DIY capsule

[saxtim]

I finally retested my capsule today.

Good news is it works, bad news is that it's not all it could be. Output is low, and the there's a bit of noise. I think the noise is comming from the capsule, not the circuit. It's below signal level, but well and truely loud enough to be unacceptable. The noise is a combination of hiss, and the sound you get when you put a microphone in a windy area - like if you had a video camera and you recording somewhere really windy. Sound wise, the signal itself doesn't sound too flash hot, missing a lot of highs and some lows. Sounds like mainly mids are coming through well. Even though it's a dual diaphragmed capsule, I wired it up as a cardoid, but (without proper testing) I'm pretty sure it's not a true cardoid pattern I'm getting.

Having said this now that I've done one I'm going to reattempt a capsule from scratch knowing I can get the thing to function. I'd mentioned I was treating this as a bit of a test to see if I could get it to work. Now that I know it can I'm hoping with some more careful working the following areas, it will sound and perform much better:

1. I'm getting some machined blanks from a machine shop - just the backplate turned down and cut to the right thickness/diameter. This one I used a hack saw on a brass rod and sanded - not convinced I got it entirely to the correct dimensions.

2. Going to get some mylar of known thickness for backplate spacers - at the moment I used some material from a A4 plastic display pocket. I calculated it to be between 40-50 microns. I need to figure out a better way for cutting out the holes for the screw to pass through. Using a drill I can't get clean holes, the material just tends to tear and not leave a properly formed hole.

3. Mechanically, the hardest thing I found to do was getting the screw holes for the retaining rings acurately spaced and drilled. I was using paper templates (printed out from computer) to drill the screw holes on the retaining rings and on the backplate, because I did them separetly not all of them lined up that well. I think what I'll do next time is make myself a metal drilling template. I'll get a piece of aluminium and drill the screw holes in that. Then I can place that over the backplate or retaining rings and use it to guide the drill. Hopefully this way I can get everything to line up correctly.

4. I think I need to rething/remake my tuning jig into something more permanent and easy to use. I'm still not convinced I accurately tuned the diaphragms either - I'll spend more time on that next time.

Anyway, over the next month or so I'm going to try again and hopefully come up with something that works closer to expected.

Tim

 

[adrianh]

A simple milling machine with a readout will give you the location.
Mill the holes out to the correct dia. then tap them by hand. Or use a tapmatic on a drill press.
With a readout move the work to the correct location by looking at the X,Y on the readout.

Every serious milling machine I have seen has a readout.
It is like a fancy pair of calipers on the machine that gives you
X and Y locations.

 

[saxtim]

I switched the capsule so I'm now taking the signal off the backplate, not the front diaphragm. I also sealed the capsule with tape.

Output went up and noise went down, though output is still quite low. I have gotten rid of that 'wind' type noise though. Remaining noise may be electronic noise, not capsule noise? I have to turn my preamp up a fare way to get a useable signal out of the capsule, so naturally noise goes up as well. Frequency response is better, but still nowhere near perfect, or really even useable at the moment.

The capsule almost responds as an omni. I'm not sure how that is possible, I can only conclude that a lot of the sound from the back reaches the front diaphragm. I'm now taking the signal off the backplate as I mentioned above. I have two diaphragms (both metallised) but only the one diaphragm is connected (as well as the the backplate) As I mentioned earlier I'm using the Royer tube circuit for a head amp.

Is there a simple way I can plot a frequency response chart? I was thinking just setting up a speaker and recording a sine wave with the mic. If I cycle through from 20-20KHz and then look at the resulting recorded level this should give me an idea? This will also plot the response of the electronics as well, but I'm not look for a perfect test, just an idea of what the response is like relative to how the capsule sounds at the moment.

I wouldn't mind trying something similiar for polar pattern too - I know this wouldn't be anything close to a scientific test, but if I took a 1Khz sine wave through a speaker and then just moved the speaker 360 around the microphone I might get some idea of how it performs.

Tim

 

[saxtim]

In terms of pics - I hope to get a digital camera by the end of the month. I'll take some pics when I get one

tim

 

[gyraf]

[quote author="saxtim"] Is there a simple way I can plot a frequency response chart? I was thinking just setting up a speaker and recording a sine wave with the mic. If I cycle through from 20-20KHz and then look at the resulting recorded level this should give me an idea? This will also plot the response of the electronics as well, but I'm not look for a perfect test, just an idea of what the response is like relative to how the capsule sounds at the moment. [/quote]

The easiest and most precise would be to compare to a known-flat microphone that is placed close to the capsule under test. Maybe a Panasonic electrect?

This will rule out frequency response error from the amp/speaker/room part.

For directivity (pattern) test, you should perform the test outdoor - to reduce room reflection errors (if you don't have a damped test chamber)

For testing software, check out "AudioTester":
http://www.sumuller.de/audiotester/maine.htm

Jakob E.