Guides, Tech Tuesday

Tech Tuesday! – Filtering fact from fiction

Filtering fact from fiction – Everything you need to know about filtering in RF wireless

By Pyers Easton 

 

In this week’s edition of Raycom Tech Tuesday I am going to cover something that I am often getting asked about and that is often misunderstood; RF Filters.  

I think possibly the easiest way of visualising this is to think of the RF spectrum in terms of the audio spectrum, they are similar in many ways. Audio operates at lower frequencies than wireless mics and propagates with air pressure waves instead of electromagnetic waves, but many of the concepts are the same.  

If I can continue the audio analogy; imagine that the RF waves hitting the antenna in your wireless system are like the sound waves hitting the diaphragm of a microphone, the pressure waves generate a small amount of electricity which needs to be amplified, in the same way as the electro-magnetic waves hitting the antenna generate a small electrical current. 

Once these waves have been converted to electrical energy this is where the problems can start as you need to amplify them to a useful level, amplifiers always have a maximum output level which can’t be exceeded, and if you hit this maximum level then serious distortion will result. The normal mode of distortion when you run out of headroom is clipping, this is where the voltage out of the amplifier cannot go any higher and ‘squares’ the waveform producing a Christmas tree of harmonics. As well as harmonics, when overloaded like this, ‘intermodulation distortion is produced, this is where multiple frequencies mix together and create lots of new, unwanted frequencies.  

You will see that all of the above really does apply just the same to RF as audio, in fact I do try my best to demystify RF, it’s not really as much of a black art as some would have you believe. If you understand audio then you will understand much about RF already, but may not realise it! 

 See the diagram below and you will see the parallel paths of an audio and an RF system: 

You can see that the microphone diaphragm is analogous to the antennayou could think of these as being passive components for the purpose of this document and therefore not susceptible to overload in the same was as a passive antenna (though in practise the microphone diaphragm will run out of physical deflection and cause distortion, but we can disregard that here). The drawing shows a configuration where the microphone has a built-in pre-amp (such as in a condenser mic) and the antenna also has an internal amp. In this configuration you can see that there are 3 potential locations for a filter, but more on that shortly.  

Now, this is the point where the RF and audio comparison starts to diverge slightly. For audio you are normally interested in hearing the entire audio spectrum whereas for RF you are interested in single spot frequencies. This is where filtering comes into play, for audio you might use a filter to reduce air-conditioning rumble, or even ultrasonic noise from distance or movement sensors. In these cases, it is possible to remove the unwanted sound whilst still passing the sounds you wish to hear.  

In the RF world the same appliesyou need to pass all the frequencies which you need for your wireless channels but filter out frequencies which might cause overload, such as high-power walkie talkies or mobile phones.  

Given that we have established the diaphragm on the mic and antenna do not introduce distortion due to overload in this example, but there are amplifiers in the path which can, it should be clear that the ideal location for the filter is BEFORE any active circuitry, in other words between the antenna and first amplifier in the chain. As you will have losses in the cable, and if you want to compensate for these losses, you must put the first amplifier in the RF path as close to the antenna as possible (otherwise you will just be amplifying thermal noise, but that’s a subject for another post!) ergo, with an amplified (active) antenna the filter would ideally go in the antenna itself (position 1 on the diagram).  

The sort of filters used for RF are basically passive themselves, and will not introduce distortion at the sort of RF input levels one is likely to encounter in a real-world situation. There are caveats, to this; filters which are electronically tuneable usually achieve this by using capacitors which are adjusted by voltage, these components (known as ‘varicap diodes’) are semiconductor based and therefore can exhibit unwanted characteristics in the presence of large RF fields. It is therefore essential to have a properly designed filter with very good quality components or the expected performance will only be achieved under undemanding conditions. 

Once again, one could compare audio and RF here, though audio filters tend to be active and based around amplifiers, thus have a maximum level above which distortion will be produced so the advice on placement in the chain is less clearly defined.  

