P.S. Mistake in label- Pole not poler, sorry for mistyping

Message:
So, I received a larger stock demand for 868M antennas where for many points the J-Pole was only candidate. What I did to produce em - I calculated it by MOUKD website calculator and got A=241mm, C=80 mm, D=8 mm at vf=0,93 (as I mean mount into the pvc 16mm whitepipe, and 8mm inner spacing into J interior. I used 1,6mm alu wire into pvc isol with 1,1 mm wall, thus total dia 4,8mm. For contacts I used an electrotech mising clamps (cubicles) 5x5x5mm with M3.5 screw, where soldered a coax going to BNC in 90 deg angle and then to N201-SA vector analyser.

From this moment my adventures begun. The resonanses (=places where reactive component of Z changes the sign) there was like forest - zillinons, for ex 939,923,916,884,849, 857,853,810,832 MHz etc etc, having swing from SWR=1 to SWR=20(!). Okay I thought, may slide up or down a coax placement, as higher as bigger ought be a resistance. Nope, here are dataset I seen at one of resonances (where XL=XC=0 thus the Z==R).
H=25mm => 16 Ohm; H=29 mm 10 Ohm, 20 mm 23 Ohm, 16 mm 30 Ohm, 9 mm 33 Ohm, 6 mm 40 Ohm etc etc. How the theory may be violated so drastically. At high up there ought be sth like 200 and 400 Ohm and small values must be as exact in the bottom!!

Okay, I thought, will scan what happens between resonances. Changing the freq, an antenna shown capacitance 2 pF -> 4 pF ->10 pF ->100 pF ->2000pF then jump to 0,01 uH ->1 uH ->5 uH etc. Both as L as C have sth like pseudorezonanse in the middle, where example 4pF 8 pF 12 pF 12 pF 10 pF 6 pF 4 pFetc etc. But anyway those semi points are not good for use as the imaginary part still is high there, say active 50 Ohm, reactive 60 Ohm etc.

So, my questions - which one of resonances I must use there, those where L jump to C or those where C jump to L (re-wording, serial or parallel resonances), or I just ought to tune on the pseudoresonances?
Other, why the height of clamp have rather hard impact on value of reson freq??
Third, why the reaction on R is just inverse as theory tells?
Fourth, what hell here is happening at all? Only true thing out of theory what worked well, is that spacing betw both wires play no role at all.

P.S. Of course I checked this J-Pole as with external pipe as without. Only the effect was few % freq shift strongly in agreement with theory.

P.P.S. Coming home and taking the supper (oh how well it alter the productiveness of mind) I begun to meditate - probably I wrongly measured length? I understood it ought be measured by horisontal projection on vertical measurer tape, taking zero from upper face of U-turn wire side. But probably its wrong interpretation and it must be measured by curvature along the wire, and then 8 mm corresponds to 5 mm - rather big mistake? What is the right method to measure wire length and connecting place length?

PPPS - why aluminium... just project deals with about kilometer of wire, and in case of copper I must buy it from my pocket while on aluminium I may put in a pocket it price. The rolls of it stays unused a half of century at least. And lightweightness plays a certain role as well.

 

Day No 2: at the http://k9erg.tripod.com/theory.htm I found a sentence going across all other I had read, but going along what I seen with my eyes - ""If Co-axial cable feed line for J-POLE is used, a BalUn MUST be used. Attempts to feed this antenna directly with co-ax have met with disastrous results."" I fed it straight from cable with thought the 10 cm long cable is short enough to be ignored.
Thus, I applied the ferrite ring over the cable near the feedpoint and immediately got the R=35 Ohms with X=-24Ohms=8 pF. Thus small coil will cure it, and squeeze the 50 Ohms out of 35 thats little sliding problem no more, hopefully.
Thanks for not answering anything , my thoughts was clarifying by itself.

 

Day No 2: at the http://k9erg.tripod.com/theory.htm I found a sentence going across all other I had read, but going along what I seen with my eyes - ""If Co-axial cable feed line'for J-POLE is used, a BalUn MUST be used. Attempts to feed this antenna directly with co-ax have met with disastrous results.""

Well, yes, you can get some serious currents in the matching section that need the isolation.

