R. Williamson capsule

Anyone build the R. Williamson capsule design? "A Professional Condenser Microphone"

I have the brass piece and plastic outside gluing up with some JB weld now. I am making it a little different (building it with what I have) than the article. I want to use a washer to set the skin to back plate distance and maybe change the blind holes to adjust for a closer skin to backplate distance.

I noted it kind of looks a little like the Sony C37/c800(not g) capsule.

Anyone measure the c37 capsule?

I am not expecting much this is my first attempt at a capsule build so I started on something easy. I used my Seig lathe and my small taig lathe I think I am going to machine the Taig Al jaws for holding capsules for the final cut before lapping.

Tim or Dale Any guess at the tuning of the skin? I am thinking about 1,200hz because of no center term but it seems to be more like a neumann style so maybe 1,200 is to high?

Lower works better. Don't expect any miracles.

where is this article? im interested

Sweet! Good luck, Gus!

What are you using for diaphragm material? Au or Al coated? How thick? I've wanted to try my hand at making a capsule for a long time, but the Au evaporation process scares the crap out of me and my wallet...

Peace,
Al.

al,
have you seen dale1176's gold depositing thread? sick stuff

[quote author="asm"]have you seen dale1176's gold depositing thread? sick stuff[/quote]

Oh yes... That's how i know there's no way I can do it! :green:

Peace,
Al.

I can run some diaphragms off for people, but it's when I get time to run some - I usually do a batch every month or two. Doing up 2 to 6 skins for a few people is no big problem. Getting the system leak-free is very frustrating so unless you are going to do tons of diaphragms, the best bet is to get someone with one of these setups to do it.

The Debenham and Williamson capsules can just use 6u film from one of the newer large film capacitors as you don't need to use the film as an insulator around the edge like on a Neumann mic.

I have been setting my diaphragms at about 800 or 900 Hz. If it's a clamped diaphragm, I usually tension it with about 200 grams of weight so it's easy to punch the holes around the ring and put the screws in, then I release the tension to about 75 grams or about 850 Hz. A glued diaphragm needs to be at final tension before the glue is applied, so at lower frequencies it is tricky to not lift the diaphragm at all with the syringe.

Oh yea, I didn't realize how close my handle was to a famous compressor till now. The 116dot7 came from where I live. Calgary's airport VOR transmitter frequency is at 116.7 MHz. A neighbouring city (Edmonton) has its VOR transmitter at 117.6 MHz.

thats cool dale.

what material are you sputtering mainly? is one harder to do than the others? have you figured out how thick micron wise its being put on?

how do they sound?

LD vs SD, which one is harder to make and why?


:green:

full of questions.
taylor

Gold. I started with nickel to try out the process but now I just do gold. Technically it's not a sputtering process. Sputtering uses a high voltage to move the material around in the chamber. Thermal evaporation (which is what this is) uses a heater.

The thickness is between about 40 and 100 angstroms. I try to judge it by the opacity of the material.

http://www.10000cows.com/AudioFiles.htm
This is my M7-like mic but through a solid-state amp head.

I would love to see some of these articles as well. Anyone have them archived somewhere?

Thank you in advance...

from the meta http://www.conditionedresponse.com/DIY/microphones/index.html

D'oh! The Meta, of course.

Thanks, Gus!

Dan and Tim I don't expect much from this build. If I get any sound I will be happy.

I was thinking this simple capsule build might be a good place to work on the machining skills needed.

Ok great thread.

Since I scanned two of those article I have 2 cents to add.

Thanks to Mr. Josephson of microphone fame who suggested I find the articles.
I talked to him later and he said both those designs do not have enough holes in the back plate. The Shulin is so bad it has a resosnse peak.
So when you machine one do several and play with more holes in the back
plate and report back here.

dale :sound realy good!

> Any guess at the tuning of the skin? I am thinking about 1,200hz because of no center term but it seems to be more like a neumann style so maybe 1,200 is to high?

??? It is nothing like a Neumann. Until the end, it is a basic omni condenser. A WE640A with different tuning and damping.

Such mikes must be tuned to the "top" of the band. WE640 was tuned to 6KHz because that was good enough for the 1930s, and because they wanted stable and repeatable rather than naturally flat, and they didn't have mylar.

The dual-diaphragm Neumann works very different. It is a dipole and basically velocity sensing. Therefore for flat response (with electrostatic sensing) it needs to be resistance controlled. Resonance is not really an issue because the damping dominates. You set the resonance around the middle of the band so the stiffness and mass equal the resistance at the two far extremes of the band. 630Hz for a first guess, though treble issues might favor a higher tuning.

> because of no center term

Entirely superficial. Since no music-recording mikes (except the original PZMs, and the quartz Altec) used diaphragm plate-stiffness, and the diaphragm is perfectly conducting (compared to 100Megs), the presense or lack of a center post does nothing to the sound.

Back to the Williamson.

Williamson likes to be a mavrick, and isn't always perfectly right, but he is never wrong enough to matter. His instructions are quite clear: "Stretching should be only just enough to remove the natural wrinkles in the plastic..."

