DIY RF Condenser Mics

[RuudNL]

The coils have their own screening can.

 

[Khron]

There's that, plus they're already inside the metallic body of the mic, which at least in theory, should be grounded...

 

[rogs]

I've posted some notes on the project prototype details here:

http://www.jp137.com/lts/RF.AMX2.Documents.pdf

EDIT: The document linked to above has been updated, as improvements have been made to the prototypes.

You can find the latest project notes and schematics here:http://www.amx.jp137.com/

As I mention in those notes, the project would become a lot easier to construct using a PCB -- so I look forward to seeing how Ruud gets on with the design layout he's listed in post # 18....

... but it's been fun trying out the prototype to prove the concept.....

 

[kingkorg]

A bit off topic, but what a lovely capsule!

I just wonder how they achieve different patterns.

https://cvp.com/pdf/sennheiser_mkh_series.pdf

EDIT:
And i'll answer my own question, they use two back to back capsules.

 

[RuudNL]

As soon as I have results, I will let you know.
I tried to keep the HF part as far as possible away from the LF stage.
Also there is some grounded copper between the oscillator and the LF emitter followers.
Let's wait and see!
Are there other forum members who are going to experiment with this circuit?

 

[rogs]

As soon as I have results, I will let you know.
I tried to keep the HF part as far as possible away from the LF stage.
Also there is some grounded copper between the oscillator and the LF emitter followers.
Let's wait and see!
Are there other forum members who are going to experiment with this circuit?

I look forward to seeing how you get on. It will certainly be interesting to see how your layout affects the  noise levels.......I'm pretty sure they should be better than my stripboard layout!

I'm not sure the oscillator circuit is optimised for low noise either...... I shall be very interested to see what you discover.

If things go well, will you be making the Gerber files available for those of us who would also like to try a  PCB version ?

 

[Khron]

I'd kinda lean towards doing a full groundplane-fill on the outside edges of the board (including the screw hole pads), instead of that thin (double) loop.

Maybe the 100nf in the middle could go parallel to the two 100K's slightly to the right? And the 10k near the bottom, 180deg rotated and parallel to the diode?

Swap positions between the two 47R and 47uH's, and the resistors can then go more towards the left, resulting in shorter total trace length?

I'm afraid my "PCB OCD" is showing, i'm sorry... ;D

Great stuff!
I can't wait to experiment with this circuit myself.
(Already ordered the inductors...)
Added is a first PCB layout I am going to use for testing.

 

[RuudNL]

It is the first try, there will be plenty of room for improvement!
First, let's see if and how it works.
I think a PCB has an advantage over stripboard because you have less parasitic capacity.
10 MHz isn't really high frequency. (I have built circuits that worked on hundreds of MHz.)
But...time will tell!

 

[Khron]

I know, i know, it's kinda like mixing - sometimes you need to take some time off, and come back and see it with new eyes
That definitely often happens with my board designs ;D And/or a "second opinion" can have similar value.

Nothing wrong with getting closer to getting it right(?) the first time though.... Right?

But i'm also on the list of people looking forward to the result of these experiments :

 

[rogs]

It is the first try, there will be plenty of room for improvement!
First, let's see if and how it works.
I think a PCB has an advantage over stripboard because you have less parasitic capacity.
10 MHz isn't really high frequency. (I have built circuits that worked on hundreds of MHz.)
But...time will tell!

I'm thinking it might be an idea to try out a 12 MHz oscillator, rather than 10MHz?.....

That would permit the omission of  the 22pF cap C3 which - being in parallel with the bridge capacitors -  will tend to attenuate  any capacitive changes from the capsule ...

Operating at 12MHz would also mean changing C4 to 68pF and C8 to 33pF, but that should (hopefully) allow the overall Q to be largely maintained, as the inductors will still be tuned near the centre of their specified range.....

I don't have a 12MHz crystal to hand at the moment to check the idea out, but it may help to improve the sensitivity a little...so maybe worth a try?....

 

[Gerard]

Hi Rogs, Ruud, For a simpleton like me - it's been too many years since I last made up a mic circuit - are there any special considerations in selection of components? Any of the caps that should be a particular type (C0G/NP0 or polystyrene or whatever)? And for the crystals, any recommendations on series resistance or load capacitance or any other consideration? And any special requirements for the inductors L1/L2.

