DIY U-47 from Cascade M20U

[CJ]

what is this, another pipe dream?

well, this mic is so noisy stock that it sits in some dark place never to be used again,

i am going to try and replace the solid state stuff with a mini 5678 raytheon tube,

also, notice the cheezy 1 gig resistors, well we have some german 100 m glass resistors from a surplus score a dozen years back, so we will give them a try.

also, we need a BV-08, plus maybe some more bass, so we can wind 2600 turns and 400 turns on some 31 UI with a stack good for 750 henries instead of 150 like the stock version,

using the small tube makes more room for hi perm metal, oh yeah...

we have 30 volts on the capsule stock, so we can run this off of 48 phantom, no need for 5 terminal connectors or power supplies,

the tube takes 65 volts max so it too will run off of 48 volts, heater is only 1.25 volts but we need 50 ma,

anybody ever mod this cascade mic?

tube data:

http://www.mif.pg.gda.pl/homepages/frank/sheets/127/5/5678.pdf

 

[SSLtech]

You'll never get anywhere NEAR 50mA from 48V through a pair of 6.8k resistors... Power supply WILL be necessary, unless you can build an efficient DC-DC stepdown.

Looks like standard multi-brand Chinese stuff... Useful for the case, for sure!

 

[CJ]

keefer where the hell!

yeah your right, we will build a special 48 vdc super clean for this, still get to use the 3 wire cord,

now if i under-heat the heater, maybe i only need 35 ma, we will check it out,

yeah, this is a free mic and free parts so i do not expect miracles without a crusty M7 capsule, but it should be better than the 100 doller POS that they sell,

the 5678 has a spray on shield for noise,

maybe build a DIY hearing aid in case the mic does not work out, 

xfmr is 150 ohms and 58 ohms dcr on a 25 EI  , shiny lams = 80 ni or a close foreign facsimile thereof,

 

[gemini86]

I tried doing this with a 5672 tube once... I failed... but you. You will succeed!

 

[CJ]

as long as it does not catch on fire i will consider it a victory, just a short winter project as we have the same weather as eugene, ever been to the slide at triangle lake?
that's my next ex wife from san diego,

http://www.facebook.com/photo.php?v=101773973277575

here is the cascade xfmr, wound sec pri sec probably to shield the pri from noise, about 2000 turns split for the sec,  and then maybe 800 turns for the pri,

 

[gemini86]

No, haven't been. I'm a bit of a shut in, last few years...

 

[CJ]

rock climbing accident at skinner buttes?

go ahead and fab the capsule mount to accept glass resistors of 100 m and 70 m since they go directly to the capsiule.

we can rid ourselves of 4 inches of noise and high z antenna by doing this, however, the resistors can be antennae also,

notice the adcom cross coil cart, another high noise device.

 

[CJ]

reference stuff:


the circuit-original:

http://www.wagner-microphones.com/images/u47.gif

vintage pics for stealing part info from

http://www.coutant.org/u47/index.html

tube data sheet which we use to correct heater voltage and cathode/plate resistors so as to match operating point of vf14 as closely as possible,

http://www.mif.pg.gda.pl/homepages/frank/sheets/127/5/5678.pdf

original tube vf14 data:

http://www.hupse.eu/radio/tubes/VF14.htm

original bv-08 blueprint:

http://www.infernalemachine.fr/TECH/NEUMANN/U47Tranny_Winding.pdf

 

[CJ]

since we are  using 750 henries instead of 150 henries like stock,

we need to calculate the resonant frequency of the tank circuit

then we can adjust the output cap to keep the same resonance,

a 0.5 uf cap and 150 henries are used in the original U47 circuit,

so we need a calculator which is right here:

http://www.deephaven.co.uk/lc.html

it says rez freq for .5 and 150 = 18.377 hertz

so for 750 henries we get
0.1 uf at 250 volts = 18 hertz.

so with the bigger transformer, we can use a smaller cap and save some space while maintaining the 18 hertz rolloff,

we can always add a bigger cap for 5 or 10 hertz, but these condenser mics can have low freq rumble problems when moved, so we might want to roll off ahead of this rumble frequency,


so why use the bigger core if you are just going to shut it down with a smaller cap?

distortion.

low freq distortion is directly proportional to the size of the core.

so we should have better specs than the original u47 at the low end.

