DIY RF Condenser Mics

[ricardo]

The BIG advantage of RF mikes is they are much less susceptible to humidity which is the bane of HiZ mikes. But modern Electrets go a long way to improving humidity resistance though still not as good as RF.
There was noise advantage in Jurassic times but today, LN FETs & zillion Gohm resistors are more readily available and there are HiZ mikes which rival and even beat the Sennheisers.
But what I want to know is the NOISE SPECTRUM of RF mikes. I should do an article on the various components of HiZ mike noise and how annoying each is. MicUlli did some detailed investigations (replacing zillion Gohms with diodes) but I'm suspicious of some of his findings.
I should really engage my single remaining brain cell & analyse Great Guru Baxandall's RF design but it's difficult to tear myself away from Beach Bum stuff .. en it mek mi head hut

 

[ricardo]

tube based RF mic seems like a very nice idea indeed ,maybe the Stephens circuit could be adapted or modified ,sub-miniturised ?

Was is das?

 

[kingkorg]

There seems to be some of the view that the Sennheiser MKH mic diaphragm is somehow floppy or loose , I can tell you by having looked at one thats not the case at all .

I can confirm this. I believe the confusion comes from the fact they have no damping from the backplates, which automatically makes them looser, but not necessarily loosely tensioned.

 

[Gerard]

Was is das?

This I guess: RF Condenser Microphone C2-OD4 Microphone/PU Stephens | Radiomuseum

 

[ricardo]

This I guess: RF Condenser Microphone C2-OD4 Microphone/PU Stephens | Radiomuseum

.. presumably someone has the circuit ..

 

[Tubetec]

There is a picture of the Rode rf mic PCB on their website , and yes the inductor they use looks very similar to Sennheiser , in fact the entire Rode circuit looks very like the Sennheiser .

Check out the details of Stephens circuit here Ricardo ,
https://archive.org/details/bbc-rd-reports-1955-24/page/n5/mode/2up
As far as the noise spectrum of the MKH416 , have a look on the SSL2 topic , you'll find an FFT plot of the mic through the interface , of course its impossible to prevent ambient noise but the undelying noise floor of the mic is visible in the high frequency range . I could try subbing in a capacitor in place of the capsule , but I'd rather not mess with it and stand the chance of hamming it up .

https://www.matidavid.com/electronic_components/03.htmjust some general info on inductor types here , clearly air cored coils are for very much higher frequencies where it says the pots cores are for frequencies upto 100khz.
Below is a close up of the inductors in the 416 , the SMD type is epoxy encapsulated 1:1 transformer . I'll get some baseline measurements with the DMM and LCR this afternoon .

 

[Tubetec]

I was able to measure the four coils on the main inductor although I have no schematic to refer to ,

coil 1
0.22ohms DC
0.208ohms @7.8khz
7.57uH

coil 2
0.14ohms dc
0.137ohms @1khz
3.44uH

coil 3
0.05ohms dc
0.035ohms @7.8khz
0.004uH

coil 4
0.1ohms
0.037 @7.8khz
0.13uH

Theres a few clues to the contruction of the inductor in the Baxandall paper , it says to be continued at the end of the article .is there part two ?

 

[Gerard]

I was able to measure the four coils on the main inductor although I have no schematic to refer to ,

coil 1
0.22ohms DC
0.208ohms @7.8khz
7.57uH

coil 2
0.14ohms dc
0.137ohms @1khz
3.44uH

coil 3
0.05ohms dc
0.035ohms @7.8khz
0.004uH

coil 4
0.1ohms
0.037 @7.8khz
0.13uH

Theres a few clues to the contruction of the inductor in the Baxandall paper , it says to be continued at the end of the article .is there part two ?

Yes. See message 408 (DIY RF Condenser Mics) in this thread for a link.

 

[Gerard]

.. presumably someone has the circuit ..

It seems the BBC analysed the circuit. See http://downloads.bbc.co.uk/rd/pubs/reports/1955-24.pdf. This appears to be the same report as that referenced by Tubetec in message 546 (DIY RF Condenser Mics), just with a link pointing to the BBC rather than to the Wayback Machine.

There is a little more in "Eargle's Microphone Book: Mono to Stereo to Surround: A Guide to Design and Application", pg. 437-438 (Ray A. Rayburn), including a very small circuit diagram on page 441. It includes the odd note that the cable between the mic head and the demodulator had to be an exact multiple of 37.5 inches, as it formed a transmission line.

There may be more to be found in the "Audio Cyclopedia" (Tremaine, 1969, 2nd ed), referenced in the attached figure; but I don't have a copy.

