Any Ideas [AM BCB overload]

I have a mini whip on my tripod on my roof also next to a TV antenna, and a Discone (all on the same tripod). The mounting is temporary using a satellite arm off the main mast. I go from BNC to RG6 down about 30 ft to an enclosure where the LNA is. I split the ground off there to a metal spike in the ground. The Coax continues another 15ft to my computer room. I get these equally distanced lines almost at all bands. I do not have any filters (yet). I will be putting an AM and FM bandpass filter on to test soon.

But anyone have any ideas what this could be? I'm a relative newb to this.



  • You're in overload (see OV at bottom right of page). You may be getting whacked by an AM station
  • Yes, unfortunately you have two massive AM broadcast signals on 760KHz and 800KHz.

    As a temporary measure try attenuating the input signal by about 10dB.

    A longer term fix is to build a notch, or high pass filter to reduce the level of the BC stations without reducing the level of other signals.

    Some notes on the subject can be found towards the top of this web page.


    Martin - G8JNJ
  • edited July 2018
    Many thanks. I'm initially just going to use the AM filter entirely. Although I do like to listen into AM radio periodically I can do without it on this unit. I think it is killing all the signals below as well. The VLF LF sections are just nothing but noise at this point.

    Ahhh the OV yes I suspected that. Thanks. Hopefully that AM filter comes in today!
  • Also @G8JNJ nice Websdr site. Very cool you have the funcube and ISS there.
  • @Lonecrow - thanks.

    I think once you have sorted out your ADC overload you may still have some problems with your Mini-Whip.

    It uses the coax as part of the antenna, so if you share the coax run with other cables, or the base of the Mini-whip is connected via the mounting metalwork to other antennas, they tend to induce unwanted noise onto the antenna feed.

    The trick is to keep the Mini-Whip coax as far away from other conducting objects as possible, run the coax down to the ground and connect it to your ground spike. Then add a ferrite cored choke Balun onto the coax between the ground spike and the point at which it enters your property. This should help prevent unwanted noise from within your property from being carried along the coax screen back up to the antenna.

    Something like this.

    More information can be found on my active antennas webpage.

    Ideally you would design the ferrite cored choke Balun to have the highest value of resistive choking impedance at the frequencies where you have the worst noise problem or where you tend to enjoy listening the most. You can add further choke baluns at other points along the coax cable, but don't add any between the antenna and ground spike.

    Steve, G3TXQ, has a good choke Balun design chart on his website

    Send me an email if you need any further help.


    Martin - G8JNJ
  • The notches I use on AMBC suck out the BIG ones at my house yet allow most of the AM band to be received
  • edited July 2018
    I actually have some ferrite on my power supply now and that did actually help a little but the actual ground wire is just a green copper wire I ran down to a metal spike I put a couple ft into the ground. I also tried mounting it in the actual electrical systems ground (the 3rd prong) without much difference.

    But adding one to the actual coax sounds like a good idea. Unfortunately I have little in the way of options at the moment to remount it somewhere else. Although this isn't the permanent location, its going to be actually place on a tower.

    Does the length of the antenna to the LNA make a difference like it does for UHF/VHF? I understand you want your LNA and filters as close to the antenna as possible. If this were the case I could mount an actual box on the tripod and just run ethernet and grounding down from the Kiwi. But like you said it uses the coax as part of the antenna correct? SO I think you'd want it long.

  • Yeah looks like I'm going to need to scrap this method. If it doesnt like power cables nearby etc I'll need to build a different type of enclosure.
  • @Lonecrow

    Two things make a difference to the Mini-Whip signal level. The height above 'ground' and the length of the actual Whip element.

    PA3FWM has some good notes on the subject.


