R/E/P > Klaus Heyne's Mic Lab Stickies

Radio Frequency Interference (RFI) in Mics

(1/6) > >>

Klaus Heyne:
Radio Frequency Interference (RFI) in microphones can be a pest- it comes and goes, then reappears as soon as one has forgotten about it, to damage a recording in progress.
RFI can be successfully and permanently eliminated in almost all cases.
I would like to share which remedies I have found are least impacting on the mic's sound while still taking care of the problem.

RFI comes in two audible forms: distinct radio station signals mixed in with the mic's audio, and noise hash, whose presence is much harder to hear and properly identify as RFI, because it appears as featureless background white noise.

How does a microphone pick up RFI?
RFI is often a result of inappropriate cable terminations, insufficient cable shielding methods or materials and/or certain designs of mic circuitry.  More rarely, RFI may be triggered by a very strong radio transmitter close by (which in turn exacerbates some of the triggers mentioned above)

Think of the microphone, including its cable, as an antenna: unlike other audio gear, the mic, as the first link in a recording chain, is open ended on one side (towards the head of the mic) and electrically connected only on the other with an amplification device which is nicely suitable to transmit what's been captured from this antenna.

Certain configurations of the mic's processor (f. ex. Nuvistors, discreet op amps, electronically balanced, rather than truly balanced, mic outputs) are often more prone to RFI than others.

How to get rid of RFI?
I use a methodical approach to curing RFI, based on my experience of what needs to be addressed first, second, and third, regardless of other potential variables. If I do not address the problem in this order, I may miss a cure and waste time.

1. Cable Terminations
Inspect the mic cable's connector terminations. I mean visually inspect, not just assume, their correctness.  Make sure that the following terminations are present:

- On phantom powered mics (and other mics with XLR three-pin connectors):
Pin #1 and the ground lug of each connector's housing must be connected together, on both cable ends!  

Pin #1 of the XLRs must then be connected to the cable's ground wire. If the cable does not have a dedicated ground wire, the cable's shield must be connected to pin #1. If the cable has both, ground wire and shield (Gotham, Neumann), connect both to pin #1 of the XLRs.

Beware of certain connectors (some Neutriks comes to mind) which do not have a separate ground lug termination point for the connector's housing. Do not use these connectors on mic cables, as the connector shells cannot be grounded.

When you are done, do the following test to confirm the correct connections:
An ohm meter's leads connected to both connector housings should read zero or a few single digit ohms. If it reads open, go back and inspect what's wrong, or you will get RFI!

On tube mics:
Connect ground wire and cable shield together to the pin of the connector that is dedicated to ground and install a wire from ground to make contact with the connector's housing. This connection is often conveniently made at the cable strain relief, where a clamp is screwed into threads of the connector housing.

Again: install this ground/shield scheme on both sides of the cable, and when done, perform the ohm meter test, as described above.

This method of termination is always applicable for microphone cables, except for rare multi-connector hook ups (I won't get into this here.)  Please do not mix up correct RFI terminations with what you have learned about terminations to prevent ground loops- there are no ground loops with mics and their cables, as they are single ended.
The terminations described above also happen to be the official Neumann recommendation.  

Gunnar Hellquist adds:
The microphone should be electrically insulated from the holder and hence from the stand.

If not, the mic is no longer ground loop-poof, as contaminating electricity, picked up from somewhere else, may flow through the conductive mic stand to the mic)

2. Cable material
I will not discuss advantages of certain cable brands, or their effect on sound. (I recommend and endorse, without payment, a specific microphone cable- Gotham, Switzerland)
I will share though, that in several independent cable tests for RFI susceptibility, one of them in an official AES paper, cables with double Reussen layer shields were provably more resistant to RFI than any other method of cable shielding.  

Reussen layer cable shields are nothing else but multi strands of copper wire twisted in a cork screw fashion around the cable conductors. A second layer is then twisted in an opposite direction over the first. This is visibly different from braided shields or foil shields.

To my knowledge, only Neumann, Berlin, and Gotham, Switzerland (whose Reussen cables are now made by Belden in Germany) use this type of shielding.

My experience has been that in nasty cases of RFI, this type of cable, in combination with proper terminations cured the problem when correct terminations alone did not.

3. Chokes
A choke is a wire coiled in such a way that certain frequencies of audio when it passes through the coil, are suppressed.
Typically these choke coils, which in mics are very small and look like resistors, are inserted in series at the balanced outputs of the mic.

This method of RFI suppression really works well- like a sledge hammer on mosquitoes: The deterioration of sound from choke coils ranges from subtle to unbearable, depending on the value of the coil chosen.

Typical values of coils factory installed in mics range between 40 and 100

Thanks for discussing this topic, Klaus. Jim Brown (whose profession is designing large sound installations) from Audio Systems Group in Chicago and I have presented several papers at the AES on this. The work is mostly Jim's.

