Using RF cables as filters for wireless microphone receivers.
There are often situations where a strong signal may interfere with the reception of wireless mic signals.
An old trick which has been largely forgotten about, is using certain lengths of coax (RF) cable to create notch filters and thus reduce this problem.
Before getting into the construction details, a quick, simple introduction for those that may not be familiar with RF design work is in order.
A coax cable is essentially just a conduit to carry RF signals from a source (e.g. an antenna)
to it's final destination, a receiver in this case.
It goes without saying, that this conduit should exhibit as low a loss as possible, least amount
of radiation (and thus ingress of interference as well) and not change the signal it is carrying in any way.
For this article, things like the insertion loss and shielding, are not important, except to mention you should always use the best quality cable possible.
One of the side effects of coax cable is, because it's made of metal and insulating materials as well, these will actually change the propagation speed of any RF signals (relative to air) it is carrying. This is known as the velocity factor and varies depending on type of cable and manufacturer.
This is important to know, the velocity factor can be thought of a correction factor for cable length that we need to apply to our calculations in order to construct these filters.
Most of the reputable cable manufacturers publish the velocity factors for their cables. Below, an example from Belden cables: (Disclaimer: I'm not affiliated to Belden in any way)
Note however that these figures are measured at a specific frequency and do vary slightly depending on which band/frequency you will be using them at.
In this application though, it's not too much of a problem as the construction of these filters is so simple that some snipping with side cutters is all that is required.
The deviation from the published figures is normally anywhere between 0 and +/- 8%.
Before moving on to the actual construction, there is one more piece of theory that needs to be understood, wavelength.
I will not go into a detailed explanation about this, will rather leave it as an exercise for the reader. All you need to know is that any frequency has a certain wavelength which can be worked out
as follows:
The wavelength (symbol Lambda) in meters is 300 divided by the frequency in MHz.
We will not be using the full wavelength but rather 1⁄4 wavelength due to it's phase characteristics. It's these characteristics that create a notch at it's design frequency and odd multiples of it.
What you will need:
– A good quality (50 Ohm) BNC T-piece (all female).
– Two lengths of good BNC to BNC cables. (Length up to you but not too long) One to connect from the T-piece to the receiver and the other to cut up one end so it can be used as the 1⁄4 filter.
You need to know the velocity factor of the piece you will be cutting up.
If you cut it exactly in half, you can get two filters out of it. We will trim them more precisely later.
– A pair of side cutters able to cope with cutting the above cable.
– A ruler.
Before giving an example, a picture of what the filter will look like.
It's as simple as it looks.
One side of the T-piece you connect to the antenna, the other to the receiver and the third will have a certain length of open ended cable connected to it.
The length of the open ended cable is what we need to work out.
Note that it is very important that the open end of the cable has no shorts on it else not only
will the filter not work, but you will short out any DC voltage that the receiver may be sending
up the cable if you are using an active antenna.
Check and double check that no bits of the outer braid are touching the inner conductor of the cable.
EXAMPLE :
Let's say you are using 650 MHz and you have a very strong problematic signal at 245 MHz. Let's also assume the cable you have has a velocity factor of 84.
Although you may think that these are too far apart to cause any problems, you will be surprised. Depending on the signal strength of the unwanted frequency and how effective your receiver's filtering is, it may well swamp the receivers front-end pre-amp and cause interference.
On to the maths:
Wavelength = 300/245 = 1.224 meters.
However, we will only be using 1⁄4 wavelength so 1.224 meters/4 = 0.306 meters (30.6 cm). Now apply the velocity factor which if you recall is just a correction factor as a ratio.
So, 30.6 cm X 0.84 = 25.71 cm.
Last thing to do is subtract from 25.71 cm, the length of the connector on it.
You don't have to be exact, just subtract around 3 cm.
The final figure for the length of the open ended cable is then 22.71 cm, you an round off to 23 cm.
Done.
To make it look neater, you can use a rubber "end cap" on the open end. Let's look at the frequency response of the filter:
"F" is the design frequency (the fundamental) which in this case is 245 MHz.
3F (3 X 245), 5F (5 X 245), etc are the odd harmonics and are also attenuated but at a decreasing rate as the frequency increases.
The even harmonics 2F, 4F, etc, will not be affected.
Always check that the wanted frequency is in the area around these.
Double checking the frequencies that will be rejected the most:
1 X 245 MHz = 245 MHz
3 X 245 MHz = 735 MHz
5 X 245 MHz = 1225 MHz (out of the wireless mic frequency range so ignore).
Our wanted frequency of 650 MHz will have very little to no loss, which is the goal.
It's very important to check how close the wanted frequency is relative to the frequencies that
the filter is rejecting.
Keep in mind that these filters are not very steep and thus the usable range around 2F, 4F and so on, will gradually start to decay and reduce your wanted signal if they are too close.
This can be better illustrated using the same graph as before:
Let's assume the frequency we want to attenuate the most is where the green circle is.
It's attenuated by about 22 dB but the wanted signal (orange circle) is also attenuated by 3 dB.
This may still be perfectly acceptable as suffering a 3 dB loss on the wanted frequencies but
having an attenuation of 22 dB at the unwanted frequency, will be enough to get rid of the problem.
This then, can be a disadvantage of this type of filter but given it's simplicity, low cost and ability to build one almost anywhere in a short time, are huge advantages.
The more you build and use them, the better acquainted you will get with their pro's and con's.
There are of course better filters (depending on situation) but those will require a deeper understanding of RF theory and test equipment.
Perhaps then, something for a future article. Happy building.