Attenuation for Long Wave ?

I have strong signal from Local LW station on 252 Khz, almost 60 0ver on the Kiwi.

Any filter I could use to filter the LW band, I am more keen to see if I can just reduce 252 Khz rather than a whole chunk of the LW or MW band or both.



  • edited June 2022

    In the past, I've managed to notch WWVB at 60 kHz and only 20 km distant, with a simple series LC filter at the 50 ohm input. I can't easily find a schematic of what I used but from memory I think it was a high permeability toroidal core with whatever turns I could fit and an appropriate C. It was pretty sharp and didn't hurt VLF or MW particularly.

    If you try it and find out-of-band insertion loss objectionable, you might try two L's in series with the second designed to be below self resonance and 50 ohms at 30 MHz

    I think I must have it in a drawer here somewhere but that too isn't immediately discovered.


    I found something. I don't know what the toroid core was but it was probably no bigger than 1/2" diameter:

  • There certainly was ! How quickly I forget...

    Apparently none of those previous suggestions were followed and the request "Any filter I could use to filter the LW band" doesn't include home constructed ones.

    oh well....

  • Yes it's a pity that most of the commercial AM bandstop filters don't include the LW broadcast band.

    It's also a pity that so many folks who have paid good money and installed KiWi's don't seem to have paid much attention to the antenna system.

    In the past I have sent filters to folks with decently sited KiWi's using good antennas, but which were suffering from overloads.

    But lack of time and diminishing returns have taken their toll, and I no longer do this.



  • edited June 2022

    Thanks for the replies lads, yes I had a thread on it a few years back but my main issue is time, I don't have the time for tinkering around with stuff. Working 5 days a week, Kids and 10-12 hrs on the road commuting, + work around the house/ big garden to maintain so sometimes I need a faster commercial solution. This is a hobby and I devote as much time to it as I can as it is which is not a lot but I will have plenty more time when my Boys are older, at this point in time family is far more important to me.

    It does look like there is no commercial solution.

  • "It's also a pity that so many folks who have paid good money and installed KiWi's don't seem to have paid much attention to the antenna system."

    The antenna is fine, don't need any solutions for the Airspy HF+, it works just fine with this antenna.

  • According to this I need 5000PF caps and 80uH inductor.

    How many caps is the question ?

  • The ratio of inductance to capacitance sets the notch bandwidth, so you can have a wide or narrow notch depending upon the values you choose.

    The key thing is to use an inductor with a high value of Q and operating well under its self resonant frequency.

    If you use a fixed inductor you may have to use a combination of fixed capacitors and a variable pre-set capacitor, so that you can fine tune the notch to the exact frequency.

    In the original thread I suggested using values of 27uH and 14.6nF.

    The inductor could be something like 75 turns on a T50-2 (red) iron powder core and this could be fine tuned by adding or removing turns as required.

    If you used a fixed 15nF capacitor, you would only have to remove turns as it would initially be resonant on too low a frequency.



  • I have a few FT-240 31 cores lying around, could I use those ?

    What wire should I use ? I read that thicker will increase the Q ?

    What about capacitors ?

  • I think you want a smaller core with lower permeability, maybe something like this (handy calculator here)

    with about .012 uF. I don't think wire loss will dominate so small wire that lets you get 20 turns and pretty much fill up the core probably would be reasonable. Maybe use .01 uF ini parallal with a trimmer to put the bottom of the notch right on .272 kHz.

    This is where a nanoVNA or similar is really handy.

  • I have made pretty effective notches using MS75085 fixed inductors and series capacitance compised of parallel silver mics fixed caps and a 35 pf trimmer cap. I followed that filter with one made with T50 core wound inductors and found little if any improvement

  • if your only concern is 252 you just need a notch

  • An FT240-31 core is not suitable. Fair-Rite #31 material is a lossy ferrite material designed to provide resistive choking.

    The suggested T50-2 toroid uses Micrometals #2 material which is powdered iron, not ferrite.

