Broadcast band filter

With the erection of my new horizontal loop antenna, I’ve found all of a sudden that I have quite a bit more signal for my transceivers to listen to. However, that’s brought with it the problem that it’s also picking up very strong broadcast band signals that are overloading at least some of my rigs front ends. In particular, my KN-Q7A has little protection from this sort of overload and is badly affected to the point of being unusable with this antenna. This is a known problem with the rig and is simply a limitation of it’s basic design.

To see how bad the interference was, I connected the antenna to a Spectrum Analyzer. The result is shown below.

BCB interference

 

As you can see, the strongest signal is ABC 1026 at only -9 dBm – that’s greater than S9+64dB! The sheer number of similarly strong signals is a bit much for many simple receivers to cope with.

The answer was fairly obvious – a broadcast band filter. I looked at the commercially available units and found that filters capable of being transmitted through at the 100W level were expensive, so I decided to build my own as they are not complex devices.

I first modeled a filter using ELSIE free version and found that a 7-pole Chebychev filter looked like it would do the job. I set the cutoff frequency to just below 3.5MHz as I am unlikely to get on 160m with the space I have available. The model indicated that I should be at least 40 dB down by 1.7MHz which should be sufficient.

The circuit is as follows (as drawn in LT SPICE IV):

Image
The following table has the component values and expected component currents and voltages at 100W. It also has the winding details for the inductors. I used type 6 material to ensure good performance to at least the top of the HF band and the T130 size is large enough to avoid saturation at 100W. The 1 mm enamelled wire keeps the RF resistance low to reduce loss. The capacitors are 500V silver mica units which have been used many times in 100W class filters over the years and proven reliable.

Components
I built the filter in a simple diecast box mounting the inductors using discs of kitchen polyethylene chopping board to separate them from the box. The other components are simply wired point-point with SO239 sockets on each end of the box. The unit is completely symmetrical, so it doesn’t matter which way round it goes.

BCB filterThis has been tested to 100W with 2 min carrier into a dummy load leading to only mild heating of the capacitors on all bands from 80m through 10m.

I put the filter on my N2PK VNA and was pleased with it’s actual performance as indicated in the following traces:

Basic trace of the filter showing more than 100dB down by the bottom of the broadcast band and a respectable 45dB down at the top

Basic trace of the filter showing more than 100dB down by the bottom of the broadcast band and a respectable 45dB down at the top

VNA1

This is an expanded view of the insertion loss across the band to 55MHz. Above 80m, it is less than 0.07dB which is excellent. on the 80m band it peaks at 0.2dB which is still quite acceptable.

 

This plot shows the SWR across the band. Apart from 80m, the HF band SWR is better than 1.2:1. On 80m, the peak SWR is  1.3:1 and on 6m it rises to 1.35:1 - still quite usable.

This plot shows the SWR across the band. Apart from 80m, the HF band SWR is better than 1.2:1. On 80m, the peak SWR is 1.3:1 and on 6m it rises to 1.35:1 – still quite usable.

The final test was to look at the broadcast band with the filter in-line. Here’s the plot:

Broadcast band interference with the BCB filter in-line.

Broadcast band interference with the BCB filter in-line.

As you can see, the interference has been dramatically reduced with the strongest signal now at -81 dBm (about S7) – about 72 dB better than without the filter. Needless to say, the KN-Q7A now works well with the loop antenna.

 

13 thoughts on “Broadcast band filter

  1. Hi again David,
    I note you have updated your site re the 100 watt capability. Good on yer!
    Now to be a bother again, I have been attempting to characterise some Jaycar toroids to use in your design, but have onus that they have a very high Al value and aren’t really suitable for the required inductance values. So, my question is where did you get the FT 130-6 Amidon toroids from? I have t68-2, but I don’t think they would handle the 100watt level.

    I am about to check out tts systems but I haven’t seen reference to the 130 size on their web site. Any advice /guidance would be most appreciated.

