Building capsules

[burdij]

The microprobe was an ePay buy. The best deals are the local pickup ones. Shipping on this would have been 10X more than what I paid for it ($50). Most of the other stuff came off ePay, too. I drove to central Wisconsin to pickup the mechanical pump from a retiring HVAC installer. He even put in new vacuum oil and painted it for me. The only parts I couldn't seem to get for a reasonable price were the feedthrus. Real ones are about $300 each. These are made from 3/4in. brass rod, a ring of delrin and sealed with two buna "O" rings. They seem to work well. I can get down to about .6 x 10(-5) torr. You only need between 10(-3) and 10(-4) so there is some room to spare.

 

[dale116dot7]

Good work!!!

here's a link to my chamber:
http://www.10000cows.com/deposition_system.htm

I just ordered a used vacuum gate valve to put between the diff pump and the chamber so I can do diaphragms without cooling the pump down for every cycle.

I have a few leaks so I get down to 10(-3) Torr but that's enough for this small chamber. I don't really know if I get further down than this - I just have a thermocouple gauge that's good to about .5*10(-3). I really hit the heater hard and deform the Mylar a bit. I've found that if I deposit at a rate so the mylar just starts to crinkle a bit in the chamber, it sticks pretty well without doing a precoat. Some manufacturers put on a nickel or aluminum precoat then top with gold using a second boat for slower coating speeds. Where did you get your mylar from? I picked mine up from www.2spi.com. I have a roll of 6u and 3u. I haven't used the 3u stuff yet - I'd like to try it someday. I made my vacuum feedthroughs 'on the cheap' but I think they are a bit leaky. I used epoxy and brass bar stock. I might just re-do them like you have. They look really nice and I can machine that pretty easily, I think.

And I just made a nasal capsule a few weeks ago. If I block the holes, it didn't actually help. I think I got the diaphragm to backplate clearance a bit too high.

Looks like you've done your homework! At this level of commitment, you'll have no trouble getting a capsule working and sounding good, I'm sure.

 

[ioaudio]

do you think this one is interesting?

http://cgi.ebay.at/Hochvakuum-Aggregat-Edwards-Isc-900_W0QQitemZ7583646536QQihZ017QQcategoryZ36788QQrdZ1QQssPageNameZWD1VQQcmdZViewItem

its called Edwards Isc 900 and the seller knows nothing about it (other than its a high vakuum pump)

 

[Gus]

cool site
http://www.belljar.net/

 

[burdij]

I ordered mylar from SPI also. They were out of stock though. They promised to ship by today. Here is a drawing of the construction of the feedthru. I had to keep it really simple as I have limited tooling for my lathe. The big problem is cutting the larger O-ring groove.

io- I looked at the pump you referred to . It looks very big. To back a diffusion pump, you don't need much capacity. The one I have is a Welch 1400 that has a capacity of 25 Liters/Minute. I bought a lot of vacuum pumps from various sources before I got a good one, that is, one that doesn't need to be rebuilt immediately. I also wasted a lot of money shipping them. You also want a two stage pump because the foreline pressure (the outlet side of the diffusion pump) needs to get down to 1 x 10(-2) torr to work efficiently with the diffusion pump. You should check out your local university surplus stores. I am also near a university accelerator laboratory. You might want to find out if CERN has a surplus outlet.

When buying surplus vacuum parts, I am very careful to clean them thoroughly with acetone followed by alcohol, wearing gloves and a charcoal filter mask and dispose of all the cleaning supplies wrapped in multiple plastic bags. You never know where this stuff has been and what it was used for.

 

[Tim Campbell]

Burdij,
What kind of tuning jig do you plan on using?

I've been asked if I'd consider selling any of my film. I'm afraid that it was such a hard battle finding and aquiring gold metalized 6micron film that I'd like to keep it for a production run of capsules that I'm planning on making soon.

I could however get my hands on some extremely high quality aluminum metalized film (the same exact film I have but with aluminum) at a decent price if anyone's interested. It would probably take about a month in that it would be made for me to my specs. Is anyone interested?

 

[burdij]

I hadn't thought much about tuning but I was intrigued by the setup shown on the Service page of this site. The capsule appears to be driven either directly by a signal generator or by a transducer in the box below it.

http://www.thiersch-mic.de/

 

[SSLtech]

Have Tim show you his tensioning jig pictures...

Then you'll know why he's 'an icon at my shrine'!

:wink:

Keith

 

[Tim Campbell]

Keith means this one:

It's been revised a few times since then.

Dale has his own that based on Neumann's early type. It uses weight to set the frequency.

 

[ioaudio]

tim, i saw your post and just had to ask - sorry, should have been clear to me.

i found this setup for tuning by accident:
http://www.htl.moedling.at/htl1/htl/Innotivanhang2.htm

 

[Tim Campbell]

That's a fantastic link. Thanks!!!

