Is this a common-mode interference? Seek for a solution ..

Hi dude,

i spent a little while trying to resolve the interference patterns shown in the picture (in the interval of 1.7~5.0 Mhz)

Currently, a Galvanic isolator has been implemented and works well for LW band noise.

A low-pass filter (DC-30Mhz) installed, and it has been proven nothing to do with this interference patterns.

No balun nor a so-called common-mode choke installed yet in my antenna.

Could a common-mode choke be useful for my problem ?? or such device would be of no use ??

just a amateur noob, not so familiar with antenna theory ....

Need your kind opinion for this very much.

Regard,

Jimmy


Comments

  • b.t.w, the "period" of these interference lines was seen precisely 220khz.

    hmm... is it an useful hint ?

  • Hi,

    It's probably interference from a swicthed mode power supply.

    A choke balun may help, but the best first step would be to determine if it is being produced by any equipment in your property. Most interference is usually made by something very close.

    What antenna are you using ?

    Regards,

    Martin

  • Hi G8JNJ,

    Thank for your kind feedback in advance.

    The advice from kiwisdr.com was taken for my power supply choice:

    Mingwell GST25A05-P1J 5V 4A type, which was said as a SMPS found to be fairly quite. It is a trustable brand and so far no eligible tester in hands to verify its quietness.

    Antenna: Apexradio 303WA-2 with a shielded coax-cable 18m in length plus a LPF + Galvanic isolator installed. With grounding at the front-end, without any lightning protection.

    I'll keep trying any possibility including balun/ choke and report back here as soon as I find something. 😃

    Regards,

    Jimmy

  • I read an article which said

    For example, for low-frequency electrodeless lamps, the switching frequency of

    the rectifier is 220KHz.

    This frequency is the fundamental frequency of interference, and other interference frequencies are basically high-order harmonics of this frequency point.

    In the initial design, the requirements of the filter can be given according to the estimate Come, when the frequency is 220KHz, the common mode insertion loss ILCM=60dB and the differential mode insertion loss ILDM=60dB. According to this requirement, the filter manufacturer can design the filter.

    Guess that the final answer seems to be getting clearer ..... 😅


  • OK, some more questions and suggestions...

    Where is the antenna mounted relative to your property ?, as it may be picking up noise directly. Even if your KiWi power supply is noise free, I can almost guarantee that other devices in your home will not be.

    Using a short whip antenna connected to your KiWi via a decent length of coax, will allow you to 'sniff' for interference around various devices in your home, and it should allow you to identify specific noise sources.

    Ideally you could run your KiWI from a battery and connect it directly to a laptop running a browser, and then drop the main breaker in your home, just to see what difference it makes to your noise floor.

    However the main problem with unbalanced antennas, is that they often rely upon the coax to provide the 'missing' half of the antenna. So if you suppress the common mode signals by adding chokes, you can also reduce the wanted signals too. The trick is to provide an additional 'noise free' signal path for the 'missing' half of the antenna and choke off any noise on the coax cable.

    This is usually achieved by adding an extra set of radial or counterpoise wires connected to the coax screen at the base of the antenna, and then adding a choke balun on the coax at the base of the antenna, but on the coax 'radio' side of the connection to the radial wires.

    You may obtain better results with a broadband loop antenna, which in addition to having better intrinsic common mode rejection, it can be rotated to provide a null in the direction of localised unwanted interference. The Wellbrook & LZ1AQ designs (and clones), and Cross Country Wireless loops being probably the best options.

    Good luck,

    Martin

  • Hi Martin,

    Thanks for your valuable opinion. It works. 😀

    Tested by battery power supply, the result revealed that the interference stripes with a period of 220khz did come from the (MingWell ) SMPS, just as you guessed (thumb up!). Well i dont plan to buy a linear power supply right away as a replacement (a bit risky..).

    Instead next, some electric improvement will be taken into account toward the SMPS DC output wire prior to KiwiSDR DC-in. Let me study again the relevant stuffs ...

    So in a way of speaking this SMPS is "very quiet" but not good enough for a SDR.

    Regards,

    Jimmy

  • MingWell or MeanWell? which model?

  • Hi, WA2ZKD

    My typo. It's MeanWell model no. GST25A05-P1J ( 5V/4A )

    FYI & regards,

    Jimmy

  • Hi Jimmy,

    Ha, perhaps it is a 'Ming well' supply after all :-)

    In my experience Meanwell supplies are usually pretty good. But with all of my Switched Mode Supplies, I tend to modify them by adding a large electrolytic (ideally greater than 2200uF) and a SMD 0.1uF directly across the DC output terminals, and I then wrap the DC cables though a ferrite ring (usually as many times as possible, which is normally limited by the wire size), to form a common mode choke, before they connect to the item to be powered. This usually bring emissions down by a further 20dB or more, so it's definately worth doing.

    In fact I've just spent the past day modifying some ex-computer switched mode supplies in a similar manner, in order to provide clean 12v & 5v DC outputs for use at one of my web sdr sites.

    Sniffing around the supplies with a spectrum analyser and some close field probes, allows you to quickly identify where RF noise is escaping from, and you can then usually cure it by adding capacitors and self adhesive screening foil or tape.

    Regards,

    Martin

  • don't overlook possible radiation from the mains input... ferrite there may help too.