    Regards
    Frank
    VK1VK

    • Frank,

      Yes, the T68-2 toroids would saturate badly at 100W. Also the type 6 material is lower loss across most of the HF bands, so a better choice in my opinion than type 2.

      I obtained my cores from kitsandparts.com in the USA which has very good prices for cores, but is offset by shipping costs. If you think you’ll be needing quite a few cores for different projects, then buying a selection from that site can be good value. Not so much if you only want 3 cores. Locally, TTS may have them or could get them, but minikits.com.au in Adelaide has them in stock.

      73

      David
      VK3IL

      • Many thanks David. I called tts and they had them so have ordered same from them.
        Sorry about the typos in my earlier email!
        I’ll keep my ears open for your callsign. I’m often on 40metres in the evenings.
        73
        Frank
        VK1VK

        • No problems. I don’t get on much of an evening due to long work hours and family commitments, but I’ll keep an ear out.

          73
          David
          VK3IL

  2. I’d like to draw your attention to an article in the August 2016 issue of Amateur Radio CQ Magazine. It’s on pages 38 – 41. “A Cheap and Easy BCI Filter” by Jack Purdum, W8TEE.

    The circuit and component values are identical to yours.

    RICH WA6KNW

    • Thanks Rich. Yes, I knew Jack was writing that article and I gave my permission to use my design. Haven’t seen the final article as I don’t subscribe to CQ, but will look out for it.

  3. A friend is having BCB problems and we are going to build a filter from your design. A quick question, I happen to have some .0015 orange dip 630V Metal-Foil Polypropylene Film caps in the stash. OK to use these or should I use Mica caps?

    Thanks!

    Jonathan – KK6RPX

    • Hi Jonathan,
      I’m not too familiar with the characteristics of this type of capacitor, so not absolutely sure, but they look like they may be OK. Voltage rating is fine. Question is how they would stand up to transmitting current. I’m guessing they would be OK. Other possible problem would be if they were too inductive, but manufacturers try to minimise inductance, so probably OK. If you can find a data sheet, that would help. In particular, look for the SRF (self resonant frequency). You’d want that to be well above your operating frequency. If you have suitable test equipment available, you can simply give it a try and measure the filter characteristic. Also check for temperature rise after a transmit carrier is applied for 30s or so.
      73
      David
      VK3IL

  4. David,

    Thanks! I don’t really have the right equipment to properly test so I probably ought to just use the mica caps. You know how it is, a project using parts out of the junk box just tastes better. 🙂 I’ll order up the cores and caps.

  5. Got the filter built yesterday and boy does it wipe out the BCB! We used 576 pF caps instead of the 560’s because that is what my friend ordered. A question: The SWR is a little high, 1.7:1, on the lower half of 80 meters. By the phone portion of the band, say 3.6mHz it drops back to good. Is this likely an effect of the change in cap values? This gives me a push to learning SPICE as I’d like to be able to answer that question myself.

    Thanks for the time and effort of your write up on this filter. We ran out of time to really characterize the filter with what primitive tools we have at hand but it clearly knocks the stuffing out of signals below 3mHz!

    • Glad to hear it’s working for you. The SWR will be higher near the bottom of 80m as you are getting very close to the cut-off frequency. The extra 16pF could well be the cause of the slightly higher value. You could either replace the 576pF with 560pF, or perhaps easier try adjust the turn spacing of the middle inductor. This will increase (closer together) or decrease (more evenly spaced) the inductance a bit and will likely change the SWR at the low end. According to a quick bit of modelling, decreasing that inductance a bit should help, but taking a whole turn off will likely be too much. You could however try removing 1 turn and then bunching up the windings. If all else fails, 1.7:1 is livable with.

  6. Thanks for the reply. I think just getting the right caps makes the most sense, but my friend is not at this point operating CW so leaving it as is would be fine. Thanks again for the write up on the filter!

Leave a Reply

Your email address will not be published. Required fields are marked *