5840 tube mic - front address

I've built a front address LDC tube mic using a 5840 tube (like the Royer circuit) and the capsule biasing like an M49 (to omit the cap between the capsule and grid). Audio Technica body.
The mic works but the patterns don't seem correct. I think it may be acoustical from how I mounted the capsule? Anyway, I thought I'd show some pictures.
My goal is to make a very accurate instrument mic that has a fig-8 pattern for rejecting vocals while mic'ing acoustic guitar. A 34mm will not fit into the body, unfortunately, so I am using a capsule from an Apex460.

Here's how it fits into the grill

And fully assembled

How can it be figure-of-eight if the back diaphragm can't recieve sound (due to the mic body's interference with soundwaves)? Cardioid-only might be your best bet for an end address design. Check out recording hacks website to see how RE-20's are constructed. Although it is a dynamic mic, this may help you with capsule placement, rear venting, etc.

There is a 1/4" between the rubber mount and the capsule. Maybe this design is blocking a significant amount of sound?
My understanding of mic design (RE-20 for instance), is that the mic would be omni, and the rear venting is designed to cancel and create a cardiod. The venting is also done to even the frequency response (on and off axis). I've been slowly reading a book on microphones by lou burroughs that goes into some detail.

Yep. That is common dynamic mic design. The main difficulty with your capsule placement is that the body of the mic blocks direct sound from entering the back. Moving the capsule closer to the end you want to address may help some, but I doubt it will fix the problem. If you look into the Oktava MK-012, the way they allow figure-of-eight on an end-address mic is to make it side-address for that particular pattern. Maybe someone knows how to accomplish what you are looking for. I hope it works out.

-James-

So I switched the capsule to be vertical and the patterns still didn't work, then I discovered that I had a mistake in the backside capsule polarization. I had a 0.1uf connected to the backside of the capsule as well as the 1000pf across the capsule (C7 & C1 on the M49b schematic) without a resistor in between. So, any sound was being damped out by the big 0.1uf cap. Added a 1G resistor between them and a 10k feeding the 0.1uf, and viola, patterns work.
Now to try front address again...
But the mic sounds really good as it is...

Let me know it goes. I have more 5840's laying around just looking for a home...

-James-

ya man that sweet. love the home brew approach. I need to start building some mics too! (into an old m421 body for starts)

The 5840 seems to be an awesome tube for mics... I bought a bunch on ebay and have a few rejects, but most are quiet and sound great.

5840 jan's are the way to go and fairly inexpensive.

Nice work!
I'm just getting into my first tube mic project attempts..
and researching already for future projects, which got me looking back to my notes on the 5840.
I have one of those old MXL capsules (with the non-tube circuit still attached), plus maybe an Equinox U47-style body kit left over to play around with.  Keep us updated on how this project comes to conclude?
(Post some sound files?)
Evan.