 

[RuudNL]

Strange that this microphone is so noisy. I just read that the specifications say that :
"Equivalent noise level: (A-Weighted per IEC 286-4) <16dB"
I have some "el cheapo" McCrypt BM-700 microphones (sold by Conrad) that look more or less the same, but are not noisy at all.
(It seems they are also sold as ISK BM-700)
In fact the microphones produce less noise than my Neumann U87!
The 1 G resistors are indeed very fragile and there are better alternatives.
However: 1 G and 30 pF (usually more with larger size diaphragms) form a low pass filter, that will eliminate most of the noise from the resistors.
But good luck with your project!!

 

[CJ]

maybe one of the fets went bad because i had 50 percent niagra falls in the headphones,

ok we have R1 R2 and C1 installed point to point plus the tube and a new circuit board,

be sure to guard the capsule with some foil jic you drip solder, 

R1 and R2 are pre roderstein glass at 100 smeg 1% and 50 smeg at 0.5 %
we have a 0.01 at 1000 volts, 1950's berlin, ero kt,  epoxy end caps, for low leakage as this is a 100 smeg circuit

who will be the wise guy who says i cause phase distortion by having r2 in front of the capsule? 

 

[bockaudio]

your first mic build?

 

[CJ]

how can you tell?

yes i know the grid wire is missing,

 

[tchgtr]

This thread is awesome! Most fun thing I've seen here in a while.

 

[CJ]

danke gut!  un dir?

ok, we added r3,4,5,6,7,8 and c2,3,

star ground is the top mounting plate, direct solder job, point to point,

notice r3, this was specailly flown in from the rectumstein institute after under going severe noise testing at the seimens plant before they sold out to landis and gear..

r4,5 and 6 are 1940's resistors from IRT, before the kaiser had all of them destroyed in favor of the new rottinghham resistors from steingetz,

pay particular attention to c2, the output cap. this is  molded foil which had to be stitched together with virgin pine tar which you can only find in the upper rhine valley.
somehow they survived the bombing of berlin and resurfaced at the neumann plant in the uttageist valley. the ESR is so low that it cannot be measured even with todays instruments. wima was the soul source for this part up until telefunken was bought out by the hearst company in 68.

the elko caps will blow up if you use them at anything over half their voltage rating,

so we run a 2.2 uf at 350 and hope for the best.
gotta have a leaky elko for that vintage electrolyte smell,
you say this mic looks kludgy?

ever worked on a blaupunct car stereo? 

 

[CJ]

please notice how the inter electrode capacitance has been used to simualte a ground filter which works in conjunction with dis function junction.

we are gonna try a 0.68 at 250 and see if it is too much bass,
we can always roll it back to the computed 0.1 if too much rumble.

to get a chinnese capsule to sound german, you must use vintage german parts.
the 1.5 k is gonna run hot, 1.5 watts thru a 3 watt resistor, which = about a 3 mississippi.

so we lump it together with more cooling material. it will keep the mic warm in winter,

you must keep the mic on 24/7 if you want a good noise floor, 

all we need is a honkin big transformer and the world is ours,

 

[CJ]

after holding up a twice stack of 31 UI lams to the mic frame,
we are gonna drop the stack down to

1.4 times a square stack

and reduce turns from

2600 to 2400.

this will make for a better fit in the mic body,

inductance reduce to 600 henries, still 4 times the stock value.

dcr  will be about the same as the core is bigger.

wind like utc a 26,  pri sec pri

here is the sheet, no dc to worry about,

 

[CJ]

note that R3, 29 ohm cathode resistor is there to get the cathode off the noise floor by 1 volt,

this is a grid leak circuit because of the 60 meg grid resistor.

so R3 is not a cathode bias resistor.

the capsule is the capacitor that blocks grid current from getting to ground,

cathode electrons get picked up by the grid and the 60 m slowly bleeds it off, but not before a negative charge on the grid gets set up.

R3 provides zero negative feedback or degeneration because the heater current is so much larger that a small change in plate current does not move the cathode around much,

circuit:

http://www.wagner-microphones.com/images/u47.gif

 

[bockaudio]

don't roll the output cap back too far to get rid of rumble unless you want a high-Q -3dB LF point.

 

[RuudNL]

Do I see it wrong, or did you solder the resistor directly to the membrane connection?
Can the membrane withstand the heat? Isn't there a risk of damaging it?
(Or didn't you use the 'Big Bertha' for it?...)