Now, that probably puts the naked diaphragm at a few-hundred Hz, far too low. So obviously he is relying on trapped air to give most of the spring in the system. Which is certainly a good plan: Mylar isn't very consistent and even has some hysterisis. Air is a better spring. And when I push my doughnut crumbs around, I come up with a number like 16KHz for the resonance on that air-space. Neglecting diaphragm stiffness, which we obviously can do if the diaphragm is 200-800Hz and the total system is 10KHz-20KHz (about 1,000 times stiffer!).

> I want to use a washer to set the skin to back plate distance and maybe change the blind holes to adjust for a closer skin to backplate distance.

The backplate distance will only screw things up. As published, the backplate distance is only about 1/8th of the total air volume, so you won't change the resonance or stiffness (mechanical sensitivity) much. Your electrical sensitivity will go all to hell if you vary the backplate distance much. And your damping will really go away with even slight increase of distance. Of course a lesser distance will just encourage suck-in. If you reduce polarization voltage to eliminate that, you will approach a resistance-controlled system that will (for electrostatic sensing) fall 6dB/oct.

> air leaks will make a good capsule sound crappy.

In this type design, a very slight leak will ruin the acoustic loading.

I must admit, I'm not at all sure how the last-page cardiod mod works. It is the same plan as how we cardioid dynamics and low-price electrets, but the tuning is usually very different between omni and cardioid models. I don't know if he hit on a happy combination of parameters. or if it is just a semi-cardioid.

> This single backplate design is not that good in that there is no adjustment possible on it. It sounds good or bad by chance.

No: by design and implementation. While I can not compute the complete acoustic load, it is perfectly computable and will be as accurate as you can match the plans.

> do not have enough holes in the back plate.

It does look skimpy, doesn't it? I think Williamson has favored relatively large (easily drilled) holes with most damping in the air-film under the diaphragm, instead of smaller (broken-drill) holes with more damping inside the holes. Either scheme can work. Williamson's may be more critical of backplate distance, but he gives us an elegant way to set that to spec.

Williamson has an impressive track-record. He may be an "amateur", but his plans are all technically defensible, practical, and musical to the ear.

Tim I want it as a cardiod. The design is a cardiod it has two thur holes kind of like the c37 capsule I believe.

Why I want to build it is because it is a simpler build(I need to practice how to machine metal) than other capsules and it reminds me of the sony C37 and small dia cardiod capsule like in the AT line as well as "standard" SDs

I have read about the theory. If the theory writeups had everything there would be alot of good capsules in the world. I was looking for "real life" you know someone realy built something and has real life info.

I re-skinned a C37 for someone last night, and one of the neat things about this capsule is you can tension the diaphragm in-mic. So we played with the tension for hours, and found that on that particular backplate, there was a range of tensions that worked. Too high, and the high and low end both rolled off, and too low the diaphragm sucked in, or had wrinkles in it. The best sound we found (and the closest match to his working and untouched C37) was just enough to get the last wrinkle out, plus a hair more. That capsule design is actually quite nice from that perspective, although that means there are a lot of parts in it.

I also checked out a Telefunken U47 for him, and he said that someone (a sleazy mic reseller type) tried to tell him the tube was bad, and the mic wasn't worth anything any more. Well, the tube (VF14M) is fine (34 volts on the plate, exactly what it's supposed to be), but we need to build a new cable (it was bad) and check out the capsule. The capsule looks good - it's an original M7.

From what I remember, the C37 had 4 through holes, and a handful of blind holes - a really easy backplate to make if you forget about the vent valve on the back.

> because of no center term

Entirely superficial. Since no music-recording mikes (except the original PZMs, and the quartz Altec) used diaphragm plate-stiffness, and the diaphragm is perfectly conducting (compared to 100Megs), the presense or lack of a center post does nothing to the sound.

Click to expand...

No, it does. First, with unsupported center, diaphragm has a different deflection due to its bias. Second, it has higher excursion with sound wave. And third--the most important, the resonant modes of such diaphragm are much more complex compare to supported one. In smaller diameter diaphragms this effect is less pronounced, but still plays its role.

[quote author="Gus"]Tim I want it as a cardiod. The design is a cardiod it has two thur holes kind of like the c37 capsule I believe.
[/quote]

This kind of design has been always kinda enigma for me. Except of 10mm SD, I never had a good luck convering capsules designed as cardioids into omni, without compromizing high end. Although when installed, diaphragm will have higher resonance than free air, cardioid midband tuning is not enough for omni, and as PRR has pointed out, omnies and cardioids require different tuning.
Besides of that, it seems for omni operation this capsule doesn't have enough "chamber" space for air to have enough viscousity to damp resonance of tuning. To make this particular capsule work right, the spacer thikness should be just right, which can be problematic and completely offset the seeming ease of machining.
As for cardioid operation, I am not convinced it has that great of the pattern.
But it is everything the way how it looks on paper. You never know until you actually tried it.

Besides, what was good as an "amateur professional" mic in 1960s is not the goal for "The Lab" microhone sharks in 2005.

I'd start with back plate drilled with many very small diameter holes and use a kind of aperture closing different sets of holes, and than see what works the best.