Home Brew Broadband Magnetic Loop

This LNA was discussed in an earlier thread and I decided to see how it might work with a home brew loop. https://www.amazon.com/gp/product/B01N2NJSGV/ref=oh_aui_detailpage_o00_s01?ie=UTF8&psc=1

I spent a total of $25.88 for parts. I had the PVC pipe for support and coax cable in the junk box. This little LNA connected to a loop made from 14' of coax and powered with a 9vdc battery is providing signals from 5KHz to ???, well at least 30MHz for the KiwiSDR. It has been tested, but is not currently online as I'm working to "ruggedize" the LNA in a small pipe.

I think an inexpensive antenna such as this could improve some KiwiSDR installations. It isn't perfect and displays more repeating interference spikes than I would like, but it is still better than many other antennas I've listened to. Not quite as good as the W6LVP but it is better below 500KHz and probably equal up to 18MHz. It needs more comparison testing and the loop circumference might make a difference.
Ron
KA7U

IMG_20180122_193503941.jpg
4008 x 5344 - 5M
IMG_20180122_193024086.jpg
4008 x 5344 - 4M
IMG_20180122_193330382_BURST001.jpg
4008 x 5344 - 8M

Comments

  • edited January 24
    I tested this loop quite a bit for the last two days. It is not on par with W6LVP loop, but it does work better than my miniwhip and other small antennas that I've tried. So on a scale of 1 to 10, with 10 being my 210' top dipole, the W6LVP rates a 8 and the home made loop with the HiLetgo amp rates a 4. What we need is a Low Noise Amplifier made specifically for broadband loops at a price point similar to the HiLetgo LNA.
    Ron
    KA7U
  • So today I changed the power to the LNA. I found that 9.25vdc and 10.5vdc results in the lowest SNR. At 60KHz 10.5vdc is optimum and at 15MHz 9.25vdc is optimum. 10.5vdc seems to be the good voltage overall if you want good results at VLF. This might vary for different LNA boards, but I doubt it. I used a variac to change the voltage on the wallwart to find the best voltage.

    This antenna is producing lots of interference spikes. Remove the antenna, the spikes go away. Remove the amplifier but not the antenna and the spikes still appear at a lower level. All the signals are still available just at a lower level. Add the LNA and the signals stand up and are easily received by the KiwiSDR over 100' of coax. This is more subjective than objective, but it seems to me that once the voltage input is optimized for 10.5vdc, this homemade antenna is about 2-4 dB less SNR than the W6LVP loop. So the W6LVP loop is the winner, but not by a huge SNR margin.

    I have not provided a balun to feed the amplifier and I have not followed all the "Tips", (only the voltage observation), provided by the seller, which follows:
    Tips:
    1.The amplifier at operating frequencies less than 500MHz, has a very good gain flatness, careful adjustment can be 1dB lower frequency, the higher the gain consistency
    2.The lower limit of the operating frequency of the amplifier is subject to the input and output capacitors, the default is 0.1uF, can work to 0.1MHz; can be appropriate to increase the input and output capacitors to expand the lower frequency limit,Such as: 10uF, you can work to 5KHz. For detailed calculations, pay attention to the input and output impedance of 50 ohms, calculate the required capacitance value, while enhancing power supply decoupling and voltage regulation.
    3.When the supply voltage varies from 5-8V, it can be used as a variable gain amplifier. The gain increases with the supply voltage. Ideal for RF receive front-end circuits, with DA control supply voltage, thereby controlling the amplifier gain, automatic gain control. When the supply voltage at 8-10V, the low-frequency gain of 30dB, this amplifier has a lower noise figure and better stability. 


    I might eventually change the input and output caps, which are surface mount and I have to order in, and possibly add in a balun to feed the amp, although I doubt it will add value.
    Ron
    KA7U

  • Since JKS organized the antenna and interference related threads to their own section, I found that JKS had posted this link:

    Reading that and noticing that my home brew loop is showing these spurs at a high level is spurring me to action. I'm planning to upgrade my network to shielded cable and switches and possibly relocating the KiwiSDR antenna which is most affected farther away from the shack. 

