DIY U-47 from Cascade M20U

[CJ]

6 micron, no biggy, 

put some foil over the capsule for sure,

 

[squib]

oh... I see your capsule has dust on it. Maybe thats where the noise came from. Best to give it a nice wash.

 

[CJ]

have you ever seen u47 capsule from 1951?

i am thinking of soldering the tube directly to the capsule lead,

this would make the grid wire very short.

 

[CJ]

just search "5678 images"

and you get circuit examples,

it looks like the 5678 is a perfect fit, likes to run grid leak, likes 45 volt supply, likes to be used in german amplifiers, jus sayin...

so we use around 270 k plate resistor, but allow for 30 k pi filter rc network,

so maybe 220K for starters,

this circuit looks noisy, back to back pentodes,

we have tele ef802 for backup in case this thing hisses like a snake, 

 

[bockaudio]

have you ever seen u47 capsule from 1951?

i am thinking of soldering the tube directly to the capsule lead,

this would make the grid wire very short.

Well, that k47willbe an underperformer there with it's lack of gold.
Amp input directly behind capsule IS an idea used by N-m-nn on some of their mics thesedays.....BUT- look at the C12 and some of AMI's work with the tube miles away from the capsule. No noise complaints there!

 

[bockaudio]

I should have been more specific: N*u*ann uses the amp directly behind the capsule in at least the M150, which then feeds into the tube amp a few inches below the capsule. Great idea to reduce noise in a small diaphragm environment where noise is an issue, where the mic will be used on an orchestra where the noise would be exposed during the quiet passages.

 

[CJ]

roll your own,

there goes a #40 hand wound wide pitch for low stray C , 600 turn pri for the 5678 tube output transformer,

then 200 turns of say, #36 or #38 for the sec,

then 600 more turns, so 2400:400 turns total,

twice that on some old school formers from a dissection, plus some pre soaked pcb wrappers  ;D and  we can roll one for the road, over...

 

[CJ]

we wound pri sec pri with #40, #36 and #36 so we get 292 ohms dcr for the primary and 33 ohms dcr on the sec.

looks like it will fit sideways, so we can name this the UTC47 if it works,

 

[CJ]

pump it up with 700 henries with the super q core,

44 k ohms pri z at 10 hz,

38 lams = 1/2 inch stack,

can only handle 7 volts primary  19db@10hz,  but the capsule can only generate how much into the 5678 tube?

 

[gemini86]

CJ, would love to see a pic of your winding setup. Do you have a machine with cam actuated side to side feed? or do you do it all by hand?

I have a pile of lams and a couple spools of copper just waiting for my lazy ass to build a winding rig.

 

[CJ]

here it is,

i have a stepper traverse, but it only has one speed,

so i was gonna divide the pulse output with 4790 chips so i can wind fine pitch,

but i just figured out a plan, instead of the logic chips, just put the traverse on a rotating platform, if you need fine pitch you would rotate the traverse up to 45 degrees,

then maybe have a divide by 2 chip and you would have two ranges, each micro adjustable for different gauges,

i just hooked up the mic to 50 volts to check the amp and it works!

no transformer installed yet,

but i get 35 ma heater current at 0.8 volts instead of 1.25, ind of like the vf14, starving the heater and the plate a bit, this tube warms up in about 10 nano seconds,

i have a 150k plate resistor that provides 30 volts to the plate and the capsule is sitting at 34 volts, 4 volts higher than stock so right in the ball park,

i got 0.8 volts ac by talking loudly into the condenser, so figure 120 mv after then transformer,  (6:1 ratio)
cant wait to wire it up,

 

[gemini86]

thanks cj!

 

[CJ]

no problem.

ok this tube starts full conduction at about 25 ma of heater current,

we are running about 35 ma, the cathode low side is about 1 volt, cathode hi side is about 2 volts, so we are running 1 volt out of 1.25 for heating.

plate current with the 150 k plate resistor is 120 microamps
plate voltage is 30 volts,

all the voltages are like stock u47 except we re running 120 ua instead of 900 ua,

so heater and plate current are being run low, wwe should have less noise with the drop in plate current over the stock u47, the transformer should work with the lower current as we have jacked up the Z by a factor of 4, and the lams are low loss so miniscule exciting current needed,

 

[grantlack]

look at the C12 and some of AMI's work with the tube miles away from the capsule. No noise complaints there!

i was less astonished by the blackspade um17r demo than the "golden ears" gents in the room, but would agree that noise was far from problematic.  as i recall, the short glance i snuck under the hood looked like a pretty typical submini circuit w/ the tube positioned further toward the connector side.  anyway, i digress... cj gettin' funky on the 1's and 2's!!!

 

[CJ]

this mic sounds gorgeous!

dead quiet hiss wise, i di have a little 60 cycle but i am using an HP dc supply to power it so there might be some grounding issues between the preamp and the pwr supply.


great midrange, no low end rumble, seems real flat, great output,

the tube does ring a bit when the mic is set down on the bench,

i think this is normal, remember lennon twapping the u47 with his finger and the engineer ripping his phones of due to the ringing?

ok, we will bolt this thing up, i soldered the screen on already so that i know is well grounded, the 2 small screws seemed wimpy for such an important ground,

oh wow, have not laced up the xfmr with i bars yet, so much more inductance on the way! 

 

[pasarski]

...dead quiet hiss wise...

Was waiting for this info, thanks! Now shopping..

It may sound weird, but I love the look of that tube, and every other component in your mic too. The home "rolled" tranny is awesome!!

 

[tchgtr]

Thanks CJ. I always enjoyed your transformer deconstruction series of posts, now we have a full-fledged reconstruction in improvised form. Great chops!

 

[gemini86]

Great, now I'm going to have to tear my g7 apart and redo everything. I've learned a ton from every one of your threads, cj. Many thanks.

 

[CJ]

thanks for the props you guys,  ;D

i never built a G7, it might be better than this,

keep the G7 and build another mic,

i wanna try analag's 6ba6 mic,

how big is the M7 capsule?

is this cascade a medium or large?

tons of 5678 on evilbay,

here are 160 of them for $149.95,

http://www.ebay.com/itm/Large-Lot-of-160-Tung-sol-Raytheon-5678-Subminiature-Tubes-from-Bulk-/261069942217?pt=Vintage_Electronics_R2&hash=item3cc8fb2dc9

you could weed them out just like the vf14,

only 1 dollah instead of 5,000 dollah,


could try pentode but why> plenty of gain already and more gain = more noise,


plate resistance in troide for the 5678 at 30 volts Vp worked out to 250,000 ohms,
they do not have this data on the cut sheet, used a 150k plate resistor,

i might rewind the xfmr and put in some copper foil between the windings for heat dissipation, 

i will do a schematic on this thing,  a better layout could be done, i did not plan this out on graph paper like a good student, just started wiring in parts, a pc board would be diy friendly as the part count is low,the tube does not glow unless you overamp the current by about 5 ma, then you can see it thru the shield,

 

[CJ]

ok  one schematic, hold the french fries,

Rev 2 schematic with R1 changed from 680k to 100k