whatever few ohms there don't matter compared to the
Gigaohm input resistance of the Alice circuit.
Look elsewhere. Surely 20 micron thickness does not help
with sensitivity, but all the parasitic capacitances are
probably the main culprit.
|
Hallo,? sorry for my mistake.? But, please let me say something to your comment. Just to clarify, you¡®re saying that Most backplates are made out of solid brass. You are completly right and this was exactly what I just wanted to say - they should be stiff. There is a different type of capsules which have a second membrahne of course. No matter of this membrahne is passive or used for pattern ?switching , this one works ?in a 180 degree? Greetings and thanks Ralf
Hi Ralf,
It's not me that's building a capsule, it's michaeljtbrooks.
However, I can make some comments:
"2- the backplate cannot be stiff enough. That is one reason
for partial cancelation of higher frequencies. This is because
if it is not it will probably cancel out frequencies by slightly
moving by soundpressure in the opposit direction of the
membrahne."
I'm not convinced. The backplate would need to be very flexible to
react in such a way. Most of the backplates are made of a piece of
metal (often brass) that is millions of times heavier and more
rigid than the diaphragm.
"3- the backplate had to be fairly thick so that you can make
some sackholes in it for damping ?reasons "
Indeed, blind holes (as well as through-holes) are necessary for
providing damping of the diaphragm
Le 12/04/2022 ¨¤ 12:38, Ralf Falk a
¨¦crit?:
Hey, Jerry Lee,
my namens Ralf and this is the first reply ?in
this forum. I¡® m writing from Germany so please ? ? excuse my
bad language.
About 10 years ago I made a very similar
experiment wich I stopped as I got me a lath to work with. I
found out the same things that you posted here and came to the
following conclusions:
1- It is important to get the membrahn streched
the same degree to all sides?
2- the backplate cannot be stiff enough. That is
one reason for partial cancelation of higher frequencies. This
is because if it is not it will probably cancel out frequencies
by slightly moving by soundpressure in the opposit direction of
the membrahne.
3- the backplate had to be fairly thick so that
you can make some sackholes in it for damping ?reasons?
4- to get rid of most of the parasitic
capacitance, you simply should take of the copper of your
backplate piece in the airia where the frame that holds your
membrahn will be positioned?
I hope that this will bring you in a good
direction.
Good luck?
I'm pretty suspicious of the
membrane resistance which is a few ohms per cm2 as the
culprit. So any suggestions for more conductive
aluminiumised Mylar would be greatly appreciated!
Whatever few ohms there don't matter compared to the
Gigaohm input resistance of the Alice circuit.
Look elsewhere. Surely 20 micron thickness does not help
with sensitivity, but all the parasitic capacitances are
probably the main culprit.
|
I have reskinned damaged capsules with aluminized mylar unrolled from film capacitors. The lower the voltage rating, the thinner the film is.?The capsules have sounded fine, though likely not good as new.
I tried making capsules with perforated aluminum sheet as backplate, Scotch tape, and mylar film. They did pick up sound: poorly, not as well as a cheap electret. Like you, I wonder how shape and backplate geometry affects tone. Your method is better because you can isolate the active area of the backplate by etching, and reduce the parasitic fixed capacitance. Good work!
The resistance of the diaphragm doesn't matter much, especially if signal is taken off the backplate. All it has to do is hold a constant charge.
|
Though not Mylar (and not as tough), you maybe could try ultra thin aluminum like the ribbon mic folk use:?
could spray coat one side with a nonconductive lacquer if need be.
Or, if you want to go cheaper than that, you could get some ¡°imitation silver leaf¡±. ?But that is extremely fragile and frustrating to work with, so..?
|
Here's the results of a triangle shaped capsule, inspired by Ehrlund.?
I've shaved off a lot of the parasitic capacitance and the output is definitely higher. It's still sounding pretty narrowband mid-heavy. This one was done with about 150g of weight for membrane tension and the usual Scotch tape as a spacer.
I'm going to try increasing the backplate thickness with a second layer of PCB or maybe some acrylic to see if it gives better highs.
|
Dude, that is amazing! Excellent?work.
Here's the results of a triangle shaped capsule, inspired by Ehrlund.?
I've shaved off a lot of the parasitic capacitance and the output is definitely higher. It's still sounding pretty narrowband mid-heavy. This one was done with about 150g of weight for membrane tension and the usual Scotch tape as a spacer.
I'm going to try increasing the backplate thickness with a second layer of PCB or maybe some acrylic to see if it gives better highs.
-- Best Regards,
Jules Ryckebusch
214 399 0931
|
I¡¯m impressed! I look forward to your news of your progress!
Here's the results of a triangle shaped capsule, inspired by Ehrlund.?
I've shaved off a lot of the parasitic capacitance and the output is definitely higher. It's still sounding pretty narrowband mid-heavy. This one was done with about 150g of weight for membrane tension and the usual Scotch tape as a spacer.
I'm going to try increasing the backplate thickness with a second layer of PCB or maybe some acrylic to see if it gives better highs.
-- Phil Manning ~ classic recording studio
~piano tuning & serviceAjax studio: 905.231.3189
Pickering shop/office: 416.907.3562
cell: 905.626.3727
|
Taking on board what has been said about the backplate mass, I've built another capsule with a double PCB backplate. I also played around with another shape: a bowed-in square.?
Thanks to having two PCB layers I was able to add some blind-ended attenuation holes too.
I did cock up the membrane a little bit - a couple of tiny spurs from the drilled holes are tenting the membrane a little. It is much better in the high frequency end, a bit lacking in the bass (I think due to the tenting), and lower signal output, but a lot less resonant. I've used the same 20?m space blanket as the diaphragm.
Here's the design, some photos and how it sounds.
|
Brilliant!! That sounds much better than any of my home made capsules.
|
I have small confession to make...
The space blanket I've been using is not 20?m thick as originally assumed. It is actually just 6?m thick! I got hold of a proper micrometer today and was able to measure it far more accurately.?
So, single-side metalised 6?m Mylar is available in sheets of 100cm x 150cm, (folded into 10cm x ~7cm rectangles) for a few pounds each. If you don't mind edge-terminated I can't imagine a cheaper way to make condenser capsules!
|
On Sat, Apr 16, 2022 at 08:11 PM, <michaeljtbrooks@...> wrote:
So, single-side metalised 6?m Mylar is available in sheets of 100cm x 150cm, (folded into 10cm x ~7cm rectangles) for a few pounds each. If you don't mind edge-terminated I can't imagine a cheaper way to make condenser capsules!
where do you get this sheet? I bet not all space blankets are the same and this thin.
|
alloy foil from old wax capacitors as diaphragm. have used in home made ribbon mics.
toggle quoted message
Show quoted text
Though not Mylar (and not as tough), you maybe could try ultra thin aluminum like the ribbon mic folk use:?
could spray coat one side with a nonconductive lacquer if need be.
Or, if you want to go cheaper than that, you could get some ¡°imitation silver leaf¡±. ?But that is extremely fragile and frustrating to work with, so..?
|
Phil Manning
Mic Scharf