My M49 style mic

Hi, I am putting together a couple PCB's for some m49 style mic's using 5840's instead of 701's.

I am using the TAB schematic here as my guide.

I have a couple Apex 460 style mic's that will donate their bodies/capsules/psu's.
Also have a couple TAB T49 trafo's.
Plan to run heater around 5.7V with a dropping resistor from the stock supply (6.3V) and cathode bypass around 10uf. May try some values in that general 10uf range to see how it affects operation as well as different output caps. I have a handful of the large square russian PIO's @ 1.5uf which I used in some other mic experiments.

My question is regarding the 5840 and triode strapping it and general hookup of that tube.
Is it recommended to tie the plate lead and the Grid#2 lead together or with resistance between them?
Also, lead#'s 2,4,8 are tied together internally with 2&8 both being cathode and 4 being Grid#3. Is it ok to snip (for instance) Lead#'s 2-&-4 and just use#8 as my cathode? Maybe some other hookup's that would affect operating points?

Just a fun experiment that I hope turns out well.

I may experiment with some eq networks I have seen around here to attempt to make the capsule not so bright. May end up putting in a different capsule.

At the moment I am laying out component placement to make the innards as sturdy, neat, and accessible as possible while paying attention to ground connections as well as high-impedance connections.

I need an M7 to make it closer. If I can get my layout to be fairly sturdy and robust I may spring for nicer capsules.

I picked up about 20 of these 5940's so am looking forward to do up a handful of mic's while experimenting with various config's.

Best,
jonathan

About 2,4, and 8, it's ok to use just one of them:

http://www.groupdiy.com/index.php?topic=37972.msg467822#msg467822

And I just tied plate and grid 2 together. There might be other ways to hook it up though.

Thanks for the link Pasi. Some good 5840 info.

I might also try giving Grid 2 the B+ direct or with a dropping resistor as well as tying to cathode, ground, or maybe an elevated/bypassed ground like a cathode self-bias circuit. Looks like there is some room to experiment. Maybe I can use Grid2 in an eq network somehow to reduce some peakiness of the k67 style capsules I will be using initially.

Does anyone have any experience drilling/mounting for Keystone PTFE standoff's? I've got a couple different sizes. Datasheet says #44 bit for the smaller one and 4mm for the slightly larger one. I tried a couple test holes which I tried enlarging using a reamer but they don't fit secure. I've got some #44 bits at home (used as the priming hole for a 4-40 tap) so I'll try that (prolly need to pick up some 4mm bits) but was wondering if anyone had any "installation-tips/tricks". I would like these to be fairly permanently mounted. Maybe a drop of cyanocrylate or something? I don't particularly want to jeopardize the electrical isolation with a glue though.

Best,
JoBu

No #44 bits in house ATM but a 5/32nd with a little blade action to open it up slightly (just a blonde hair now. Maybe two) appears to do the trick. Nix on the glue idea. I will want to clean this with some alcohol/solvent which will most likely weaken any glue joint considerably. I believe the ptfe will expand after a bit to lock itself into the hole it was just a little large to easily slide through.

I am liking the terminal with turrets on both sides.

Cheers,
JoBu

EDIT: Here is a photo of the two PTFE turrets mounted. With the smaller one, the hole ended up a bit large and I tested using some high-impedance-craft-glue to see how well it would hold. Dried overnight and it's not too tight but it definitely doesn't fall out. I am thinking the proper bit will sort things significantly but if you need to make-do with some standard bit's then the slightly larger standoff can be made to work very well with a 5/32 bit with the added benefit of having turrets on both sides of the board.

BTW, Here is the mouser link to the larger keystone turret:
Keystone-Electronics Part# 11251 @ Mouser
-JoBu

Updating with some photo's.

Stuffing the perfboard

-jb

Something like this:

-jb

Pretty good, JB! Have you heard it yet?

The tube's on the other side, right?  ;D

Dan, Thanks!