 

[rogs]

I was able to measure the four coils on the main inductor although I have no schematic to refer to ,

coil 1
0.22ohms DC
0.208ohms @7.8khz
7.57uH

coil 2
0.14ohms dc
0.137ohms @1khz
3.44uH

coil 3
0.05ohms dc
0.035ohms @7.8khz
0.004uH

coil 4
0.1ohms
0.037 @7.8khz
0.13uH

Theres a few clues to the contruction of the inductor in the Baxandall paper , it says to be continued at the end of the article .is there part two ?

As Gerard says, both parts of Baxandall's Wireless World's paper are in the copy I posted here : https://www.jp137.com/lts/Baxandall.RF.mic.pdf
I'm not sure how much of Baxandall's concept apply directly to Sennheiser's version though?...
There doesn't seem to be much correlation, when comparing Baxandall's schematic with Sennheiser's schematics for the MKH20-70 mics.
(I'm assuming the MKH416 is of a similar construction? - although I've not seen a modern 416 schematic...)
I've attached a couple of screenshots taken from the Rode NTG5 promotion video... It seems that their inductors are all contained within the same ferrite core -- and with 9 terminations -- so it could well be similar to Sennheiser's concept?

I'm not sure why you've included the audio frequency references in your coil measurements?
The RF circuitry employed by Sennheiser (and by me -- and possibly by Rode as well ? ) all tend to be operational with an RF oscillator frequency around 8MHz. I'm not sure quite how your audio frequency references apply here?...

 

[Gerard]

Hmm. Just found a scanned copy of the Audio Cyclopedia on Scribd at . It includes also the schematic for the Schoeps CMT RF mic at page 180, the Stephens RF mic at page 182 (not much better quality image than my scan from Eargles), and for some Sennheiser MKH mics at page 183. However, all circuits are without component values.

 

[Tubetec]

https://www.jp137.com/lts/Baxandall.RF.mic.pdfpart 1
https://worldradiohistory.com/UK/Wireless-World/60s/Wireless-World-1963-12.pdfpart 2
https://www.jp137.com/lts/RF.AMX2.Documents.pdfRogs RF mic

 

[Tubetec]

The frequency refers to testing on the LCR meter , DC readings are calibrated DMM
I have the choice of 100,1k or 7.8k on the LCR , some frequencies gave unstable results on the various inductances so I chose the ones that gave the best readings
Seems like coil 2 is the centre tap in a balanced winding , the bifi wound one ,

I see lots of similarties between the Baxandall and the Sennheiser , they both have a similar transformer structure , with 4 coils
In your design Rogs , your missing the coupling as its two seperate transformers your using ,
I think the other thing is , what were doing here is in fact making a filter , to get rid of the carrier frequency , hence the use of the ferrite pot cores destined to be most effective at around 100khz ,
or am I wrong on this ?

There are some small pot cores 18x10mm on ebay , but no bobbins appear to be included , the other thing is there not the adjustable kind ,although they do have a hole through the centre , AL value compared to the Sennheiser is unknown .

Maybe theres a way I can estimate the AL value of the Sennheiser pot core in situ ,but I dont know how to do that sort of measurement .
I cant get the calipers around the magnet wires exiting the inductor , there just to neat and tidy and down in the recess of the case work and I dont want to damage anything , what I can say for certain is its a lot thicker than pickup wire more like 0.1mm diameter , the coils are all fractions of an ohm .

It seems from what I read the transistor is by far better doing the hf switching element of this circuit , I still want to dig into the Stephens idea more though ,

 

[rogs]

https://www.jp137.com/lts/Baxandall.RF.mic.pdfpart 1
https://worldradiohistory.com/UK/Wireless-World/60s/Wireless-World-1963-12.pdfpart 2
https://www.jp137.com/lts/RF.AMX2.Documents.pdfRogs RF mic

• Note that both parts of Baxandall's Wireless World notes are posted in the first link: https://www.jp137.com/lts/Baxandall.RF.mic.pdf
-- There is no need to download the second link just for the second part of the Baxandall notes (You might want something else from that copy of Wireless World of course! )

• That reference document of the early version RF.AMX2 of my project is superseded by the later AMX10 circuitry, which you can find on this page: DIY RF Microphone Schematic

I need to delete that earlier document to avoid confusion.....

 

[rogs]

The frequency refers to testing on the LCR meter , DC readings are calibrated DMM
I have the choice of 100,1k or 7.8k on the LCR , some frequencies gave unstable results on the various inductances so I chose the ones that gave the best readings
Seems like coil 2 is the centre tap in a balanced winding , the bifi wound one ,

I see lots of similarties between the Baxandall and the Sennheiser , they both have a similar transformer structure , with 4 coils
In your design Rogs , your missing the coupling as its two seperate transformers your using ,
I think the other thing is , what were doing here is in fact making a filter , to get rid of the carrier frequency , hence the use of the ferrite pot cores destined to be most effective at around 100khz ,
or am I wrong on this ?