The problem used to be that mobile high-power two-way radios would burst in on the session. Now it is more insidious. All of the digital RF devices that may visit a studio bombard the microphone with random pulse streams that can cause a very significant degradation in microphone resolution due to elevation of the noise floor. 5 GHz signals are common, high power signals up to 80 GHz (millimeter wavelengths) are coming soon. If someone is serious about high definition recording, it is necessary for them to pay close attention to this issue and investigate how RF can get into their systems, because it won't be obvious like it is if you get a good buddy in your overhead track. We have not found a condenser microphone in existence that is RF-proof.

Of course Neumann, fulfilling the specifications of the German state radio networks, pioneered RF-resistant microphones years ago with the M 2xx and KM 2xx series mics, which had additional bypass capacitors and a carry-through of the cable shield all the way around the signal conductors to the shell of the microphone.  

While I concur that your recommendations will eliminate some forms of RFI, grounding the shell of the cable connector attached to the microphone can provoke other trouble, particularly in hand-held microphones. This can result in additional hum and potentially a shock hazard if the microphone is touched by someone who happens also to be touching something at a different power ground potential. The resulting current travels through the mic pre via the shield of the cable, and depending on the design of the mic pre can cause additional hum as this current travels through common impedances such as pc board traces and wiring inside a mixer on its way to "ground."

There is an additional problem if you ground the shells to the shield in microphone extension cables; additional points are created where power fault currents can enter the shield circuit if the connectors are for example sitting on damp concrete or a metal catwalk. As a result you need to have some cables with pin 1 tied to the shield of the female connector (only) which are used for microphones with poor RF immunity, and other cables with no such connection which are safe to use as extensions.

The solution we are recommending is that the shell be connected to the shield for RF only, that is though capacitors. Neutrik has introduced a new connector with pass-through ferrite material to present some resistance to RF flowing on the signal leads, and a concentric ring of capacitors which connects the shield to the shell for high frequencies. This is an expensive connector but is an excellent solution.

Ferrites, particularly when applied without engineering the whole solution, can cause more trouble than they solve. The problem is that most ferrite materials are partly conductive. If you put a ferrite bead on a wire and anchor it in place so it cannot touch anything else, that's fine. But if the bead touches anything else while it is also touching the signal lead, you now have a variable leakage path that can cause serious trouble, particularly at the input and output of a condenser microphone impedance converter (for the same essential reason -- a large DC potential superimposed on a small audio signal). Microphone companies who successfully use ferrites for RF suppression take particular precautions in this regard to avoid such problems.

It would be better if manufacturers paid more attention to this problem. A few companies have begun to do this, but as Jim and I documented there are several famous microphones with fairly long wires between pin 1 of the connector and the shell of the microphone. These wires form antennas that radiate shield current into the circuitry of the mic. Making this path as short and low inductance as possible will help quite a bit.

Yannick Willox:
The BIPT that came to measure the RFI in the concert hall where my problems started (they measured no abnormal RF, just very high levels of the transmitter at 104 MHz) state that the best way is connecting a ceramic capacitor between the conductors of the mic cable and the earth (10 nanofarad or 100 nanofarad).

I think I'll try those Neutrik XLRs first.

Yannick Willox

Doc B.:
Hi Klaus,

If I may offer a couple more techniques for RFI/EMI reduction that we have experimented with as regards small signal tube amplifier circuits that use tubes with VHF type bandwidth. Caveat emptor, you get what you pay for, etc., and please take this post down if it's getting too nuts and bolts for the discussion-

A small value 1/4W carbon composition resistor in the 50 ohm to 1K ohm range can be effective as a "grid stopper" and also as a "plate stopper" to block RFI. In each case the resistor must be installed with the body of the resistor right at the grid or plate terminal of the tube socket, in series. Generally the lower the value, the less influence it has on the sound, and one can thus season to taste by changing the value. I suspect it might be worth trying in a case where an RF choke is killing the RFI but also killing the sound, to see if it has less negative impact on the sound than the choke. FWIW we use 220 ohm stoppers with 6DJ8 type circuits and 50 ohm stoppers with tubes like the 417A or 6C45pi that want to see a low impedance grid load. Something like a 12AX7 or 12AT7 or 6CW4/13CW4 might even be OK with as high as 500 to 1K ohms.

Also, 1uF @ 50V X7R or Z5U ceramic caps can be effective RFI blockers when one is installed as a shunt across the heater pins, and one each is connected from a heater pin to chassis ground, as close to the tube socket as possible.

Rob Darling:

is the Neutrik product you are speaking of the xcc series xlr?


[0] Message Index

[#] Next page

Go to full version