    Martin - G8JNJ
  • edited July 2018
    I found here an earth at the base, Low Pass Filter (lots of VHF here) CCTV balun, and CAT5 back to the house (buried slightly), another short RF earth, CCTV balun etc. tamed it.
    The higher frequencies were nothing special so I looked for something else but a MiniWhip here was only usable in this layout. Oh and Lonecrow an earth for mains and an earth at RF require different things, I use short braid from RG213 at the antenna end to a buried copper (clad) spike and wide earth bonding braid at the house end, everything as short to ground as possible.
  • I would think your Cat5 feedline might suffer from RFI ingress
  • WA2ZKD From real world testing it was markedly better. Admittedly the label is wrong (not STP CAT5, more correctly S/UTP - single shield) but it was created mostly as notes for recording changes.
    I have a broadband RF transmitter within 30ft (As I mention in nearly every post), each incremental change in noise pickup is visible instantly.
  • @Powernumpty

    Using CAT5 should be fine, the big problem with Screened CAT5 is what to do with the screen connections at each end of the cable run. Sometimes this can just make things worse.

    Balanced twisted pair depends upon the two ends being connected to properly balanced terminations. Using a video Balun is a good starting point, but make sure they are of good quality. Some of the cheap generic Chinese CCTV Baluns are real junk, especially in terms of insertion loss. Some I have measured had an insertion loss of >10dB at >4MHz, so if you had two connected back to back, the performance at 20 to 30MHz would be appalling.


    Martin - G8JNJ
  • edited July 2018
    We measured the frequency cover and insertion loss with a MiniVNA (my own) against a calibrated rented VNA at work (with rented quality baluns for same) - for passing data over 600m twisted pair.
    The performance was so similar we returned the rented items short of time and continued with the CCTV baluns and Min VNA.
    I read much of the notes you linked to from your excellent site and tried various combinations. The biggest single improvement I made was the short RF earth at the house and the Kiwi in an all metal case. If you look at the image on my SDR the earth strap is just visible bottom left, the other earths were tried and for the MiniWhip, in my location, both ends grounded work well.
    Oh and you may be right about the affect of back to back on higher frequencies as the MiniWhip was not a great performer up high and with few signals (other than QRM) is was hard to quantify (I don't recall earth shattering losses but I don't remember exactly what I compared it to). I know swapping between 75ohm double shielded and the CAT5+balun run didn't seem to kill signals just affect SNR and most of my testing was done by ear rather than test device.
  • @Powernumpty

    Yes most of the Mini-Whips 'run out of steam' on frequencies above about 10MHz as the input capacitance of the FET starts to present significant loading at the amplifier input.

    They seem particularly popular on the German KiWi's, but don't really work too well other than perhaps on the LF and MF bands.

    In many cases using a 1m whip in place of the Mini-Whip 'plate' antenna can help (providing you don't have any really strong BC stations in the vicinity) as it increases the antenna element capacity from about 2.5pF up to around 12pF. This makes the ratio of FET input capacitance to whip capacitance much more favourable, and improves the overall S/N ratio that can be achieved on the HF bands.


    Martin - G8JNJ
  • Ok so I applied my AM filter to it last night and it did improve things but I'm still in overload. If anyone wants to check.. I highly doubt it but I actually have 2 AM bandstops... and wonder if I put both of the inline if it would help. I'm also going to add the ferrite tonight. But those two AM stations are killing me. 760 has always had the nickname the "Blowtorch of the midwest" over 70k Watts of power.
  • Stick a t-piece and a 50ohm termination on it (purists splutter as they compose replies) mines like that.
    Sure it affects the antenna match and probably degrades reception but by slightly less than overload does (here).
  • I wonder if the active stage in the whip is getting pickled?
  • Powernumpty Yeah I heard those braided grounds are much better. I'll look into those. The layout of my house isn't too great for vertical poles unless I get an 8ft Tripod and add another 16ft mast or something. I may need to put another separate tripod out there at the edge of my house but then it would block half the signals as the cable would run down the side of the house.