Let’s stick with the RF domain from now and look at the pros and cons of the three filter locations I have identified in the diagram:  

  1. This is the ideal location when an active antenna is used as it provides filtering before any active circuitry which might overload. The downside is that the filter might be on the end of a long cable and maybe in an inaccessible position, which could be awkward if it needs to be tuned to a different frequency, though there are antennas now which can be remotely tuned by communicating through the same coax cable which carries the RF and DC power. The filtered active antennas Raycom sell have the filter correctly located at position 1. If a passive antenna is used then there is no benefit to putting the filter here, and the filter might as well be at the RX end, position 3, and the only important thing is for the filter to be before any active circuitry.
  2. If you have an amplified antenna without a built-in filter then one might be tempted to put the filter here, but, as it’s already too late as it’s after the amplifier, then you might as well have it at the RX end, position 3.
  3. This is the most convenient location for your filter if you are using a passive antenna or an active antenna with no filter. 

Filter types: 

There are several different types of filter which can be used individually or together. Bear in mind that the aim of the filter is to pass all the frequencies you want to receive, and to reject all the frequencies you don’t want. The ideal choice of filter(s) will depend on how wide the range of intended RX frequencies is and how close these frequencies are to other potentially interfering sources. Here’s the main filter types:  

  • High pass/Low pass – As the names suggest, these types of filter pass frequencies either above or below a certain frequency. These are handy if you want to receive a wide range of frequencies but reject signals above and/or below the range you want. A typical example for high-pass application might be where you have many wireless mics between say 500 and 700MHz, but have walkie talkies pushing out 5 Watts on 469Mhz.  Lowpass can be added to reject frequencies above say 700MHz where there might be high power mobile phone transmitters. Some antennas have a range of switchable low/high pass frequencies which can be chosen to produce the best compromise between bandwidth needed for the channels you wish to use, with best rejection of frequencies outside that range.  
  • Bandpass Filter – This is a type of filter which passes a range of frequencies and rejects outside this band. Bandpass filters are most useful when they are tuneable so you can tune to the range of interest in a given location. Obviously the width of the band it passes needs to be wide enough to pass all the channels you need, if this is not the case the options mentioned above for high/low pass may be a better choice. Electronically adjustable, remote controlled bandpass filters are now available in active antennas and inline amplifiers. For electronically tuned filters please be aware that good quality is essential as you might get unexpected results, also a poor quality filter might have a lumpy response as you tune it (passband ripple). 
  • Notch Filter – These filters are useful if you have a specific small range of frequencies you wish to reject. They have the advantage of passing a wide range of frequencies whilst rejecting only a focussed range. This type of filter is less prevalent in our industry, but is sometimes specified for specific installations where a particular threat needs to be mitigated.  

 Some myths:  

  • Filtering will make my receiver more sensitive  Actually, if the filter is used as suggested, before any amplifier, then the opposite is truePassive filters have a loss, even in the intended passband. A well designed filter will have a low loss of say 0.3dB to 2dB which is not much, but it will reduce sensitivity by increasing the noise figure of the system by its loss. Having said that, if your system becomes overloaded for any reason by unwanted signals, then you are much better off with the small loss in the filter than the range-reducing effects, or even total obliteration, from overload.  
  • Always tune your filter centre frequency to the desired signal  Whilst it might seem logical to always tune your filter dead centre to the frequencies you are wanting to receive, this might not always be the best option. Let us say, for example, that you want to receive 485MHz and have a band-pass filter bandwidth of 40MHz, but have a high power walkie talkie on around 470MHz. If you were to tune the filter to 485MHz then the passband would be 465 to 505MHz, and would thus not attenuate 470MHz. If you were to offset tune the centre to 505MHz then your passband would be 485 to 525MHz, so you would pass the desired frequency, but the unwanted signals would be well in the ‘stop band’ of the filter, thus significantly attenuated. 
  • Gain, the more the merrier! – WRONG! As we have seen in the signal flow discussed here we are typically faced with a number of amplifiers at different points in the chain, just as with audio, gain structure is critically important and too much is often worse than not enough. The quality of the amplifiers and their ability to handle large levels of RF without distortion is critical. Once again it is really important to use the best quality equipment or risk introducing a weak link in the chain. For some more reading about intermods take a look at my earlier post here: https://www.raycom.co.uk/so-just-what-is-an-intermod-then/

I hope this has answered as many questions as possible and made the subject a little clearer. I and the team at Raycom are highly skilled in all things RF, so, if you have any further questions, problems which need solving, or need info on specific products then please do get in touch.

Email: Sales@raycom.co.uk 

Telephone: 01789 777 040

Facebook: https://www.facebook.com/raycomltd

 

 

 

Leave a Reply