    Now back to why would the home brew loop produce higher amplitude spurs than the W6LVP loop? I am thinking it might be because the electrical characteristic of a shielded magnetic loop ( one end has shield and center conductor both common to the ground, and the other end uses the center conductor only to the input of the amplifier) should be an unbalanced antenna. The HiLetgo LNA is an unbalanced amplifier, the net result might be less cancellation of the 60KHz network spurs on the coax feedline?  The W6LVP loop amplifier is potted and a schematic is not provided but, George Smart https://www.george-smart.co.uk/projects/wellgood_loop/ , has a schematic of balanced LNA and the input to that is a tube which is a balanced input. The attached image shows that George Smart's KiwiSDR does have 60KHz spurs displayed at various amplitude, but they are on the order of magnitude as the ones produced by the W6LVP loop, which causes me to think a balanced LNA and a balanced loop will not be as strongly affected by the network spurs.

    I'll follow up on this topic after I upgrade the local network here at KA7U.
    Ron
    KA7U
    6KHzspurs03.png
    1422 x 835 - 1M
  • Hi Ron,

    I think that the problem is almost certainly loop imbalance. 

    You can check this by rotating the loop to peak a known station. Note the signal level then rotate the loop 180 degrees and note the signal level of the peak in that direction. 

    If the loop is properly balanced the signal levels should be the same. If it isn't the peaks will be different as the imbalance means that it is now configured as an Aperiodic loop. 


    Which are often used for phased receive arrays like this commercial version.


    I found that the LZ1AQ design was easier to build than the 'Wellgood' (Wellbrook copy) and performed slightly better.


    I fed mine with coax and left out the input filter network and voltage regulator (works OK on 13.8v) 

    The Wellgood and ZL1AQ work OK up to about 10MHz then run out of steam. if you want better performance on the HF bands the only solution is to make a 'fatter' loop or connect multiple loops in parallel in order to reduce the loop inductance. 



    More stuff about loops etc. on my active antennas webpage http://www.g8jnj.net/activeantennas.htm

    Regards,

    Martin - G8JNJ



    KA7U
  • edited January 30
    Martin,
    Now that I've read through the information you provided and even more that Google was listing, I think I'm getting the scope of loop antennas. Hi Hi. 

    The Hiletgo amplifier is actually doing a reasonable job on the home brew shielded magnetic coax loop that I've been experimenting with. The gain is more than the W6LVP loop but it does not have as good a noise figure. I find that if I open the top of the loop and cut the shield and center conductor, cut coax in half, and then solder the center conductor which becomes the LNA input to the shield which becomes the LNA ground on the opposite side of the loop, make the shield continuous at the bottom of the loop, then the loop quiets down and the various spurs and images disappear. I think a better way to feed the Hiletgo LNA would be to make a loop out of copper tubing, place a T in the bottom of this loop and then cut out 2 or 3 inches from the top of the loop. feed a coax inside the tubing, solder the braid to one side of the tube and the center conductor to the other side of the tube and then connect that coax to the Hiletgo amplifier. I might give that a try using 1" copper water pipe. 

    I've ordered 3 Wellgood boards from Mr. Smart and the parts from Mouser Electronics. After those are assembled and working I'll report back. 

    My quest for an inexpensive loop antenna is getting expensive.
    Ron
    KA7U
  • The home brew loop with the Hiletgo amplifier is finally working adequately. So here is the setup that finally put it into respectable service. The idea that a working amplified loop antenna can be made with retail parts for under $50 is now demonstrated.
    • 1 meter diameter coax loop.
    • coax is cut in half and the center conductor that will feed the input to the Hiletgo amp is soldered to the braid of the half that will connect to the ground of the Hiletgo amp.
    • The center conductor on the side that is to be ground is soldered to the coax braid at the point it connects to the Hiletgo amp.
    • The center conductor that connects to the input of the Hiletgo amp is not connected to the braid on the coax that shields it in the coax.
    • The braids of both coax sides are soldered together at the feed point of the Hiletgo amp.
    • The power supply is critical and I now find 12.13vdc is optimum and needs to be a clean DC source.
    • check the attached images for better clarity.
    • The screenshot of a waterfall on the left is a 210' top dipole without an amplifier, the waterfall on the right is the shielded magnetic loop with the Hiletgo LNA. The settings for waterfall gain are the same for both displays.
    I'll leave it in service at http://ka7u.no-ip.org:8073 if you have an interest, try it out. 
    homemadeloop06.png
    1586 x 720 - 815K
    IMG_20180130_170754655.jpg
    4008 x 5344 - 5M
    IMG_20180130_170716366.jpg
    4008 x 5344 - 4M
    IMG_20180130_170736269.jpg
    4008 x 5344 - 5M
    IMG_20180130_170724817.jpg
    4008 x 5344 - 5M
  • Here is a schematic of the loop. Please forgive my artwork.
    Ron
    KA7U
    New Doc 2018-01-31.jpg
    1608 x 1600 - 63K
  • A note about the split in the shield at the top of the loop antenna shown in the illustration above. The center conductor can be left continuous. It does not seem to make much difference in the antenna if it is connected to the shield or continues around the loop without a cross connection.
    Ron
    KA7U
  • Hi Ron,

    I think the loop is still unbalanced.