    Well the permanent location is going at work on a tower away from the 2 stations that are causing the issues. I'm just trying to use / learn while at home. I was more thinking of putting the box with the beagle at the roof but since you need the long antenna to help the signal that wouldn't help.

    So I thought a magnetic loop may be a better all around LF-HF alternative for receive only. Cost isn't really an issue for the permanent location but while I am learning I thought a mini whip ($30) would be a good start to learn.

    WA2ZKD - All the ends are nicely done but it goes from BNC to Type F Rg6 down to another type F to BNC into the box I posted there. I've tried both 5v and 12v power supplies. Although I wonder if 5v may be better?
  • Hi Jim,

    "I wonder if the active stage in the whip is getting pickled?"

    I don't think so, from memory the two BC stations were peaking about -20dB on the waterfall, so I'd have thought that a pair of 10dB notches would have fixed it.

    Unfortunately Lonecrow's KiWi doesn't seem to be on at the moment, so I can't double check.


    Martin - G8JNJ
  • yeah it been rebooting. Going to get the wife to unplug
  • Ok its back up today. Was using it last night. AM filter is in and I put ferrite on the RG6. The OV is less prominent but still popping off at several spots. I noticed around 50Khz there is a massive unknown signal especially at night. Also at 40Khz (combo of 2 stations). So I have zero electronics gear here to make anything of my own. Anyone suggest a kit or something to get started? I mean I have soldering iron and wire. But literally no experience in electronics. I do follow instructions and learn fast though :) Is there some way to target am 800 and 760?
  • I posted some howto in another thread for making AMBC notch filters
  • I think I saw that. I really need an electronics kit and a 101 on this. I mean I understand the principles of resistance and volts/amps. But never took any classes. The schematics usually are a little over my head.
  • what AM station is causing you fits?
  • Suppose you want to notch 800KHz >10dBm. You could place a series LC circuit consisting of a 68pF capacitor and a 575uH inductor from the signal input lead to ground. A calculator to help with this determination is here:

    You can buy parts at and many other places. For example check out the part #143-10J12L on the data sheet located here: and the capacitor could be a ceramic disk like this one:

    It is good to either use a variable inductor or a variable capacitor to help center the notch on the strong signal. You do this by adjusting the variable while observing the amplitude of the signal on the KiwiSDR and stop when it is at the lowest point.

    You can change the values of inductance and capacitance to select parts that fit into your enclosure better. for example a 2000pF capacitor and a 20uH inductor. In this case you would probably wind the inductor on ferrite toroid and adjust the winding spacing or number of turns to find the inductance value. The capacitors will be +/- if they are fixed capacitors and it helps to have a LC meter to see what they are to get close to the resonant frequency.
  • adjusting high-Q notches with the kiwi is a bit tricky due to latency
  • @Lonecrow

    Send me a private email.


    Martin - G8JNJ
  • Being the main "RF" person for the Northern Utah WebSDR, we had a significant AM BCB issue at our site with four 50kW stations (most of them daytime) within about 100kM or so, resulting in some of their signals being in the -20 dBm area on the receiver end of the antenna coax with the total aggregate signal power being higher than this.

    To address this, I probably went a bit overboard and built this:

    While targeted mainly for the 8-bit RTL-SDR, it applies to the Kiwi as well (It's been there and accessible via the web, but not yet "public" on due to networking issues that make even a reverse proxy or DDNS unworkable). The intent was to allow a cheap RTL-SDR cover 630 and 160 meters with sufficient gain to hear the band's noise floor yet allow the reception of the local AM BCB as well - a tall order since this represents more or less 70dB of total range. As can be seen from that article, several tricks were pulled, including:

    - A band-stop L/C network that provided >30dB attenuation over the entire AM BCB
    - Individual notches for the "powerhouse" stations. We can afford to "lose" 30-50dB on some of the strongest stations and still be able to hear them fine.
    - A filter "bypass" adjustment to limit the absolute attenuation. While the band-stop provided about 30dB at the edges of the BCB, there were places in the middle that were attenuated >70dB and any AM stations there would be lost. This "bypass" allows one to set the levels of the weaker local stations so that they can be heard and in our case, it is set at around -17 to -20dB.

