DIY U-47 from Cascade M20U

[desol]

this mic sounds gorgeous!

Why am i not surprised. How much do you want for it Chris...? 

 

[RuudNL]

According to the schematic, the polarisation voltage is 34 V.
But IMHO it can't be more than 28.6 V ( 1000/1680 * 48 V)
Isn't that voltage a bit low? Or better: wouldn't the microphone perform even better if you would raise the polarisation voltage?
Modern capsules don't have any problem with 60 V.
More output + same noise = better S/N ratio!

 

[CJ]

it is hard to measure the actual polarization voltage at the membrane without some expensive gear, the 100M resistor is upset when any meter under 10 gig ohm input is used, maybe a scope could be used on dc,

the hp supply went up to 50 volts, so we were a bit above normal phantom 48 when some mesurements were taken, current might have been  38 ma 

i measured the stock cascade m20u capsule at the 1 gig resistor  before i took out the old board, and it was 30 volts, so i did not want to warp the membrane into the backplate, but i thought we could have a 5% factor of safety so i tried to raise it a tad bit over 30,

it is very quiet right now, i want to side by side it with a audio technica 4035,

we backed the 0.68 plate cap down to 0.33, same foil wima, still have 43,000 ohms inductance and 43,000 ohms cap to offer 8 microamp excitation, might need a termination resistor to draw a bit more current to reduce distortion, plenty of headroom, you can shout into the thing and no overload, using api preamp for testing,

the old cap was worked from all the jangling around if you check all the pics it was just hanging around, so i do not trust the leads anymore, thus the change to 0.33 the bass came up when i bars were added to the xfmr, 31UI will lose a ton of perm without the i bars to complete the mag path,

a neve transformer could work here also, it would fit easier,

if using utc lams, reduce lam count to 34 like ha100x so you do not have such a tight fit, you will still have a ton more henries than stock bv8,


the xlr melted when i tried to solder,  :-[

so i took the guts out of a pre wired neutric with ferrite beads and 10 pf caps installed, it bolts right into the cascade jack plate, 

 

[CJ]

other side of mic pictured below

now we have to figure out the phantom thing, :'(

i did some math on the u47 stock vf14 heater circuit,

that 1780 ohm resistor is putting out about 3 watts so a 6 watt resistor would be practical, is that what the big bosch can is at the bottom of some of those mics?

ok so what we were trying to do here on this build:

use funky german parts from the 40's and 50's,
improve grounding by soldering direct connections,
get rid of as much wire as possible
use the classic u47 circuit
use weird techniques that commercial companies can not use to get lower noise,
used same starve plate and heater technique for low hiss as used on vf14,
use large resistors to drop resistor noise,
use a crazy large transformer for  low distortion, hi headroom, and extended bottom
make it all run on 48 vdc.

oh yeah, the most important thing, make it diy dirt cheap, 

a battery pack might be nice here as we only need 34 ma, 34,000 mili-amp hour battery  = 1000 hours of ray charles gumming up your capsule with cheap azz cigar smoker and corn liquor, 

 

[CJ]

Ruund your a Genius!


i put the capsule t 50 volts like Rund says and the s/n ratio instantly gets better, more punch, more bottom,

maybe that is the whole problem with this cascade mic in the first place, the capsule is at 30 volts compared to 63 volts on the stock u47,
it really likes the langevin am16, it show a different character with each preamp, dead quiet with the api, super mid-range with the v76, over the top high end with the quad 8,

it sounds a ton better with the capsule at 48-50 volts,

maybe just add a cap at the 48 volt input to reject anything going on up stream at the preamp,

 

[bockaudio]

you need an electrometer to measure the polarization volts. Ebay. Even then measure before  the resistor.
FWIW when our model 195 fet mic has the pad on it's about 21v polarization, normal Vp is 45v, w/healthy phantom.

 

[0dbfs]

CJ, That's an inspiring build as we seem to be approaching the event horizon. And as we get closer I realized that I can't see your capsule/grid connection... Is it going through he 50 smeg?

-j

 

[pasarski]

you need an electrometer to measure the polarization volts. Ebay. Even then measure before  the resistor.
...

Couple of observations that puzzle me:

1) In these forums I've learned: measure before the voltage divider with a regular DMM and do the the math.

2) Every time (actually only few times) I have measured with regular DMM before the polarization resistor I get ~ the same reading than in case 1.