    The split is to avoid a continuous 'screen' (Faraday screen to minimise E-Field pickup) around the loop. But if one end of the center conductor is connected to ground you may have a problem.

    Check signal levels with the loop normal and 180 degrees as suggested previously.

    A 1:1 transformer on the loop input may help improve balance. Try six bifilar turns through a BN73-202 core as a starting point. One winding as the (balanced) loop input (primary) the other as the output (secondary) unbalanced connected to the amplifier input, with one end of the winding connected to ground. This will also provide galvanic isolation.

    You can also try a few more turns on the secondary as an experiment as this may improve reception on the LF bands.

    If you made a balanced loop amplifier with a much lower value of input impedance (LZ1AQ or Wellbrook copy) you wouldn't need the loop outer screen to minimise E-Field pickup.

    Regards,

    Martin - G8JNJ
    KA7U
  • Martin,
    I believe the loop is unbalanced as currently connected. My next iteration will feature a balun to feed the LNA, the center conductor won't be grounded, and I'll check to see if it has a null in the axle directions.  My parts from Mouser Electronics have arrived, and the three Wellgood boards are in the mail and I'm looking forward to them arriving. Mr. Smart is sending his Bias-T boards as well, although I haven't yet ordered parts for those boards, when I do I'll order in several BN73-202 ferrites. Mouser has a 5 week delivery on them, but I do find them for twice the price elsewhere.

    My goal on this loop discussion is to build it for under $50.00/35.40£.  I'm at $42 now, so there is room for a balun. Hi Hi The Wellgood boards with the Bias-T boards and parts will be under $50 as well, as long as a few are ordered to reduce the postage, or a batch of boards could be locally etched I suppose.
    Ron
    KA7U


  • Related stuff here. I imagine the low end of that balun doesn't do well at LW but the board could be the base for alternate toroids


    KA7U
  • This photo is the cleaned up version of the loop being discussed. The 1/2" PVC pipe is not glued, relying on compression to hold together. This makes it easy to break down and use portable. The Hiletgo LNA works fine with a 9vdc battery.
    Ron
    KA7U
    IMG_20180206_093343063.jpg
    4008 x 5344 - 7M
  • edited February 10
    It occurred to me to check the relative antenna / LNA sensitivity this afternoon. I connected a 50 ohm load to the transmitter and set it for 5 watt carrier output. Transmitter is 100' from the loop antenna. The loop and Hiletgo amp provided these results into the KiwiSDR. The second column of numbers is a commercial loop for relative comparison. Remember the numbers are negative so the smaller number is more gain, the larger number is less gain. The smaller the noise floor "NF" number the better, the higher the NF number the worse.
                          Hiletgo       W6LVP
    28.5MHz    -   -69dBm       -60dBm   note: The average NF on the Hiletgo is -122dBm, the average NF on the W6LVP is -115
    24.9MHz    -   -63dBm       -70dBm
    21.2MHz    -   -48dBm       -50dBm
    18.1MHz    -   -44dBm       -44dBm
    14.2MHz    -   -54dBm       -55dBm
    10.12MHz  -   -53dBm       -60dBm
    7.15MHz    -   -57dBm       -75dBm
    5.33MHz    -   -59dBm       -76dBm
    3.75MHz    -   -67dBm       -77dBm
    1.9MHz      -   -90dBm       -90dBm
    WWVB       -   -98dBm      -94dBm   note: NF Hiletgo -127, NF W6LVP -122

