M4/8/7 - U47&48 style circuit / switched three polar patterns / single 6028

[ln76d]

Nothing new since 1947.
I like to draw schematics, i have few 6028 tubes so why not
It's proposition of u47/48 based circuit for different microphones bodies, not u47 clone.
Different tube, different transformer, different headbasket (size, mesh), chinese capsule so it wouldn't be U47 anyway.
Only making a good microphone based on a good circuit
In general a cheap build.
To use the circuit in a small body, the best way is to use small output transformer and direct heating from 20V PSU, to remove 1,8k "warmer".
First i had in plan to build it in blue baby bottle style body, but right now i think about MXL 770.
Transformer which i want to try is 1:7 Sennheiser.
Since 6028/408a tubes are small it can fit many bodies.

Schematic updated, 3P3T switch topology was mistake.
Right now is DP3T, removed one 100M resistor and additional filtering 10nF capacitor for figure of eight pattern.
Since plate/polarisation voltage isn't common with filament,  most of filtering can be done in PSU.

Right now am thinking about new switching schematic for MXL770 body. Since there are two cutouts for switches, maybe i will use two switches for polar pattern change
Then it can use additional filtering inside microphone.
We will see.

 

[ln76d]

R4 - R6  - all 100Mohm, for easier sourcing.
R6 was also 100M in some versions of original U47.
R5 shouldn't affect  figure eight pattern.
Heaters direct from 18V supply - to remove dropping voltage resistor, which should be "big" enough to dissipate heat.
For low end extend C4 2uF. I would leave C5 0.5uF anyway.
Other tweaks (if some will be necessary) for use 6028 i will post after building real circuit and making measurements.
When?  I don't know.
Someday i will post also PSU schematic

After i make it i will try to design (or redesign) some circuit to use 6028/408a in SDC microphones.
One of the biggest advantages of these tubes is that those are subminiatures!
Why only use it for boring "cloning" U47 ;P

R1 2Mohm
R2 3Mohm
R3 1Mohm
R4,  R5, R6 100Mohm
R7 29ohm
R8 100kohm
R9 430ohm (440ohm accurate for 18V supply)
R10 30kohm
C1 1nF
C2 10nF
C3 2uF
C4 0.5uF
S1 - DP3T  ON - ON -ON type

 

[Khron]

Not that i have any technical or moral objections, but... Isn't there some way to achieve the pattern-switching with an SPDT or DPDT switch?

I would imagine 3P3T switches are almost as common and easy to find as hens' teeth...

Wouldn't something like what's in the CAD Trion 8000 work, perhaps?
http://recordinghacks.com/pdf/cad/trion-8000-schematic.pdf

Also a tube mic, also using a resistive divider to obtain the polarisation voltage(s), and a simple SPDT center-off is all it needs

I'm just not a fan of complicating things more than needed, or "esoteric" components ;D Just a thought...

 

[ln76d]

Not that i have any technical or moral objections, but... Isn't there some way to achieve the pattern-switching with an SPDT or DPDT switch?

I would imagine 3P3T switches are almost as common and easy to find as hens' teeth...

Wouldn't something like what's in the CAD Trion 8000 work, perhaps?
http://recordinghacks.com/pdf/cad/trion-8000-schematic.pdf

Also a tube mic, also using a resistive divider to obtain the polarisation voltage(s), and a simple SPDT center-off is all it needs

I'm just not a fan of complicating things more than needed, or "esoteric" components ;D Just a thought...

There's nothing esoteric here.
3p3t are easy to find in toggle and slide versions.

Main idea was use u47 and u48 topologies not to make something "new" (yes, sounds funny here ).
I didn't found over the net that before anyone use three polar patterns in this circuit so maybe this is something "new"

Why that way?
Because there's difference between U47 and U48.
If you look on schematics of both mikes, cardiod sensitivity is different due to lower polarisation voltage in U48.
U47 - ca. 63V
U48 - ca. 53V
Also there is additional filtering for figure of eight pattern.
Using 3p3t switch allows to keep in one mike both topologies - sensitivity for cardioid/omni like in U47 and sensitivity for figure of eight like in U48.

Using Omnitronic/CAD/T.Bone/Great Hon (and how the hell they will call it) topology you will had to use lower polarisation for cardioid/omni (without tweaking it). Also look for differences in resistors values. I think that the capsule "prefer" higher impedances.

There's nothing complicated and to be honest i prefer to take inspiration from Neumann schematics than from Omnitronic

 

[Khron]

So that's an 18% difference in bias voltage, which ends up translating into a sensitivity difference of... 1.43dB, if my math isn't terribly "broken"?

 

[RuudNL]

I didn't do any calculations but... isn't the 27R resistor a bit low for the correct -Vg?