    The result was that we could use the lowly RTL-SDR dongle on both AM BCB and 160 meters - and that can be seen today if one goes to the WebSDR. During the day there are a few IMD products scattered around, but these seem mostly to be from nonlinear junctions on the antenna/guying itself and the rusty barb-wire fence nearby as evidenced by their changing if they are shaken.

    The HF RTL-SDRs being used "directly" (no converter) are connected to the output of this filter/amplifier, along with the KiwiSDR and the sensitivity of the Kiwi is "noise limited" at the antenna port (e.g. swapping the antenna for a dummy load results in several dB lower noise on 10 meters when the band is quiet) indicating that we should be able to hear anything that there is - but it took a bit of amplification to achieve this.

    Another element that I had to install is described here:

    (No attempt was to optimize the return loss as it was placed after amplification and connected with fairly short cables.)

    This effectively reduces signals below 8 MHz or so by 2 s-units (12dB) to suit the very broadband and "flat" antenna like the one that we are using on site. This problem stems from the fact that the total signal power at the lower frequencies (say, below 8-10 MHz) is very much higher than they are above this - at least during this (low activity) part of the solar cycle.

    The result was that when the system gain was adjusted to just allow the natural 10 meter noise floor to be heard, on summer nights the very loud static crashes - which were strongest below about 8 MHz - would often saturate the ADC, resulting in these crashes being apparent across the entire spectrum as artifacts.

    The addition of the above filter has largely eliminated this - except in the case of local storms (nothing to be done about those, anyway...) and there is no apparent down-side to this "selective attenuation" as all signals (even the already-reduced AM BCB) are still perfectly audible.

    It is likely that with something like a Mini-Whip, which tends to lose steam above 10-15 MHz, could benefit from something like this same problem could be occur: Already-strong signals at the low end, too-weak signals at the top end and additional gain to make up the deficit at the high end would be out of the question: Knocking down comparatively strong lower-frequency signal levels quite a bit will still keep them above the system noise floor, but "hog" less of the A/D's total RF power handling capability.

  • edited August 2018
    Hi Clint,

    Interesting reading, and I 100% agree, as it's pretty much what I'm doing too.

    I use a genuine broadband antenna, the TC2M, HF Vertical G8JNJ.pdf

    I don't have the luxury of selective resonances on the amateur bands, I get everything at full blast, including all the shortwave broadcast stations that tend to be particularly loud in Europe during the evening and night.

    The KiWi Noise figure is about 17dB

    So it ideally needs something to 'pep' it up on the higher frequencies. Especially if you are lucky enough to have a quiet receive site and a good antenna to start with.

    I use this pre-amplifier

    And this equaliser circuit.

    More info about notches can be found on this webpage

    and more info on active antennas can be found on this webpage.

    One other thing worth considering if you use a good active antenna or pre-amplifier with good intermodulation and strong signal handling performance. They can generate very high levels of RF at their output under saturated power conditions often more than 1 watt. So if you are an amateur who occasionally transmits, it's well worthwhile putting some form of RF limiter or receiver protector on the KiWi input to limit the maximum RF power that can be applied to somewhere around +10dBm. Normally a pair of back to back 1N4148 diodes will do the trick, and if used in a 50 Ohm system they will not generate any significant intermodualtion products that wouldn't already be noticeable on the KiWi if the signals were that high to start with.

    Generally speaking it's a good idea to try and keep the highest level signals seen on the KiWi spectrum display to be just below the -30dB as this give a bit of margin for changes in propagation conditions, anything peaking up to -20dB is likely to start causing problems.


    Martin - G8JNJ
Sign In or Register to comment.