I've got to finish this up still so have not heard it yet and I need to get mic#2 to the same point as mic#1. Tube is on the other side and I may re-solder that with some heat-shrink over the leads (no heat-shrink on grid lead).

I'll make more pictures as I go. This a a leisure project for sure but I finally settled on the perf-board layout.

I am using the stock Chinese PSU.. LM317 for 6.3V so will need to readjust that for 5.7V in circuit... Don't expect to require additional filtering since this is a self-bias circuit.

Cheers,
jb

EDIT:
If these turn out really well I might get some heavier duty fiberglass board and turrets for a more "robust" installation.... I have noticed some "tensioning" of the perfboard as I solder up components... Will likely need to "reflow" everything once it's all mounted to hopefully prevent undesirable stress on the joints.

Wired it up to test. Here are some photo's.

cheers,
jb

Sounds quite nice with the stock capsule and a voice test.
I also measured some voltages and need to tweak a bit to get it all into spec.
Cardioid and fig-8 patterns work as expected but omni provides silence.
B+ is sitting at about 150V so need to tweak that down to 120V.
Pattern selector provides about 120V - 0V... I think I'll take a look there and see if I can get more accurate voltages.
Lastly, I need to drop the heater voltage to target of 5,7V...
I'm going to mess with this one for a bit and see if I can get all the issues sorted out to have a good starting point before burning it for a while.
All that said, it's sounding nice already. I'll keep the stock capsules in until I can get all that sorted on both mic's then swap one with the C12 copy (the recent GB from WGT) and see how they compare.

Cheers,
jb

These photo's were after I wired it up. I use cat5 (pre-twisted) for internal wiring....

Here's a shot of the PSU from the top. This is from the tnc group-buy acm1200 mic which is the same as an apex460 I guess.

-jb

And the schematic for the PSU:

-jb

EDIT: That's an L7806 not 7602... Regulator output: 5.76V - 6.24V....

So...

Looks like I can get 120V B+ by re-placing the B+ resistor string to drop / filter significant voltage and replace the 75V zeners w/60V zeners... I'll just measure drop across plate/cathode to figure total impedance of tube in this operating range and then solve for a divider to get 120V on top of Rp/100k. Not even sure if the zeners are really that necessary.

Dropping the heater down to the 5.0-5.7V range may be a little different. The L7806 data sheet looks like it should provide 6V yet it measures 6.35V @ the heater... I could replace it with an L7605 which would then drop to maybe 5.35V but that's probably too low. I could also insert a dropping resistor in series after the output pin of the existing L7806 (or perhaps a 5W 25R pot) and tweak.

If I used a pot/resistor would another filter cap after the dropping-R be advisable? 1000uF/16V?

Looks like this PSU will need to be completely disassembled to get access to underneath to resolder/modify... At that point it's probably not too hard to re-work in a more "linear" topology... I would like to better understand their design approach strengths/weaknesses and see where that goes..

Any comments appreciated.

Best,
JoBu

Hi,
very nice project! I had to repair the PSU for my KM56 and used an LM317 regulator for the heater supply. The mic was much noisier with this psu. Ioaudio recomended to add a RC stage right after the LM317 to drop exactly 1V. I did so and turned the pot up to get back this 1V. My Neumann KM56 is dead silent now. I suppose you can do the same with your L7806 since you only need about 5,5V heater voltage. I always wanted to do a M49 circuit as well. Very inspiring and motivating thread.
regards
Bernd

0dbfs said:

Dropping the heater down to the 5.0-5.7V range may be a little different. The L7806 data sheet looks like it should provide 6V yet it measures 6.35V @ the heater... I could replace it with an L7605 which would then drop to maybe 5.35V but that's probably too low. I could also insert a dropping resistor in series after the output pin of the existing L7806 (or perhaps a 5W 25R pot) and tweak.

Click to expand...

You could put a diode in series after the regulator to drop .7 volt.