There are some small pot cores 18x10mm on ebay , but no bobbins appear to be included , the other thing is there not the adjustable kind ,although they do have a hole through the centre , AL value compared to the Sennheiser is unknown .

Maybe theres a way I can estimate the AL value of the Sennheiser pot core in situ ,but I dont know how to do that sort of measurement .
I cant get the calipers around the magnet wires exiting the inductor , there just to neat and tidy and down in the recess of the case work and I dont want to damage anything , what I can say for certain is its a lot thicker than pickup wire more like 0.1mm diameter , the coils are all fractions of an ohm .

It seems from what I read the transistor is by far better doing the hf switching element of this circuit , I still want to dig into the Stephens idea more though ,

I'm not sure you're addressing the RF formats being used here in quite the correct way?...... All the RF circuitry is designed to work around the oscillator frequency, which is normally around 8MHz. The only filtering performed by the coil formers will be from the 'Q' of the coil assemblies enhancing the response at the resonant carrier frequcncy. The coils and their associated components do not operate within the audio range -- or even at the 100KHz you mentioned. All the inductive interactions will be at the RF carrier frequency which, as I say, is normally around 8MHz for the Sennheiser mics...

The variations of the capacitive value of the capsule created by audio inputs are used to modulate the RF carrier. .... Either using Phase Moduation ( the earlier Sennheiser way of doing things) or using Amplitude Modulation (AM).
My project, Baxandall's project - and the later MKH Sennheiser 'push pull' mics - all use AM.

It needs to be remembered that the changes in capsule capacitive values are tiny -- something like 0.001pF for normal speech at 30cm, according to Baxandall.
Translating these tiny signals into usable audio, without introducing too much noise is a real challenge. The optimisation of the 'Q' factors in the inductor assemblies play a big part in getting results that can compare to conventional mics.

Designing these inductor assemblies from scratch is not easy . I have enjoyed my experiments using the Spectrum coils (I decided they would be worth a try, having read Baxandall's notes) but, as I have mentioned before, I'm not an academic and rely on instinct and experience for much of my experiemental work.

I wish you luck with your inductor experiments --- I've discovered it can be both rewarding and frustrating working at RF frequencies, with the signal levels encountered in these mic projects...

 

[kingkorg]

@rogs i confirm another successful build, it definitely has to be BC549, won't work with 2n5551.

I do get some HF roll off which is absolutely acceptable and musical. I listened to your clips and i noticed RK12 sounding warmer than Røde mic. It should be actually the other way around RK12 is brighter than Røde capsules. Absolutely nothing wrong with that, most people prefer this.

Now on to trying to make the circuit work with Sennheiser capsule.

 

[Tubetec]

The Rode mic from what I can see also uses 7 transistors , and a very similar style pot core to Sennheisers .

I was intrigued by one reference to RF mics in the Baxandall papers from I think Philips in Holland , maybe Ruud might be able to find something about it




below is a pic of the MKH416 p48 same as the one I have .
I still cant believe the price I got these things for , was 40 euros for the two mics and a pile of bantam to xlr cables thrown in .

 

[Tubetec]

I remember back in the day we had a couple of Sennheiser MKH105's at the back of the mic closet in the studio I worked in , we didnt have any T-12 also known as A-B power ,so they remained unused ,
but still even now a very capable mic by anyones standards .
A 12 volt A23 type battery in an XLR barrel might be just the trick to bring these T powered mics back into service.

 

[rogs]

.....I was intrigued by one reference to RF mics in the Baxandall papers from I think Philips in Holland , maybe Ruud might be able to find something about it


View attachment 111815

Gerard addressed these notes -- and provided an English translation of some parts - in his post on this thread a couple of years ago.
See here: DIY RF Condenser Mics

 

[rogs]

The most challenging part of this type of project - of indeed probably any RF bias microphone project? - is the specification of the inductors.
The inductors originally specified for this project are the Spectrum Communications type 5u3HH .
Those inductors can perform quite well, but suffer from two potential drawbacks:

• They don't perform as well where the selected microphone capsule has a capacitive value smaller the 65pF
• If you are based outside the UK, the Spectrum shipping costs can make them quite expensive.

I have prepared some notes here : https://www.jp137.com/lts/AMX10.Inductors.pdf
which include both coil winding details, and an alternative cheap source of similar coil formers.

These notes should allow experimenters to create their own transformers -- either as copies of the Spectum devices, or - by altering the coil winding details - to make alternative transformers more suited to a specific project requirement.
Some examples of how that aspect might progress can be found in the posts after #43 in THIS THREAD