    The Hiletgo antenna impedance measures low < 50ohms on 21.2 and 18.1MHz. This would make me think the Hiletgo does better with a low antenna impedance.This antenna is dropping dramatically at 1.9MHz which is the lowest transmitted signal used. The antenna does receive the MF broadcast band quite well and so it might be a spot in the system that attenuates more than others or lower radiated power from the 50ohm load resistor and cables.  The W6LVP output impedance over the band has not yet been measured because the antenna is connected and potted and there is not a practical way to remove the LNA.
    Ron
    KA7U
  • Martin,
    I tested this Hiletgo loop for balance tonight and it is balanced. My test was with RNZI @ 15720KHz so it was skywave and not totally reliable, but the null on either side was deep and the gain off both ends was pretty much equal. So at least on this signal the antenna exhibits balance. I need to buy a rotator so I don't have to run the stairs. Hi Hi
    Ron
    KA7U
  • edited February 19
    Martin,
    I wound a 7 turn bifilar on a bn-73-202 core and connected yellow to yellow on the antenna, and red to red on the Hiletgo LNA (red and yellow being the wire colors). I'm seeing between 5 and 10dBm gain with the Hiletgo LNA versus the good Wellgood balanced LNA. There is a static buildup that discharges with significant consequence when the wind blows across the antenna. I speculate that is caused because the loop is not directly DC grounded. I know you have significant experience with baluns and transformers so, what should I do different? Should I use an auto-transformer? Or should I go back to my original design with the loop split at the top, and the shields bonded and grounded with the inside wire also grounded?  https://goo.gl/JHqTe7  Now that I'm into this, I find another Ham that is following that design.

    I'm thinking an auto-transformer might be the cleanest and simplest way to do this, but the static has to be drained off somehow.  I'm not sure what is happening with the antenna balance, but suspect that if a transformer is used, the balance should be good.
    Ron
    KA7U
  • Hi Ron,

    I think you could either center tap the input transformer on the loop side (prefered method), or add two 1K Ohm value resistors (or similar) from either side of the loop / transformer connection to ground.

    Regards,

    Martin - G8JNJ
    KA7U
  • Martin,
    Thank you for bearing me on this quest. I'm a bit compulsive. So I've now tried several balun types with the Hiletgo and while it achieves good gain the noise floor and images are over the top unless it is connected as a shielded magnetic loop, shield split at the top and shield and one side of the inner conductor grounded at the bottom on one side and the other side of the inner conductor used to feed the Hiletgo LNA. Then the noise floor and images drop down to an acceptable level. So my conclusion is to use a balanced LNA with a simple loop and a shielded magnetic configuration with an unbalanced LNA.

    While I've been experimenting I've learned that the CPVC pipe fittings are not going to hold up with the heavy copper loop, so I'll need to find a more secure support system for the Wellgood loops. PVC works great with a loop made from coax but then there isn't any stress with that material.  

    I'm not ready to post my conclusions about this loop comparison project.., 

    but I'm starting to think about saying that the Hiletgo LNA used with a 1 meter diameter coax loop, shield continuously bonded except at the top of the loop where the shield is opened, the feed point at the bottom of the loop with the shield and one side of the center conductor connected to the ground side of the Hiletgo LNA and the opposite center conductor connected to the Hiletgo LNA input, is a low cost, low stress way to get a decent loop antenna on the air.

    The Wellgood LNA while exhibiting balance, low noise, good feedline isolation, and good gain, is an expensive and stressful design to make. If you are not equipped with a good soldering station complete with a desoldering tool that works well the frustration might significantly increase. I say this because I have one successful build and one sub-par build and tomorrow the suspect transformer will have 8 leads desoldered and the core rewound and reinstalled. This will add significant time to the hobby project. Fortunately my friend K7RHB will do the desoldering work for me.

    The 1" (25.4 mm) copper tubing formed into a 1.2 meter loop adds quite a bit of gain to the loop antenna. I suppose this to be the result of increased loop area and reduced loop inductance. Besides PA3GZK claims 1.2 meter diameter is optimum in practice and who am I to doubt that? If I use the PA3GZK construction technique and use brass fittings instead of CPVC, a sturdy self supporting antenna base will result so add the significant cost of those fittings.

    Long story short, if you build a Shielded Magnetic Loop out of coax and use the Hiletgo LNA, reception should be good if not excellent depending on the coax diameter.  I'm still considering all of the above however.
    Ron Morell
    KA7U
  • Hi Ron,

    I find your comment "Then the noise floor and images drop down to an acceptable level." interesting.

    I suspect that the wide bandwidth of the Hiletgo LNA may be a problem and that it is VHF and UHF broadcast stations that are overloading the amplifier.

    It's not easy to filter these out without adding additional unwanted inductance into the input of the amplifier stage.

    I would take a look at the filter on the input of the LZ1AQ loop amplifier. Although this messes up the LZ1AQ loop performance, it may work OK with the Hiletgo LNA as the input impedance is much higher than the LZ1AQ design, and you may be able to get away with it.

    Regards,

    Martin - G8JNJ
Sign In or Register to comment.