THE EXTINCT BATTERY
As CM-300 owners we all know the original battery is no longer made. The modern replacement used to be an Eveready 206 but I haven't seen those since 2000. The modern solution is to stack small value cell batteries to achieve 9V but one caveat is that you should always use the same brand and voltage (or multiples of). The reason is that if you use different voltages (ex: 5v and 4v ) the mismatch can cause minor current flow and drain the batteries faster. Different brands can have different resistances which can also cause a drain. So for best results build a stack using identical voltages from the same brand.

For my CM-300s I use three "CR-1/3N" batteries. These 3V lithium-ion cell batteries are far superior to the old mercury/lead acid types and together produce about 9.3V which is what this mic is spec'd for. The batteries' OD is smaller than the mic's so I went to the hardware store and bought some clear flexible tubing the ID of the mic. I wrapped the batteries with a few turns of electrical tape to fit the inside of the tube. I used a spring between them to pass the voltage. So this method looks strange and isn't very elegant but it worked for me.

CIRCUIT MODS
I wanted to retain the stock functionality and decided replacing the 30+ year old parts on the circuit board was the best plan. If you pry up the switch plate you can unscrew the mike. Inside I noticed something strange: the parts values were different from what was printed in the owners manual and even what was printed on the PCB board(!) The 100pf cap was actually .068uf and the 47uf coupling cap was a 10uf. Not unexpected was that the 10uf coupling cap was a tantalum type. Tantalums are classed as electrolytics and have been used for decades as input caps in mics because of their tiny size. Photos of some modern Neumann mics still show tantalums in places. However by modern standards they are long obsolete and widely considered to produce poor sonics. Tantalum is also a conflict mineral which is unfortunate. You really can't blame Nak as back in 1990 a tantalum would have been the only cap that would fit. My 21st century opinion is there is no valid reason to be using a tantalum in an audio circuit as critical as a microphone.

I changed the two original carbon 5% resistors to 1% Dale CMF metal films to take advantage of their very low noise. The .068 mylar became an Epcos MKT of Polyethylene Terephthalate (PET). Because of limited space for the 10uf coupling cap there was only one choice: a modern audio-quality electrolytic.

I decided to use a 10uf/16V Elna Silmic II. To my ears Silmics provide an open, dynamically textured sound which I find musical. Interestingly I find Silmics can sound slightly soft at first but after several hours of use they suddenly open up; one of the few times I've heard an audible break-in of a board component. Low quality caps can have a bright sound which is misleading. This brighter glaze may sound clean at first but produces flatter, less realistic dynamics to my ears.

BELOW: To get inside the capsule body remove the C-clip from the bottom and press down on the brass pickup to push the plastic insert out the bottom. Once out the insert snaps apart. Just as before the values inside were wrong compared to the schematic: the 1.2K resistor was a 2.2K which I replaced with a Dale CMF. Next to the FET is a mysterious resistor not labeled in the schematic: Brown-Black-Violet calculates as 100M ohm which makes it a bias resistor to pass DC to the FET. The FET is a tiny 2SK118 transistor listed for "General Purpose and Impedance Converter and Condenser Microphone Applications". This FET may have been hand-selected originally and modern day replacements are scarce and questionable so I left it in place.

Hard to see but Nak connected the 100M resistor to one leg of the FET using a shared solder joint in the center of the board. This is a bad practice as critical parts touching the board can introduce noise into the circuit. Air is the perfect dielectric. So when I replaced the 100M ohm resistor with a 1% MOX200F100 resistor (not shown) I bent up the FET lead in mid air and soldered the new resistor to it.

The ceramic disc cap is the -10db circuit and was labeled '151' instead of 220pf as in the schematic. I don't use the -10db circuit so I left it alone. NOTE: the foil trace on the bottom of the circular board is super thin and heats up instantly. So I used a heat sink on the FET when soldering to protect it. I finalized things by using Chemtronics flux-off to clean any residue from the board.

RESULTS
I had to tilt the Silmic slightly to clear a rib in the mic shell. Everything else fit perfectly. The new blue 15K Dale is a Mil-spec CMF/RN50, the 2.2K inside the capsule (above) is also a CMF/RN50. While I was in there I replaced the red/white/black wires that go to the XLR jack with new stranded Canare 22g.

Using the same preamp, headphones, batteries, and levels from before the mod I was flat out astonished when I powered the Naks back up. Before I had to get reasonably close to the microphones to pickup a clear signal, they had a dull, closed-in sound. To be honest I had never been too impressed. Now, the noise floor floor is vanishingly silent, with an articulate, spacious sound. I can hear creaks in the room, someone walking in another part of the house, my whispered breath, my computer fan, actual room ambiance. Switching the low cut filter on didn't diminish the magic.

I never expected such a difference from 3 Dale resistors and a Silmic. I can only conclude that the original tantalum was pretty horrific and that the original carbon film resistors were poor sonic choices. Isolating the FET from the circuit board may also have lowered the overall noise levels. Take note.