Also, would you, Ruud, be willing to make some PCBs for sale? I no longer have the facility to make them myself.

 

[rogs]

I built the prototype with 'standard' stock parts I already had in my  hobby electronics kit ...So..
• Vishay K series MLCC with class 1 C0G  diecletric for C1,2,3,4,and 8
•Multicomp polyester 63V  MCPBSFC  series for C5,6,9,10,11, and 12
• Crystal is an IQD LF A143K -- (HC49  -30pf - parallel resonant)
• Inductors are type 5u3H  (10mm coils) from Spectrum Communications here in the UK.  (I notice the Ebay stock level of these devices is now 0, so you may need to buy from Spectrum directly - see here:  http://www.spectrumcomms.co.uk/Components.htm

 

[RuudNL]

First I will do some experiments myself.
I have crystals for 8, 10 and 12 MHz available. (Sennheiser used an 8 MHz crystal in their MKHmicrophones.)
If the result is successful, I can publish the Gerber files. (At https://jlcpcb.com you can get 10 PCB's for $2.)

For those who can't wait: here are the Gerber files I used for the prototype. (Rename .JPG to .ZIP!)

 

[Gerard]

I have already ordered some coils. The eBay vendor says he has thousands of them! But for bulk orders to contact his website.

My question concerning any specific requirements for L1/L2 actually referred to the two inductors at the connection to J1, the XLR connector.

 

[Gerard]

Ruud, thanks for posting the Gerber files. I'll see if I can get some boards made here in Ireland.

 

[Khron]

I'm quite sure those are only against (conducted) EMI / RFI, so they don't need to be anything special.

My question concerning any specific requirements for L1/L2 actually referred to the two inductors at the connection to J1, the XLR connector.

 

[RuudNL]

Ruud, thanks for posting the Gerber files. I'll see if I can get some boards made here in Ireland.

The PCB fits this microphone body: https://www.wgtcenter.com/microphone-accessories/shell/diy-accessories-for-u87-microphone-condenser-classic-neumann-vintage-body.html

Type "B" (the smaller one.)

 

[rogs]

Sennheiser used an 8 MHz crystal in their MKHmicrophones....

That was my starting frequency too, but  - unlike Sennheiser - I don't have the expertise to design and build custom inductors! 

The Spectrum 5u3H has so far been the most useful 'off the shelf' IF can  I have found, so the oscillator frequency needed to be  adapted to match those, as far as possible.

I've had quite useful results at 10MHz, but - if my calculations are right - it should be possible to improve the sensitivity by  some 6dB  by using  12MHz, and the revised capacitor values?
I shall be interested to see what you discover....

Thanks for the preliminary Gerber files...

 

[rogs]

My question concerning any specific requirements for L1/L2 actually referred to the two inductors at the connection to J1, the XLR connector.

The 47uH inductors were fitted because that's what I had to hand.....The value is not critical -- anywhere from 47uH to 220uH should do. 
The potential dividers formed with C11 and C12 will attentuate any conducted stray RF pretty effectively...
The dominate source of any extra stray RF is likely to be ground borne - or possibly radiated -  and  hopefully not a significant amount.

As I stated in my notes, this is not a finished design, but rather a 'proof of concept' ,  to hopefully allow a practical design to be developed.  Ruud has already taken up the mantle! 

I think we've probably already advanced further in these few posts than the contributors did in the  70+ posts in the Yahoo thread you started a couple of years ago? 

 

[RuudNL]

The first thing I would like to have a closer look at is the biasing of the oscillator transistor.
Assuming the voltage over C6 is 12 Volts, the base of Q1 is at 10.9 V.
This would mean a DC voltage of ~10.3 V at the emitter of Q1.
This DC voltage theoretically causes a DC current of 10.3/470 = 21.9 mA.
But since the microphone is phantom powered, this current can't be delivered.
I would suggest a bias of Q1 in such a way that the voltage at the emitter is in the order of ~ 1/2 * U C6.
Maybe it would be a good thing to add a capacitor in series with R5, so there is only AC over T1.
(Any DC current through T1 is a waste and because T1 is tuned to resonance, the impedance as seen by the oscillator should be pretty high.)