 

[burdij]

Yes, io, thanks for the link. I am curious if someone might know if that frequency (1500Hz) is a natural resonance and the object is to maximize the amplitude, or is the object to vary the resonance frequency and set it to that frequency. It does say "When the maximum of the magnitude is reached, the attunement process is finished" which I take to mean that the frequency is a natural resonance of the system. I re-read the Bore & Peus article, but the information on diaphragm resonance is sketchy.

 

[gyraf]

The other way around. You choose a frequency to tune to - say 1200Hz - get the speaker to sound at this frequency, measure with a mic, and then strech the material untill maximum amplitude is obtained. Now you're tuned to say 1200Hz.

The frequency the membrane is tuned to tells us something about it's mechanical properties - not necessarily that that frequency is part of any real-life equation.

Jakob E.

 

[burdij]

Some progress on the capsule build.

The capsule has a diaphragm and is mounted in the test mike. Here are pictures of the front and rear.

The first audio sounded a bit like a pre-cellphone telephone. Plugged up about half the holes, as predicted by Dale. Sounds better but the output level dropped by 10-15db. The capsule capacitance measures about 45pf. That seemed low so I checked the thickness of the capsule to diaphragm spacer material. That measured three thousandths. Need to reduce it to 1 thousandth. Made a new diaphragm and rebuilt the capsule. This produce a major improvement in the output level and the high end is clear but maybe a little lacking compared with a TLM103. The capsule capacitance is now 115pf. Removed some of the hole plugs. This improves the highs but now the level has fallen off again. The back plate needs to be remade with a flatter back so that I can try a delay/phasing plate like the TLM.

 

[rodabod]

Excellent work.

I notice some SMD on your mic PCBs too. Brave man!

 

[dale116dot7]

To raise the high frequency output, you can try increasing the blind hole size by one drill bit size (I went from #58 to #57 and the high end comes up a bit). Also, what is the depth of the blind holes? If you make them a bit deeper, that can increase the HF response as well. I find 4mm a good depth for any of my capsules. One of my first capsules I tried had a 2.5mm blind hole and that didn't work at all - it sounded like a telephone. It seems as though one wrong move with a capsule and you get a telephone frequency response.

Yes, I use weight to tension my capsules, but I also will check the capacitance with and without polarization to check for tension. The weight works well if your initial tension is accurate. Mine isn't so I add the capacitance check. I used to check for diaphragm suck-in voltage except that if there are any fingerprints on the backplate, you can have a problem with the diaphragm sticking to the backplate and not wanting to release due to the finger oil. I quit doing that now, although I now added a few cleaning steps to the capsule making process in addition to not doing the diaphragm collapse test. Neumann glass-beads their newer capsules (K67/K103 - maybe more) to avoid this problem.

Most of this is from observation and experimentation, so I don't know how much is 'fact' and how much is just coincidence, but....

Highly damped capsules like the M7 and C37 seem to be less fussy as to diaphragm tension. The K67 and CK12 capsules are quite fussy from what I've seen. In a damped capsule, mostly you set the diaphragm to be at a decent tension, but the exact frequency is not quite as critical. There is a bit of a peak there so you can 'voice' the mic differently. But I find it subtle, there seems to be far more effect on a slightly inaccurate hole drilling.

The K67 requires identical tuning for both sides, or the sound isn't right in any pattern, except possibly cardioid on one side. Cardioid on the other side won't sound right though. It usually takes me two or three tries on each side to get a K87 to sound right.

The CK12 capsules have the diaphragm resonance as part of the acoustic circuit so they need to be tuned not only to a specific tension, but a specific resonant frequency. I've only tried re-skinning a nylon CK12. I got it to sound about right but it's hard to deal with. I also tried making a CK12 clone. I might try it again some day, but I've been quite busy making M7-like capsules. I think Tim has quite a bit of experience with the chambered capsules and I don't really want to duplicate all of his efforts.

That is a very impressive amount of progress you've made in relatively little time!

-Dale

 

[burdij]

Thanks, Dale, for the suggestions. I am not sure I am at the "tweaking the sound in" stage yet. I am still looking for the street the ballpark is located on. I have noticed that the tension has to be fairly accurate and repeatable. To that end I have made a tensioning jig similar in concept with the one in the R. Williamson article and the one Tim Campbell built. In experimenting with it, I have been able to fairly easily set the resonance by setting the jig on a speaker connected to an audio oscillator.

 

[Tim Campbell]

Try soldering all of the holes in the back of your capsule closed. Then try opening them up one at a time until you get a cardiod pattern. This is a pressure gradient capsule. You want to have half the sound pressure of a figure 8 design reaching the back of the membrane.

Try tuning your membrane to 1000hz.

 

[Hayman]

What tools do you guys use for machining the parts on your capsules?

Richard

 

[burdij]

I use a Harbor Freight 9x20 cheapie lathe, a Harbor Freight mini-mill that I have recently converted to CNC and a 20 year old Ryobi drill press. The big investment has been in tooling. I was machining the acrylic for the capsule bodies on the lathe but since I have made the acrylic rings for the jig on the mill, I think I am going to try running some capsule bodies on that. Then I don't have to do everything manually. Here is a picture of the mill, post conversion. It looks really clean here. Of course, it doesn't look like that anymore: