Adding RF filters to mic builds

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Delta Sigma

Well-known member
GDIY Supporter
Joined
Dec 4, 2010
Messages
616
Location
Winnipeg
A friend moved into a new studio that has a history of pickup a radio station with certain gear, especially vintage mics, in one section of the studio. I lent him my EF800 U47 build and it picked up the station pretty clear.

Would a simple filter like C12/C13 in a U67 help?
http://recordinghacks.com/microphones/Neumann/U-67

Anyone have better examples of RF filters in mics?
 

Attachments

  • U67 RF Caps.jpg
    U67 RF Caps.jpg
    27.8 KB · Views: 44
If you look close at the photos of the u67 the caps are feed though caps. One side of the cap is the metal body

You would want to add them at the B+, heater, pattern if present and ground as well.

If you do try ceramic or mica at the XLR(or what ever connector you are using) you will want solder one lead of the cap at the pins and attach the other lead as close as possible to the body XLR(or what ever connector you are using) area with the shortest lead length.

You could try ferrite beads for each of the wires

How long is the wire from the ground of the XLR(or what ever connector you are using)? to the body
 
Thanks Gus!

I'll have to crack it open when I get it back. If I remember right, I grounded the connector immediately. I used one of Dany's bodies and it's threaded for the heat sink resistor for the parallel 408a or VF14 circuit. Pretty sure I grounded there.

I'm putting in a Mouser order so I'll take your advice and see if it makes a difference. Fingers crossed.

The owner of the studio has a U87 that is bad for picking up the station. I'm guessing it has the XLR ground going to the PCB before the mic body.
 
Is the radio station AM, FM, TV, or some other service?  Interference control puts bread on my table.  Do you know the frequency of the radio station?  You should, because that determines in part how to reduce or eliminate the interference.  How far is your studio from the radio station transmitter?  Do you know the call sign of the radio station?  If you do, you can look up its frequency, power, etc. on line at www.fcc.gov using their AMQUERY, FMQUERY, TVQUERY, etc.  Be aware that FCC regs only protect receivers; they don't protect non-receivers, audio and recording gear, etc. from "blanketing interference" from a nearby transmitter.

If your studio is near a radio station and your equipment demodulates (rectifies) the incoming RF, you should first find the source and frequency.  This is the starting point for any RFI solution.

1.  Invest in some AC line filters.  I use the HUM-X ones.  (They can be ordered from B&H in NYC if unobtainable locally for around $60 each.)  Note they are rated for 6 amps so don't overload.  You should really filter AC to all of your studio gear.  Plugging a "clean" mic into a "dirty" preamp or console doesn't fix the problem!  You may want to try an "isolation transformer" on the incoming AC to your equipment.
2.  Housekeeping:  Are your mic cables in good shape (i.e. no floating shields, etc.)? No "ground loops" (AC or RF)?
3.  I have used a stock Neumann U67 around broadcast stations with no problem.  Some mic preamps ARE a problem.  Use a filter on the preamp AC plug also.  A good quality shielded 1:1 transformer between the mic and preamp should stop most RF.  RF does not like iron.  Try the transformer at the preamp end first.  Terminate the transformer secondary with a 200 ohm resistor.  Ferrite beads on the input of the preamp may help, check the schematic if this has been done.
4.  Some mic preamps "dump" the incoming audio from the XLR connector directly into solid state devices, which in turn rectify the RF from the nearby radio station, which you then hear clearly.  The input impedance of the preamp may be 1200 ohms or higher.  If you are using a U67 (or other tube mic), terminate the output (at the preamp input) with a 200 ohm resistor, so the mic output transformer secondary "sees" the correct load for its turns ratio, especially if using a "unknown" preamp.
5.  Are any cables or AC cord lengths approximately 1/4 wave length of the radio station frequency? Your cables can act as an antenna if resonant at the RF frequency.
6.  Resolution of "audio rectification" can be a multi-step procedure.  Start with housekeeping, then filter the AC, and work from there.  Avoid at all costs making modifications to the mic...or connecting capacitors across the mic output.  If you want to mess with filters, get some in line shielded "barrel" fittings such as the Switchcraft S3FM type and build in line filters...and LABEL THEM....

I made a remote recording at a chapel where a spectrum analysis revealed a defective MIDI box in a large pipe organ console, which was radiating a strong signal around 2.15 MHz and getting into everything.  The organ console was a fixed installation, and the wiring to the pipes and controllers acted like a large antenna.  I used multiple mics, an AC isolation transformer and multiple AC line filters, coiling a few turns of mic cable over ferrite core, the whole 9 meters, to get a suitable recording.  Moral of the story, other devices with clocking at RF frequencies can also trash a recording.
 
I am not familiar with Canada's Dept. of Transport (or the Canadian equivalent of the FCC).  I took a quick look at the AM station data for Winnipeg Manitoba CANADA; most of the AM stations there operate with a minimum of 10 kilowatts power full time and some are 50 kilowatts full time.  Many are directional (multi tower) arrays.  If that studio is near one of those transmitter sites, it could be receiving "the Royal RF treatment"... AM or other LF/MW interference can be resolved using correct procedures.
 
Great stuff rmburrow!

Unfortunately, I haven't visited the studio. Things are pretty locked down here with COVID. I have asked my buddy to figure out the station but he hasn't had the time to set the mic back up to listen.

Industry Canada (I think they're called something else now) handles such things in Canada. I found their site and told him to contact them but it's kinda pointless if he can't figure out the station.

I love your inline filter idea.

Funny thing, is that I actually work in telecom for the power utility up here. It's probably been more than ten years since I've done any real RF stuff though.
 
Your studio friend should be able to go out to his car or truck and check the radio there...the "front end" overload and audible distortion should be obvious....and "splatter" on the adjacent channels should also be obvious...and then check with the mic...
 
rmburrow said:
Is the radio station AM, FM, TV, or some other service?  Interference control puts bread on my table.  Do you know the frequency of the radio station?  You should, because that determines in part how to reduce or eliminate the interference.  How far is your studio from the radio station transmitter?  Do you know the call sign of the radio station?  If you do, you can look up its frequency, power, etc. on line at www.fcc.gov using their AMQUERY, FMQUERY, TVQUERY, etc.  Be aware that FCC regs only protect receivers; they don't protect non-receivers, audio and recording gear, etc. from "blanketing interference" from a nearby transmitter.

If your studio is near a radio station and your equipment demodulates (rectifies) the incoming RF, you should first find the source and frequency.  This is the starting point for any RFI solution.

1.  Invest in some AC line filters.  I use the HUM-X ones.  (They can be ordered from B&H in NYC if unobtainable locally for around $60 each.)  Note they are rated for 6 amps so don't overload.  You should really filter AC to all of your studio gear.  Plugging a "clean" mic into a "dirty" preamp or console doesn't fix the problem!  You may want to try an "isolation transformer" on the incoming AC to your equipment.
2.  Housekeeping:  Are your mic cables in good shape (i.e. no floating shields, etc.)? No "ground loops" (AC or RF)?
3.  I have used a stock Neumann U67 around broadcast stations with no problem.  Some mic preamps ARE a problem.  Use a filter on the preamp AC plug also.  A good quality shielded 1:1 transformer between the mic and preamp should stop most RF.  RF does not like iron.  Try the transformer at the preamp end first.  Terminate the transformer secondary with a 200 ohm resistor.  Ferrite beads on the input of the preamp may help, check the schematic if this has been done.
4.  Some mic preamps "dump" the incoming audio from the XLR connector directly into solid state devices, which in turn rectify the RF from the nearby radio station, which you then hear clearly.  The input impedance of the preamp may be 1200 ohms or higher.  If you are using a U67 (or other tube mic), terminate the output (at the preamp input) with a 200 ohm resistor, so the mic output transformer secondary "sees" the correct load for its turns ratio, especially if using a "unknown" preamp.
5.  Are any cables or AC cord lengths approximately 1/4 wave length of the radio station frequency? Your cables can act as an antenna if resonant at the RF frequency.
6.  Resolution of "audio rectification" can be a multi-step procedure.  Start with housekeeping, then filter the AC, and work from there.  Avoid at all costs making modifications to the mic...or connecting capacitors across the mic output.  If you want to mess with filters, get some in line shielded "barrel" fittings such as the Switchcraft S3FM type and build in line filters...and LABEL THEM....

I made a remote recording at a chapel where a spectrum analysis revealed a defective MIDI box in a large pipe organ console, which was radiating a strong signal around 2.15 MHz and getting into everything.  The organ console was a fixed installation, and the wiring to the pipes and controllers acted like a large antenna.  I used multiple mics, an AC isolation transformer and multiple AC line filters, coiling a few turns of mic cable over ferrite core, the whole 9 meters, to get a suitable recording.  Moral of the story, other devices with clocking at RF frequencies can also trash a recording.


This a great post it should be a sticky
 
It's  a good post for sure. Lots of useful info.
One thing that I question though is the use of a 200 ohm load resistor.  A microphone such as the U67 mentioned wasn't really intended to operate in a matched line situation.
Nominal min. load as specified by Neumann for both original and reissue U67's is 1000 ohms.  Ignoring dcr etc., this reflects back through the U67 BV12 transformer as an approx 40K load to the triode'ed EF86.  Which is fine.
A 200 load would reflect as about 10K and your wee EF86 isn't too happy with such a low load.  Your -3dB point with the output cap will also be off.

About the same issues would be present if loading a BV8 type transformer in a U47 with 200 ohms.

Anyway, not wanting to trash on the post here, just saying.   

Edit:  If we're talking about an old U67 or U47 which was sold to the US market through Gotham Audio, then the bets have different odds.  Those 67's were strapped for 50 ohms output and had  pads installed in the power supply.  This because most American desks of the time were expecting lower impedance mics at much lower level. 
Most everyone has re-strapped the transformer for 200 ohms and removed the pad on those mics by now.




 
The Gotham NYC sticker is still on the bottom of the power supplies for the two U67's I use...
 
rmburrow said:
The Gotham NYC sticker is still on the bottom of the power supplies for the two U67's I use...

OK got it.  If those ol' boys are still strapped at 50 ohms output then a 200 ohm load on the line wouldn't have the same negatives in that respect. 

Thanks.


 
 
Also look at some transformerless solid state microphone schematics for some clues by companies that have engineers and real designers.
 
A lot of the transformerless gear isn't "hardened" for use around RF.  Common mode rejection works at reasonable desired/undesired signal levels.  Common mode (CMRR) and ferrite beads may be all one gets with audio gear these days.  Keeping high level RF out of linear equipment is a challenge when the desired level (mic level) is millivolts and the undesired (RF) is in volts.  At one time, there was a "unofficial" standard of 0.5 V/m maximum RF field specified where various computers and peripherals would operate error-free.  FCC Part 15 specifies maximum RF emission limits from "clocked" or other RF emitting devices.  There IS no FCC standard for RF susceptibility for most commercial or consumer equipment.  Equipment for military or official use must comply with various standards per the contract (which makes it very expensive).

I don't own or use any transformerless mics.  One could compare (with cable length approx. 3 m to the preamp/console) a 200 ohm termination (resistor), a EV RE20 (or other) dynamic mic, a Neumann U67 (or other transformer coupled tube mic), a Neumann U87 (transformer solid state mic), and a TLM103 or other transformerless mic, using the same preamp or console, within 0.5 to 1 km of a AM transmitter site, then repeat for a FM transmitter site.  The results one gets can be interesting.  AM stations often drop power at night, so test during the daytime.
 
101.5! It took a while but my friend finally found the radio station that is picked up in his studio. He was laying down a bass track (D/I) and picked it up clearly. It's strange that it's 101.5 because that's a low power university station.

So now that I know I need to filter 100MHz, where do I start in ordering ferrites for filters? I'm definitely taking your advice and building XLR filters that he could pop into any circuit. Not sure I'll modify my mics unless this doesn't work.

I started reading Jim Brown's paper. A lot of good info but I'm not absorbing it too quickly; might take a few reads. From another thread, the Laird part 28L0138-10R-10 was mentioned but I was hoping to find something from Mouser as I have an order to put in. What do I look for? A leaded single turn ferrite?

Do I factor in the impedance of the circuit when picking values or just plop in simple LC low pass with the ferrites and caps? If I want to filter 100MHz, what would I pick for my filter 3dB down point? 100kHz? 10MHz?
 
Delta Sigma said:
I started reading Jim Brown's paper. A lot of good info but I'm not absorbing it too quickly; might take a few reads. From another thread, the Laird part 28L0138-10R-10 was mentioned but I was hoping to find something from Mouser as I have an order to put in. What do I look for? A leaded single turn ferrite?
A multi-turn is more efficient. Check https://www.mouser.com/ProductDetail/Bourns/FB20011-3B-RC?qs=QVmd2DPn3usU2rFzlHEenw%3D%3D
It's the one that ofefrs good performance at 100MHz and wide enough range.

Do I factor in the impedance of the circuit when picking values or just plop in simple LC low pass with the ferrites and caps? If I want to filter 100MHz, what would I pick for my filter 3dB down point? 100kHz? 10MHz?
You can't really calculate, because for 100kHz, you'd end up with capacitor values that would also suck treble from the mic. Just use capacitors of convenient value (I use 22nf in this position). The resulting -3dB point will be around 100kHz, but it will be the result of the source resistance, niot the inductance of the ferrite, which action will be at much higher frequencies. It is very difficult to calculate the filter, because the parasitic elements contribute more than the nominal value. For example, the parasitic capacitance of the ferrite introduce an anti resonance, as the paristic inductance of the capacitors does, also. Even the length of the connections have an effect.

 

Latest posts

Back
Top