 

[ln76d]

So that's an 18% difference in bias voltage, which ends up translating into a sensitivity difference of... 1.43dB, if my math isn't terribly "broken"?

You can always post your equations here
It's based on informations over the net.
Some owners of U48 with interchangable "capsule head" 47/48 noticed that there's audible difference in cardioid "sound", when using U47 head in that circuit and comparing to the normal U47.
Please remember, that with 10V of difference in polarisation voltage, diaphragm stiffness would be different which affect response in different way. 
Sensitivity would be different for several frequencies, it wouldn't be "linear" drop i think.

 

[Khron]

The diaphragm tension thing is indeed true, good point.

Not sure if it was you, or someone else who said they liked the Rode(?) or some other chinese K67's with 70-80V on it - kinda makes sense, if it pushes that resonant peak higher in frequency (and makes it less obnoxious in the process).

 

[ln76d]

I stated somwhere that i prefer RODE NT1a with K47 and 75V than K67 and 75V - or something like this
My last statement is that for k67 in NT1a i prefer alctron 32mm than rode version
But for that, circuit need to be modified a little. Other thing is to use 75V in schoeps style circuit as we talk about in other thread.
In fact from that time i changed my "preferences" little and right now am using 65V in NT1a microphones.
For many capsules (especially RODE center terminated) this voltage is too high.
In some (like mentioned rode capsule) i do a little modification inside the capsule.

In my molested G7 am using K47 with 50,5V polaristion voltage.
I tried from 80V to 45V and 50,5V fits the best for me.
In different circuits am using 48V - 75V
There's also a little difference between double sided and single sided chinese K47!
For higher voltages, single sided sounds a little bit better in cardioid than double sided.
All depends on the circuit and what for microphone is.

Also there's possibility that i will go down with the polaristion voltage if the chinese K47 will sound better in circuit from this topic

 

[ln76d]

In

I didn't do any calculations but... isn't the 27R resistor a bit low for the correct -Vg?

RuudNL  and that's why you are my hero!!!

I drew schematic without thinking about how the circuit works.
So  tube bias had no chance to proper work. To use that way as in first versions of schematic it should be larger value of cathode resistor, so there could be bypass capacitor and all togheter it change too much.
When i build "something" and do some tests i will correct schematic,
Right schematic is updated - extra resistor from filament as Oliver Archut did on his "U47 Alternate Tube Schematic"
http://www.tab-funkenwerk.com/sitebuildercontent/sitebuilderpictures/U47ArchutAltTubewithRemote.jpg
Filament changed to 18V for a little underheating.
Added to the schematic  R9 resistor for 1.1V at cathode, using Oilver method.

Again Mr. Rudolph - thank you

 

[ln76d]

Most of the parts which will be used for this project.
Only need to buy a switch and wiating for 100M resistors.
Drilling boards when i get the switch.
Since there's no place for MP output capacitor i choosed Siemens MKL 0,5uF (measured 0,65uF) from 1962.
Rated for 60V -  i hope it wouldn't explode
For R9  - 470R and 3.9K in parallel - 420R and cathode R7 27.5R - to get 1.1V will be heater adjustement (with little swing - unregulated supply).
From the start C3 2uF -  I'm too lazy to check and it wouldn't hurt

 

[gary o]

Im no expert but yr 60Volt cap sounds bit low..... hope doesnt blow up too.....\i built a circuit very similar to yrs Im not sure where I got it from tho all the best

 

[ln76d]

Im no expert but yr 60Volt cap sounds bit low..... hope doesnt blow up too.....\i built a circuit very similar to yrs Im not sure where I got it from tho all the best

For unregulated supply there could be ca. 100V for less than 1s
Then it should be ca. 34V so it's safe.
In some tube mikes original were low voltage rated caps.

Here are some missing parts on previous picture.
If there will be a problem with fitting transformer i will remove the can

Btw one inner mesh is removed, if the second one will affect the sound i will remove it or replace with something better.
Unfortunately MXL, painted, fine mesh affect sound in not good way.

 

[ln76d]

A little progress
Something i could do without switch and 100M resistors.
I had to pull  transformer from the can.
It's so tiny that will fit into square housing on the picture.
More space for the parts
Unsoldered connector is for R10 30K.
For a start it will be potentiometer to adjust proper plate voltage, then i will replace it with proper resistor.
I think that there will be a little drop in resistance, so that's why C3 will be 2uF from the beginning.
Tube will be cooling a little by the xlr base.

Project abandoned , due too many reasons!
Most important 6028/408a tube doesn't fit 6:1 or 7:1 transformer - if "telefunken" usa really use it in their "U47"...

New circuit with single 408a you can find here:
http://groupdiy.com/index.php?topic=62423.0