 

Hello,

The J-Pole is a type of antenna used by many radio amatateurs.
Here is an other page with a calculator:
http://hamuniverse.com/jpole.html

For 868 Mhz it gives the following results:



The page also tells me that the outer side to the short side and the center to the long side of the antenne.
There are also other models derived from the j-pole, like the slim jim.
The wiki shows some of these other models:
https://en.wikipedia.org/wiki/J-pole_antenna

Bertus

 

Thanks, Bertus, but those are sizes I tried, one by one milimeter between 210 and 270 mm for long and from 70 up to 90 for short. My wire is exactly 1,7 mm and spacing 6 (it is determined by contact place eballage and is not changeable.
So, after number of experiments, the first conclusion - resonaces are NOT depending on length of antenna. Hurrr! So, I put the two 100 Ohm resistors instead of antenna and got the same resonances. Cut off cable and got the same. Took off the adapter BNC to SMA-K and it normalized. It means the plug adapter is those damn problem-maker. However I made it very well soldering stick to stick the SMA-K with BNC, some 5 mm long zone where diameters may be not exactly passed, and rolled copper plate tubulus between both gnd. Thus it looked very well, and I was sure the much shorter than lambda/40 is acceptable. Seems I ought calc now a best diameter of this tubulus, let the soldering emballage have the same 50 Ohm by air volume.

P.S. Your recomended site is less flexible as those MOUKD site, as only there one may adjust the actual Vf. For thin wire one, for thick wire another and for wire in pvc pipe more another. Its not a secret, the lambda/1000 is thin with 0,98; more than lambda*0,008 is thick with 0,958 and pvc is 0,93. However if the vector analyser shows well, the adjustment-making may be provided with a smile - not my case.

P.P.S. Today I visited all dozen of component chain-shops here and no any result for SMA-BNC adapter. To wait from China - its month or more. Thought I`ll solder cable straight on the SMA, but there was no even those, only pressable kind. Means the story will have a longer agony in the future.

PPPS - the slim Jim takes actly twice as much copper, but gives 0,2 dB better result. Its nonsense.

 

Hello,

There do exist solder type SMA connectors.
https://www.amazon.com/Eightwood-10...er+connector&qid=1553090430&s=gateway&sr=8-15

Bertus

 

Well, yes, you can get some serious currents in the matching section that need the isolation.

At least after applying a computer network cable ferrite core over the coaxial as near a possible to antenna changed the picture cardinally, thus I was able to see that something is wrong with vector analyser.... leading to idea the troublemaker part was cable adapter from BNC to SMA-K. When I shall find it new, the game will go ahead again.

 

Hello,

There do exist solder type SMA connectors.
https://www.amazon.com/Eightwood-10...er+connector&qid=1553090430&s=gateway&sr=8-15

Bertus

Yepsssk, but 11 USD for tiny mess and 14 for postals its little bit excessive, isnt? The best would be to get the BNC (female electric male by emballage) to SMA-K adapter (male electric, female by emballage), but even Octopart and Farnell shows none.

 

Hello,

The shown add has a pack of 10 pieces of the connector.

Bertus

 

For bit better explanation - the sliding contact-block over J lover setion is looking like this https://by.all.biz/en/terminal-blocks-of-12-couples-2-5-4-0-6-0-10-0-16-g476523 (this is not a advert, only explanation how it looks)- I used to grind into those double-block the 4 mm deep gap with a thin disc, where happens place for coaxial "meat", which is soldered into drillings in inner metallic brass, later cleaned by corresponding drill from inside of main-wire channel. Thus factually in 15 mm long part the J-Pole wire is severe thicker than overall, but I dont think much the crucial reflections may be caused by such short fat-section. Coaxial is then going 90 deg away to aside, thus nothing metallic happens on the axis of both J-wires - ought not to create any interferences.

 

Hello,

The shown add has a pack of 10 pieces of the connector.

Bertus

Ahoi, didn cached. Yes, then its about those 2,68 Eur per piece I pay here in local shop, but wrong type.

 

Wow! Victory!! I worked bit on the turner bench, thus the pressable SMA turned to be solderable type, put the cable straight to vector analyser, checked cable (far from ideal but roughly OK) and adjusted the antenna. So, if any wants to repeat I declare here the exact sizes for free air installment. U-turn made on 6mm drill, from inner surface of U turn the cable soldering drilling heigth 7.5 mm, shortest dia 1,7 mm bar from the same place 86mm, longest 259mm. It means that velocity factor was 1,000 (?! wtf?) Measure data: rezon freq 868; BW correspnding to SWR 1,2 are 871-861; BW by 1,5 SWR is 855-877; SWR@868=1,07 (51 Ohm). Im happy
P.S. Ferrite core is obligatory
P.P.S. Still is fundamental contradiction to theory. As LOWER slide the contact place, as BIGGER is the R, and vice versa. Midway to 86mm is so small as 10 Ohm, instead of theoretical 1000-2500 Ohm. So, my mind still is bit tremolous.

 

Hour later: No, I am not 100% happy. I pulled an insulation over against the atmospheric corrosion, 4,7 mm out.dia or 2mm thick soft pvc, and now I have two fundamental resonances 841 MHz with R=426 Ohm and 909 MHz with 4,5 Ohm. Both are catastrophic, and may be corrected only by sliding contacts faaaaar away from calculated placement. WHY this is happening?
So, I got the insulation off and put a wires naked into 16/13 mm el-tech plastic paipe. Result: 912 MHz with 7 Ohm and 848 MHz with 158 Ohm. So... antenna is f** sensitive about plastic emballage.
Next experiment, I substituted the plastic pipe with the glass one, 13/11 mm dia, and got two resonances at 841MHz R=376 Ohm and 909 MHz with 4,6 Ohm. So, at least I may make a sure that the plastic material is innocent and not produce the huge loss of energy.

