K47 capsule with EF86

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Joined
Oct 10, 2012
Messages
11
Hi everyone!
I have one idea, to connect k47 type capsule (Chinese made, single diphragm ) to G7 output stage as on the picture.
How do you think, will it work well? Or is it wrong way?
 

Attachments

  • IMG_2668.JPG
    IMG_2668.JPG
    1.5 MB · Views: 68
Thats quite similar to an EF86 based mic I built, back plate connected to polarising voltage ,capsule front direct to grid like the U47. 1gigaohm grid resistor does seem a bit high ,my guess is you'd need to very carefully select your ef86 to work with that value .U67 uses a similar value but ef86's dont live all that long in them in my experience . 40 megohms is the listed as the highest grid resistor for the ef86 I believe ,I ended up using an 80 or 120 meg resistor in mine .Worth using a ptfe tube socket in this application also if you can get it .
 
Tubetec said:
Thats quite similar to an EF86 based mic I built, back plate connected to polarising voltage ,capsule front direct to grid like the U47. 1gigaohm grid resistor does seem a bit high ,my guess is you'd need to very carefully select your ef86 to work with that value .U67 uses a similar value but ef86's dont live all that long in them in my experience . 40 megohms is the listed as the highest grid resistor for the ef86 I believe ,I ended up using an 80 or 120 meg resistor in mine .Worth using a ptfe tube socket in this application also if you can get it .

Thank you so much!
So ..hope I'll get results:)
And ..
If I'll get double diaphragms capsule, can i have omni pattern like u47?
 

Attachments

  • IMG_2670.JPG
    IMG_2670.JPG
    1.7 MB · Views: 36
Here you have two options of EF86 mic which fit k47 capsule

https://groupdiy.com/index.php?topic=64687.msg819492#msg819492
https://groupdiy.com/index.php?topic=62441.0

S316 use psu switch to change polarisation voltage, you can skip it - for k47 c3 isn't needed.
High impedance resistors from 100M to 470M.

For your schematic - don't need to install two plate resistors and filtering cap - there's marginal benefit from it, but the phase is much better without it.
For cathode - G7 arrangement was experimental to use feedback. Without any feedback this arrangement doesn't work good.
Choose between one cap and one resistor or just one  resistor.
Output capacitor 0.5uF is enough and there's positive audible benefit from using lower value. 2.2uF is totally to high. In both cases low end level is the same. Proper cathode biasing will give proper low end level ;)
Polarisation voltage - you need to use divider! Currently your capsule will die quickly with 160V. Max what you can use is 3M/2M (this side to ground) and add high impedance resistor + filtering cap.
Hope this help somehow ;)
 
I learned from Oliver Archut to use a trimpot to dial in cathode bias exactly. He did this on CS1 and CS4.
 
AusTex64 said:
I learned from Oliver Archut to use a trimpot to dial in cathode bias exactly. He did this on CS1 and CS4.

Yeap, that's the way to do that (if we use resistor to bias) ;)
 
ln76d said:
Here you have two options of EF86 mic which fit k47 capsule

https://groupdiy.com/index.php?topic=64687.msg819492#msg819492
https://groupdiy.com/index.php?topic=62441.0

S316 use psu switch to change polarisation voltage, you can skip it - for k47 c3 isn't needed.
High impedance resistors from 100M to 470M.

For your schematic - don't need to install two plate resistors and filtering cap - there's marginal benefit from it, but the phase is much better without it.
For cathode - G7 arrangement was experimental to use feedback. Without any feedback this arrangement doesn't work good.
Choose between one cap and one resistor or just one  resistor.
Output capacitor 0.5uF is enough and there's positive audible benefit from using lower value. 2.2uF is totally to high. In both cases low end level is the same. Proper cathode biasing will give proper low end level ;)
Polarisation voltage - you need to use divider! Currently your capsule will die quickly with 160V. Max what you can use is 3M/2M (this side to ground) and add high impedance resistor + filtering cap.
Hope this help somehow ;)

Thank you a lot!
I done test layout. Circuit is working.
But I can't understand one thing..
After divider 3m3  and 1M series resistor i got 75v. But after 100M in the backplate point only 1v..
so.. do I need to use 100M ?
 
