I recently build a Loop antenna for CB radio, or at least i tried.
Its made out of a 80cm diameter Loop of RG58 Coax (shield and core connected at the ends), a Coax stub condensator and a unshielded wire primary loop.
When i put my SDR on it, it seams to have way to much of a wide reception (calculator said it would have only like 40-50khz wide reception band).
When i put my analog power/swr meter on it, it claims to have a SWR of 1.2 and takes about 3.5W of power (compared to my dipole taking 4W).
But when i put the NanoVNA on it to get a more accurate reading of SWR, all i see is a flat line that claims a SWR of about 50.
When i pump up the stimulus frequency up to 300+Mhz i get some SWR dips there down to 1.6, but i assume thats just the Primary loop resonating.
Any idea why i get results on my analog SWR meter but not on the NanoVNA? Is the NanoVNA maybe putting to few power into the loop to make it resonate?
All good points, especially calibration. Without, it still gives you a general idea, but it’s more a “meh looks OK” than a serious value.
Also, as fullsquare said, take a picture of yhe smith chart / impedance display as a second check, if it shows something like alot + j toomuch, you know your SWR is actually bad, and you know what direction you need to go for (add more capacity, inductance, or some resistive transformation).
Here is the measurement with smith chart:
There is 2m of BNC RG58 coax between the antenna and the NanoVNA. I calibrated the NanoVNA without the coax line because i was under the impression the line to the antenna is considered part of the radiating system and hence should not be calibrated out. But here is the reading with the coax line to the antenna calibrated out:
I’d argue that your capacitor and loop are OK (otherwise there would not be a dip at 27 MHz).
But my guess is that the match to 50 Ohm is bad. Try playing around with your feeding loop.
You can ignore the cable if 1) it is properly matched to 50 Ohms on both sides and b) there is no field outside of the cable, meaning the cable does not act as an antenna itself.
You are clearly working on the 1) part, and for the 2) part, it depends on how you feed the loop, unbalanced vs balanced. For coax you either need an unbalanced feed loop, or a balun (“balanced-unbalanced” converter) or a ferrite or similar “Mantelwellensperre” (don’t know the English word, sorry).
I love the word Mantelwellensperre, peak german language :D
I build a Air-Core choke from a part of feed line right below the antenna now, and got some slight improvements.
I will try to improve the feeding loop position to see if it gets the swr down a bit more. By now i am already quite happy with the antenna, during a test i managed to make a ~15km Thor8 connection while the antenna was still inside the house (with 4W AM Tx and the receiver in SSB/USB).
Yeah this lower one looks better but still probably your capacitor value in loop is way off, try to find frequency where impedance is real (purely resistive; green line on smith chart crosses horizontal line in the middle) and work from there, then you’ll know whether to increase or decrease it.what LH0ezVT said makes more way sense than that, i forgot how magloops work. but you still might want variable capacitorresonance is narrow so you might miss it. there’s a reason why magloops are made with variable capacitors (sometimes retuning is required due to changes in ex. humidity)
how have you made your capacitor anyway?
you can put some ferrite beads on your coax close to feedpoint in order to eliminate common mode currents. better yet, use a balun. this might help you in getting more reproductible results
e: note how swr gets much higher when off resonance with properly calibrated nanovna. when measuring antenna with cable, you’re seeing loss in cable as a degree of lowered swr but only with high swr, because energy is lost in cable when it bounces around and never goes back to nanovna
This is my capacitor (RG58):
I dont have ferrite beads right now sadly, but i will try to make a Air-Core choke from parts of the coax feed line, maybe that helps.
If i would use a balun between feed line and antenna, i need a 1:1 balun i assume?
I will try to find the frequency where the impedance gets real. Also gona try to maybe build a variable capacitor from 2 metal pipes going into each other depth regulated by a screw, but i mostly wana use this antenna on a single frequency so i hope i dont need much adjustment.
Just a remark, remember that everything that you connect to the radio and which is not matched on both sides will have an effect that the radio “sees”.
So if you get a different result with and without calibrating in the cable, and the cable is used for the radio connection as well, the vna result that doesn’t include cable calibration shows what the radio will see.
Ferrite beads allow you to use old calibration. If you make 1:1 balun just by threading coax through toroid, you can use old calibration as well provided it’s the same coax. Keep in mind minimum bending radius of coax. There are other designs, like using twisted pair on toroid, then you have to include balun in calibration as well (it adds some electrical lenght). If you noticed changes after making air core, this suggests that you do have some common mode current, this will make your measurements sensitive to random changes as rf current flows on the outside of cable where it shouldn’t
I’ve seen people using PE-Al-PE pipe for variables, this gives you layer of good dielectric (polyethylene) (but not as good as air) in dimensionally stable form. One connection is aluminum layer inside the pipe, and for the other you’ll have to figure it out on your own. Retuning might be required anyway within the band (magloops are narrowband) Common way to make variables is to bolt two of them in series, so that no sliding contact is used, moving part is the same for both. This is good for high voltages also but i’m not sure if you’ll need it
i’ll add that in a way SWR chart is more resistant to misuse, because if nanovna is calibrated with wrong length of 50 ohm feedline, or without feedline at all, then smith chart will be rotated by angle depending on difference in length of that feedline, while SWR chart should look the same. for example, if real part of impedance at resonance is too low (ex. 20 ohm), and feedline is quarter wavelength different from what nanovna was calibrated with, then impedance will be still real but too high (ex. 125 ohm), while SWR chart should look the same (1:2.5 SWR minimum) (barring losses in feedline). (this works the same way as quarterwave long feedline impedance matching scheme). for different feedline length differences (non-multiple quarterwave) impedance will be complex at antenna resonance. this problem is avoided by calibrating nanovna with feedline