    To get 75 turns without overlapping them you will need to use a thin wire of around 0.25mm diameter (approx 30 AWG). This will provide plenty of Q and you may even want to reduce it with a series resistor.

    Capacitors - ideally polystyrene or silvered mica if you can get them. Younger readers may be able to recommend a modern substitute.

  • edited June 2022

    I take your point about lack of time and family commitments, having long gone though that stage of life, and now having the luxury of retirement. I do tend to forget how demanding it was at the time, so you have my sympathies.

    My comments about the poor performance of some KiWi's was really directed at the very poor ones that are in the bottom 1/3 of the 600 or so KiWi's listed on this site.

    They typically only mange to achieve a S/N score of 10dB or less, and some don't seem to have any antenna connected.

    The next mid range block of about 200, manage to achieve scores between 10dB to 15dB, which is better, but still not really quite good enough to be useful.

    The top 200, who regularly manage to achieve more than 15dB are generally usable, but a lot of the really top rated ones either have large chunks of the spectrum masked off, or are experiencing frequent ADC Overloads. It's the KiWi's with ADC Overload, but with otherwise good reception or useful locations for TDoA, that I would have tried to assist in the past.

    The 75 turns on a T50-2 core don't have to be neatly wound and overwinds are acceptable. I used some PTFE coated wire intended for wire wrapping prototypes (I have a lot of it from clearing out labs), and I find that it's ideal for winding all sorts of inductors and transformers.

    You could use an FT50-61 core (low permeability ferrite) and would only require 20 turns to achieve 27uH, but it it would not be as predictable as one wound on T50-2 material. Don't consider using any higher permeability ferrite such as 43 or 31 mix, as these are even less predictable and much more lossy.

    Some moulded axial inductors with fixed values will work, but many do not, a lot depends upon the materials used for their construction. Generally speaking they are quite low Q, and not at all predictable.

    If you don't have any test kit (apart from the KiWi which is actually quite useful in the workshop), I'd stick with using T50-2 red iron powder cores for tuned circuits. However if you have something like a NanoVNA, and can measure the circuits RF performance, that opens up the possibility of further experimentation with different components and methods of construction.



  • I forgot to mention.

    If you wish to calculate inductors would on various core materials, this software is pretty useful as a starting point.



  • Thanks, I will see if I can get the parts locally.

  • If you get really stuck send me a PM.



  • Will do Martin Cheers.

  • edited June 2022

    Hello all :)

    I've designed a few band-stop filters with this on-line tool:

    I live in Europe and these are the only solutions, for me, to get rid of -mostly- China Radio Int. on various SW bands that make KiwiSDr overload and being unusable.

    I've already built one for 49 mts, another for 41mts is being soldered, and others for 25 and 31 are in the plan.

    I then hope to fully enjoy the 120 mts long horizontal loop that lays over my head :)

    Thanks to anyone spending time to share his knowledge and experiences on this forum.



  • Based on the above link this is the filter it designs for me.

    This is the circuit that was created by me, I based it on an lower and upper cut off based on the bandwidth of the 252 Khz signal or should I just input 252 Khz with no upper and lower ?

    Also, if I go for commercial inductors what type should I look at buying, I would love to make my own but haven't the equipment to measure the inductance and it would be nice to make it variable to make it more convenient to be able to move the null up and down a little rather than having to remove some wire of each toroid and try again, where would be the best place to put a trimmer ?

  • Although a complex filter like this can no doubt be made to work just fine, it reallyl is over-kill for your problem. All you really need to make the Kiwi happy in your situation is a few dB of attenuatiom right at 272 kHz. For this, the simple series L/C will work well and be much easier to build and tune.

    I suggest a simple approach may be best here.

  • nearly 60 dB over so probably want at least 20 dB reduction ?

  • "This is the circuit that was created by me, I based it on an lower and upper cut off based on the bandwidth of the 252 Khz signal or should I just input 252 Khz with no upper and lower ?"

    I meant calculated by the software on the website not created by me lol.

  • Ceramic vs Mica ?

  • there are good and bad ceramic..... their stability over temp can vary.

Sign In or Register to comment.