Your multimeter has an input impedance of 1M, or maaaaaybe 10M if you're lucky. Together with that 100M, that forms a (huge) voltage divider ;) That's why you're measuring next-to-nothing "after" it.

PS: For whatever reason, i can't see the attachment. But if you did that 3.3meg / 1meg divider from 160v, unless my math is seriously wrong, that would only result in about 37.  ???

160v x 1meg / 4.3meg (total) = 160v * 0.23 = 37.2

 
Khron said:
Your multimeter has an input impedance of 1M, or maaaaaybe 10M if you're lucky. Together with that 100M, that forms a (huge) voltage divider ;) That's why you're measuring next-to-nothing "after" it.

PS: For whatever reason, i can't see the attachment. But if you did that 3.3meg / 1meg divider from 160v, unless my math is seriously wrong, that would only result in about 37.  ???

160v x 1meg / 4.3meg (total) = 160v * 0.23 = 37.2

Sorry, reload attachment
Is this correct circuit, or need change something?
 

Attachments

  • IMG_2715.PNG
    IMG_2715.PNG
    201.6 KB · Views: 30
Umm..... In short, yes, changes ARE required.

First of all: https://en.wikipedia.org/wiki/Voltage_divider

(Hint: the "Vout" from the above link, goes to the backplate)

https://groupdiy.com/index.php?action=dlattach;topic=62441.0;attach=46826

Looking at the capsule side only, in your case, your R3 (the "top" resistor of the voltage divider) would be the 3.3meg, but to get about 60v at the capsule, your R6 ("bottom" resistor) would need to be around 1.2-1.3meg. Using (more) standard values,  2.7meg and 1meg would get you there. That is assuming a 160v "starting" voltage.

With 137v or so, you miiiiiiiight get away with a 1:1 resistor ratio (say, two 3.3meg ones), which would give you about 66-67V or so. A bit on the high side, but workable, i think. Maybe if you bumped up your R5 from 10k to 12-15k, that would bring the "B+" (137v) down a few volts, and with it, the capsule bias voltage as well.
 
Thanks everybody for clarification and inspiration! Project is almost done!
I have a question ..
For cathode bias adjustment i use 5k trimmer parallel 100u cap. During my experiments without cap i got the same sound just with little less gain.. so what reason to use this cap?
 

Attachments

  • image.jpg
    image.jpg
    1.5 MB · Views: 25
Hi,
Using the capacitor bypass the 'natural' feedback that occurs (sometimes called degeneration feedback) when you put a resistor in series with cathode.
The cap indeed increase gain, but raise distortion too and cathode being low impedance 'the sound' of capacitor is audible (one of the reason you sometimes see some film cap 1/100 the value of electrolitycs parraleled).
If you like the sound without cap keep it like this... or add a jumper to have both options.

A little bit of theory (easy to understand) about unbypassed and bypassed grounded cathode amplifier:

https://www.google.fr/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwiopqa6kcPXAhVHwBQKHUleAn8QFggnMAA&url=http%3A%2F%2Fken-gilbert.com%2Fimages%2Fpdf%2Facpart1.pdf&usg=AOvVaw2wR3eUwBpksZ-uIGXpSdpL

Edit: if link don t work google 'Crowhurst Audio Classroom part1'
 
For chinese K47 remove C1 and R2 - you will get better response.

In case of cathode cap - with EF86 you are able to get decent level without it, so try both options but without the cap you need to re-bias circuit to get similar level in comparison. Another thing to try is wiring output transformer.  Especially for unbypassed cathode - negative node connection - cathode vs. ground (you can use simple switch to make comparison).
 
Back
Top