So I wonder, what's the point of procedure 1), let alone an electrometer?




P.S. Big up for CJ, and also Mr. Bock, I'm so glad you're contributing to this forum, big up for you too!

And BTW, It's already 21.12.12 in Finland and world didn't end, so no worries there over seas!

 

[bockaudio]

there are at least a few reasons to bother measuring rather than relying upon calculations.
1) you can confirm a voltage actually exists where it supposed to exist.
As to the electrometer, though the reading may be wrong after a high z resistor, you can still confirm (1) from above.
Example: we had a defective new mic. Using the electrometer I was able to determine that a brand new 0.01uF cap was shorting the rear polarization voltage (0-120v) to ground, though a static ohmeter measurement of that cap showed no such short. A regular DMM was ineffective at finding this short in the two resistor one cap network.
There are plenty of other circumstances where the EM tell the true voltage that the DMM cannot. But I understand why a hobbyist wouldn't have one.

 

[CJ]

capacitor leakage will drop the capsule voltage down in a hurry because of the hi value resistor 100m to 1 gig,

so an electro-meter would show how much current, if any is being leaked by the cap and the capsule.

electro-meter just like color tv picture tube anode voltage checker on reverse, 

this thing blows away the audio terchnica 4035 that i thought was so great,
nice to finally have a good vocal mic, 4035 sounds like brittle mid-range with harsh distortion compared to the utc47, 

maybe i can do some sound clips with this tascam dr07 pcm recorder, has 24 bit wav.

and i will fix the schematic to include a 100k inplace of the 680k on the voltage divider for increaded capsule voltage,

maybe jack up the phantom supply to 60 volts, have to pick the best mic pre and wire it special for this mic,

 

[pasarski]

Just to be sure, we're not talking about "The Hubbard  electrometer"? 

Seriously, thank you for answering, learned something new again. I see there are used electrometers for sale starting for about 200 euros, so not out of reach for a hobbyist. There actually have been situations for me also when a DMM tells everything is ok but apparently isn't.

 

[CJ]

here is a pic of the completed Cascade mic,

we just spliced in a bannana pug thingy for power so we can run it off a regular DC supply which means we can get the capsule voltage up off the noise floor by a greater amount,

it sounds fantastic, put it up against a beta87a and it sounds richer in the mid without the hi end bite,

 

[CJ]

here is the final schematic with the voltage divider adjustment,

added a 100K in place of the 680K  R1 for more capsule voltage,

no problems were encountered with this change,

capsule voltage goes up from 34 to 42 with 48 VDC input,

 

[fum]

It's always great to see CJ doin what CJ does best!

Back on the grid I take it? 

 

[kato]

I secretly admire dudes who can roll their own cigs. But I openly admire dudes who can roll their own transfos.

I'd bow to you but I'm afraid you'd kick me in the face. So please accept this virtual metal sign as my expression of goddamn. 

 

[CJ]

used to ride horses on the beach at Neskowin, Oregon Territory,

this old sourdough who owned the stables, had a beat up hat and a 5 oclock shadow whips out a bag of bll durham, holds the reins in one hand,  and rolls one up one handed in a stiff wind, no gummed papers, just a good snot wad and he was smokin like a chimney, a lost art i tell ya, i can't even keep this hash pipe lit and i have  crack torch,

Fum, where the hell, whats new in the rack?

 

[fum]

Fum, where the hell, whats new in the rack?

Too much secret squirrel stuff on the electronics front, not allowed to talk about it.  :-X 

Worked on this if you want to catch the drift(have to scroll down a bit):

http://www.purpleaudio.com/ox/

Will share something DIY in a bit.

I'm still out here, ribbon mics keep me busy.

It's been a while, miss my DIY roots.  Good to see you here CJ!

ju

 

[CJ]

cool on the mixer!  looks like the api format,

good to see you and remember, don't eat the food at the space needle, 

come see us at the Tacoma Dome next weekend,

let me know if you need any ribbon transformers,

South Sounder makin sounds,

cj

 

[kato]

What for is the white plastic insulator thing at the butt end of R1?

Also, what supply voltage did you settle on - 60v?  Or is it good at 48v with R7 dropped to 100k?

 

[CJ]

that is a nylon standoff,

yes, 100K for R 1 brings the capsule voltage from  30 to about 40 volts and change if you use 48 volt phantom,

you must use the 48 right off the supply, not after  5.11 k resistors