  • Hi Martin & WA2ZKD,

    As suggested, four ferrite rings and a very long DC 5.5/2.1 extension cord (not received yet) are on the way. Plan to check the attenuation for high-order harmonic noises by these first.

    Then, connecting a large capacitor in parallel would be an addition also... if being not satisfied.

    Thank for your help again !

    For safety, an arrester was connected at the front-end ground last night. Neither perceivable signal attenuation nor introducing of new noises have been observed. Just keep it going on, hmm.

    There is truly a lot of knowledge in optimizing reception. Will update my status for everyone here while my CMC trial-n-error is done for the DC supply wire.

    B/R,

    Jimmy

  • edited November 24

    Hi Jimmy,

    I may be wrong, but to me the green ring in your photo doesn't look like a 'ferrite' ring, but it is more likely to be 'iron powder', as they are normally coated in a shiny epoxy paint and colour coded according to their properties. They are commonly used as energy storage inductors in switched mode power supplies, and often mistakenly get sold as being ferrite when they are not.

    The problem with nearly all cores is that unless they have some markings on them, you can't tell what they really are unless you can somehow measure their properties.

    Unfortunately not all manufacturers use the same colour codes, but the most common types found in power supplies are coloured Yellow (or Yellow/White or Yellow/Red) which is type 26 Iron Powder (not to be confused with simailr numbers used for ferrite core materials), and Lime Green (Lime Green/Red) which is type 52 Iron powder. Dark green is sometimes used in mains input filters (to reduce the amount of noise from the switching circuit on frequencies typically <100kHz being fed back into the mains supply), but I'm not sure what the material actually is, or how suitable it would be as a common mode choke on frequencies >1MHz.

    Ferrite cores are usually not painted, and are various shades of grey. Some manufacturers do coat them, but the paint is usually a matt epoxy and the colours tend to be pastel shades such as white, pink or light blue.

    The best types of ferrite to make common mode chokes from are mix type 43, 31 or 77 (for lower frequencies).

    The type of material and number of turns can be optimised for particular bands of frequencies, and the chart on this webpage can help you decide what is likely to work the best.

    Although the design data is intended for antennas, it also applies to mains and DC cables, in fact it will work for anything you wish to apply a common mode choke to.

    As nearly all ferrite rings look the same, they can easily get mixed up, even by distributors and re-sellers. You can determine the type of material, by using an antenna analyser, or better still a cheap VNA, but that's a story for another day.

    Regards,

    Martin

  • yes Martin, I was going to say the same about powered iron vs. ferrite

  • things like this, which are incorrect, confuse matters


  • Ha yes, their description is confusing too.

    Green Iron Power Ferrite Toroid Core 63mm Outside Diameter

    Specifications:

    Product Name Ferrite Core

    Material Iron

    Color Green

    Overall Size63 x 38 x 25mm/ 2.5" x 1.5" x 1"(Outside Dia.*Inner Dia.")

    Net Weight 233gPackage

    Content 1 x Toroid Ferrite Core

  • Hi Marti & WA2ZKD

    Thanks.

    An Nd-Fe-B magnet can hold the green ring easily, but not so attractive as that to an iron plate.

    As both of you told these rings should not work, luckily they were really cheap ones. Well, forget it.

    There is lot of pictures of chokes, indeed, the correct ring body is seen to be mostly dark gray.

    I decided to go to an electronic material retailer today rather than a net-shop.

    If the improvement of the DC power supply takes effective, i will update here asap.

    check the progress out real time for 110, 330, ... 1430khz and all these stripe lines in 1.6~4.5Mhz

    @ http://jjm0311.ddns.net:8073/


    best regards,

    Jimmy

  • edited November 25

    Hi Jimmy,

    OK sounds good.

    Iron powder ring cores do have their uses, and may make some difference to your level of interference. However if you can get hold of the correct type of ferrite rings, they are much more likely to be effective on the short wave bands.

    Update - I just took a listen to your KiWi and it's already pretty good. I can see some broadband noise around 3MHz, but even so, the Signal to Noise ratio is pretty good, and better than a lot of the other KiWi's I have tried in in the Pacific region around Japan.

    I've been trying to use a lot of these in the past few days, in an attempt, using the TDoA extension, to identify Chinese coastal radars. (For example I can hear at least four of them on your KiWi, buzzing away and using up about 60kHz bandwidth each on frequencies between 3.5 & 3.8MHz). But for this to work I need to use KiWi's that have got good GPS fixes.

    It would be great if you could also somehow improve your GPS reception. Placing the GPS puck antenna on a metal (radio reflective) surface such as a metal biscuit tin (or similar) can often make a huge improvement.

    Regards,

    Martin

  • Hi Jimmy,

    Regarding the interference that can be seem around 300kHz (and some other frequencies).



    I think the two light blue wiggly signals are from the same source, as the changes in frequency occur at around the same time as each other. The periodic step changes in frequency are typical of something like a phone charger, when the battery is nearly fully charged they tend to exhibit this sort of behaviour, as it flips between being on and off load.

    The bright green signal has a 60Hz (I think) component, so it's another switched mode power supply, but this time the load it is powering is much more stable.

    Regards,

    Martin

Sign In or Register to comment.