David

David, Thanks... I spent some time this morning measuring voltages and figuring currents.
With the stock supply and Oliver's M49 conversion circuit populated on the perfboard I found I was able to fairly easily swap in different values, measure and come up with a plan.

I was looking at other methods and decided to try putting a diode in series w/heater because:
1. It was easy to do in the mic-body and with that method I didn't have to disassemble the PSU and order more parts.
2. I have some diodes (extra 1N914's from some other build).
3. I was able to insert the diode in circuit and quickly re-measure voltages/currents and better understand what significance dropping the heater voltage has on the 5840w operating points without considering even sonic attributes (of course I listened though!).

At this point I considered oliver's schematic & voltage spec's  and came up with a plate voltage requirement of about 45V. With my heater at 6.35V and 100k plate resistor I had about 63V on the plate due to a measured B+ of 148V and .89mA idle current.

B+ provides capsule polarization and pattern select voltages as well as supplying the plate through the dropping resistor so I guess a 75V capsule polarization is fine with the stock capsule and most capsules are fine up to what 80V right? Slightly better s/n with a slightly higher possibility of a stuck diaphragm.

So, I reasoned that if I change the plate dropping resistor to 150k i'll end up with about 40V on the plate and capsule polarization of about 75V. Without having to mod the PSU at all.

Works great and is quiet! At least on first listen.

With the diode in series and a 150k plate dropping resistor, I now end up with 5.45V on the heater, about 43V on the plate, and with a 2k2 on the cathode I measure about 1.5V of bias.
10uF cathode bypass recommended and installed to extend bass response when subbing 5840 for ac701k in this m49c conversion.
I am going to let it burn in for the rest of the weekend and clean up my work a bit to make it more robust.
And I need to swap in the C12 style capsule to compare with the stock capsule. First I need to build up mic#2 so I can get to a reference between the two with stock capsules.
I'm hoping to get these two happening for a session next week to test them out in action. We'll see how the burn-in testing, assembly of #2, and capsule swap goes.

Cheers,
jb

I put another of these together over the weekend using #1 as my template. It came together fairly quickly and without issue.

I kept the stock capsule in both and they sound "the same". I then swapped in the c12 style from the recent group buy. I am liking it much better but am reserving judgement until after I listen tonight and tomorro and compare a little side by side.

I am wondering what effect raising the plate resistor by 50k will have on the sound regarding interactions with the output-cap/transformer-primary/reflected-impedance, etc...

When I switch towards omni (or power-up$ the capsule mutes until it charges. This takes several seconds. Maybe 30. Anyway, I can hear a buzz during that time when there  is no acoustic signal. Must be from the heater. Definitely in the noise floor once capsule is charged but wondering if others hear this in other builds.

Cheers,
jobu

0dbfs said:

When I switch towards omni (or power-up$ the capsule mutes until it charges. This takes several seconds. Maybe 30. Anyway, I can hear a buzz during that time when there  is no acoustic signal. Must be from the heater. Definitely in the noise floor once capsule is charged but wondering if others hear this in other builds.

Click to expand...

I have had the same problem using an LM317 to regulate the heater supply.  I think you need more filtering between the transformer and the regulator (that was my problem).  If you use an LM317, you can also get an improvement in hum by putting a C across the resistor the goes from the ADJ pin to ground.  See the LM317 datasheet.  (In fact, the whole problem of getting the right voltage at the heater may be solved by using an LM317 rather than the 7812 I see in your schematic above.)

David

Hey Guys, Thanks so much for all the info.

I am getting ready to make a mouser order for some other mic's and want to get some parts for these as well as I would like to experiment a bit.

In order to get the B+ down to 120V I will need to modify the PSU. It currently uses two 75V zeners which I could easily replace with 60V zeners. Any pointers on which ones may "work better"? ie; less noise or perhaps other characteristics that may be helpful? I've got some 317's already but will order a couple more since I am ordering.

Thx,
jb