TO FIX FLAWS IN THE OKTAVA MK-219 MICROPHONE. 

Scott Dorsey's version of the MK-219 schematic.

I decided to make changes to six components, prioritised as red and orange in his diagram, but before I did that I modified some physical details to improve the sound. I have my own ideas about that. Some people cut away most of the grille to leave only the basket, but apart from cosmetic improvements I think this risks more than it gains.

PHYSICAL MODIFICATION:

Oktava. Probably the most austere microphone in the world.

Ok, that was unfair. Its sound is not austere, it's more like having superhuman hearing. I like the tank-like aesthetic of these mics, so I leave that well alone. They look as uncompromising as the kind of train that did battle with a tank in the James Bond film Goldeneye. And just as black. That casing does have some small problems, and this is what can be fixed.

View of the internal surfaces of the shell and basket.

The basket is in four parts, and to cut away the grille leaves little anchorage for those parts, and in work I have seen published, another more serious source of resonance is left unsolved! The capsule itself is on a mount that oscillates at low frequencies. It seemed better to me to use a damped coupling to the shell at the top end of the capsule, and to damp the entire shell especially around the capsule to prevent any oscillation remaining undamped, at any frequency. The reason I have seen for cutting away the entire grille is that interference, nodes reinforcing and cancelling, are formed by the array of thick metal struts. I think this is not a significant problem because the gaps are too small, any overall diffusion caused there will be on a much smaller scale than the wavefront as seen by the large diaphragm, so the smoothing of the audio caused by it is far from being a problem, but is one of the nice things people who choose large diaphragm microphones want them for. Much more significant is that these struts are stiff, metal, and resonate along with the entire shell. Also, the internal corners will cause a much greater disturbance to the internal wave propagation than is caused by the grille.

Black silicone sealant and thin double-sided foam tape.

First, details of damping materials: Black low modulus silicone sealant is cheap, easy to find online. Don't use the normal domestic stuff, it cures with release of acetic acid! That's a very bad idea close to a delicate microphone capsule. Low modulus sealant cures by releasing alcohol vapour which won't damage any part of the microphone. The foam tape came from a motor vehicle parts shop. It's about 0.8 mm thick, sticky on both sides, very stable over periods of many years, and unlike PVC electrical tape its adhesive does not dry out or ooze or seep after long periods of heating or pressure.

Internal damping with low modulus silicone sealant.

It's not easy to see exactly where and how the sealant is laid into the shell, even by careful comparison between the two pictures of the shell, and there are several important details to watch for, so I'll spell them out in detail now. First, the woven wire basket parts have a line of sealant about 4 mm thick along every edge. This helps to anchor them as well as to damp the metal shell and grille, and it also absorbs and diffuses objectionable resonances caused by the internal corners. To gauge the thickness of the sealant, examine the placement of the circuit and capsule mount before applying any of it. The main area to keep clear is that covering the transformer. I add sealant there too, at first, to about 4 mm depth, having first worked out where to scoop out excess later with a fingertip. Take care not to apply any at points where other parts must directly contact the shell halves. It doesn't matter if the top ends of some of the standing components push into the silicone when closing the shell, that will help reduce resonance in those parts as well as in the circuit board, but avoid any amount that would cause significant pressure, as it would if it pressed onto the transformer near the connector end. A small extra bit of damping may be had by running a line of sealant down the vertical struts of the grille, behind the basket, but I did not do this because an open chamber around the capsule is a better way to maintain a good wavefront reaching the diaphragm. It would also do little good unless it penetrated the basket weave to bond it to the grille. A carefully laid line of a glue like Evo-Stik might work because it is runny when fresh but partly elastic and partly damped when cured, and it forms a strong bond, but I decided the gain was not worth the trouble of avoiding the likely mess. That stuff tends to get everywhere it can, and is hard to clean up. When the sealant is in place, leave it to cure for at least 24 hours, preferably at least two days. This is a very good time to replace some components on the circuit board, but I hadn't got mine yet so I solved another problem first...

Detail of capsule damping with thin layers of foam tape.

The low frequency resonance in the capsule mount became obvious once the metallic ringing of the shell was strongly reduced. The obvious cause was the mass of the capsule assembly and the spring tension in the base of its mount. In turn, the obvious place to fix it was at the other end, using something to press between it and the inside of the top of the shell while minimising any obstruction to sound coming through the grille. This would also lower the resonant frequency of that shell and add further damping. The method is much easier to explain than the reason for it: stick a few layers of foam tape on the top end of the capsule, enough for gentle but firm pressure, and leave the top layer covered by its protective film to make it easy to disassemble the microphone in future if needed. The microphone photographed on its box is already modified, and shows how completely undetectable this modification is, by external examination. It's ideal for people who want to preserve the original state of it as much as possible.

ELECTRONIC MODIFICATION:

New components. The 1.1 giga-ohm resistors are Russian.

The FET is a Toshiba 2SK170BL, bought as 'new old stock' on eBay. The resistors, also from eBay, are from Russia, they're 1.1G, slightly more than the originals, and they work very well. They were described by the seller as "Modern production, excellent quality, Small size:10,8mm 4,2mm, High-value, high voltage thin-metal-film resistors, All climate version - stable at wide range of atmospheric pressure and humidity - Operating temperature range from -60°C to 125°C, Great for measurement equipment and for many DIY projects like tube condenser microphone builds". I like them, they remind me of the best military spec parts I've seen from various locations and ages. The other metal film resistors are standard metal film, 0.6W, again bought on eBay, which is the fastest and cheapest way to find the unusual values of 1.78K and 2K in small quantities. The capacitor is a COG dielectric ceramic type. I stayed with the original 680pF value instead of using 820pF, partly because I could only find a wired capacitor of this type in that value. Most of them are now SMT types, and it's worth the trouble to find wired capacitors for this work.

The original board, before any modifications are done.

New resistors, FET, and capacitor for the input stage.

Instead of writing lots of detail about these parts, I put those two pictures close to each other because careful observation will tell more than I can explain if I doubled the amount of text on this page. One detail that may not be clearly visible is the middle lead on the new FET, which is the one that connects to the new capacitor's solder pad near the capsule end of the board.
EXAMPLE RECORDINGS:
This RAR file contains three samples: undamped, with shell lining, and with capsule mount damping. Two days after doing this, I recorded some sound capable of showing the clarity and dynamics these microphones can capture. There's a text file in the RAR file with more details. I put it there because if the file is found wandering loose, it makes little sense to have those details here.