 

Just I found a few articles with surname of Rojas between others (as this - "Discussion of series and parallel resonance phenomena in the input impedance of antennas" by Khaled A. Obeidat,Bryan D. Raines, Roberto G. Rojas) explaining which one of those resonances is the true one.
Thanks to them from full hearth for opening my eyes.

Just any X axis crossing, having pF at right and nH at left is the sure wrong alias parallel resonance i.e. transmitting current minimum; and from rest (having pF at left and uH at right) those lowest is those resonance for actual using, namely serial, or that highest one is it`s third harmonic.

I wonder why only at 2010 such instruction was firstly spoken out - and how it can be the radio-ham society had lived 100 years wrongly singing a song about that resonance is one alone when elementary measurement shows a bunch. Even me a week ago was singing the same song as I was teached.

 

To illustrate the problem: I calculated and made very accurate +/-0,5 mm dipole antenna with idea, its most well explored and repeatable. Each of 1,7 mm thick branch H=84 mm, no metallic objects in nearest 5 lambda, expectations to have a resonance at 868 MHz and expected to have a 73 Ohm.

So now here is the resonace frequencies list, obtaind with vector analyser by criterion that at resonance any reactance is changing a sign, from inductive stays capacitative and vice versa.

137 MHz - lowest possible by analyser, inductive, non-resonant, active R=15 Ohm
230 MHz - sign changes from C to L, R=2,6 Ohm (=serial resonance indication)
310 MHz - sign changes from L to C, R=870 Ohm (=parallel resonance)
384, C to L 2,8 Ohm
462, L to C 509 Ohm
537, C to L 4,9 Ohm
613, L to C 390 Ohm
685, C to L 11 Ohm
756, L to C 102 Ohm
880, C->L 23 Ohm
951, L->C 151 Ohm
1022, C->L 14 Ohm
Note none of frequencies havent anything near the R=73 Ohm, and all are going far beyound and far above the calculated resonance.
WTF??? Chinese instrument??? Marsians??? ill fate??? Curvy fingers? Solar paches?
Ach Ya, near antenna I applied a ferrite tubulus from computer mains wire, to stop any currents going at 50 Ohm cable braiding. Probably it isnt designed for so high frequencies, but... havent in shelve a better choice for a while.

The cabling is going perpedicularly to resonators, thus braiding currents ought not to change any resonant frequencies. Fixture for resonators - the same electrotech contacture with white polyethylene emballage 12x12x6 mm.

 

Wait wait wait, may be those resonances are NOT an antenna produced but as the cable is crucially non-SWR~ed, those may be the resonances in the cable itself??? May be I ought to pass the cable length beforehand somehow???

P.S. After two days:
Hey, people, do You everyone are died? Or I am telling something too much new?

 

So, today I visited one local firm what every day makes a adjustments and repair works on hundreds of civil, military, seaman`s and airflight radios, so the job with vector analyzer is their ultimate bread. And they brainwashed me well.

The main mistake was that I took an "any length" of cable. But must take a lambda*V(f) or it`s whole multiplicates, what all has eternal R at resonance. Because all those bunch of resonances I seen was coming from (group A - highest) from antenna resonators, (group B, the middle freq) inside the cable between plug and place where 50 Ohms jump to antenna current resistance, and (group C) - what resonates from plug up to antenna most far end - the most longest wavelength.

So, my only claim this moment is that R of dipole is 13 Ohm instead of 68 expected at resonance, however good news is that thick pvc insulation over resonator slows down a frequency so small as 932 to 925 MHz (then R diminishes to 10 Ohms).

And one more claim, that 175 mm length of dipole from side to side gives so high frequency, what means that V(f) is negative.... Cannot be true. The next frequency is 692 with 260 Ohms, but that is antiresonace; and next resonance is 520 MHz what is serial again but the R=3 Ohms - so its wrong. More over, as those resonances stays far away from frequency determined by cable length, they are not to be trusted much. It means the 932/925 was the real and true.

P.S. Experimented with 227 mm what gave a 919 MHz @ 7,7 Ohm and antireson 763@397 Ohm and last one which I believe is the true one - 583 MHz @30 Ohm. Thus the V(f)=227 mm / 257 mm= 0,883 what incoorporates the insulation effect. I was expecting rather similar. At least now I know my insulator velocity factor.
And R is now exactly right, 73 Ohms, when insulated wite length is 142 mm from end to end, or iniside of one resonator 67 mm with 8mm gap, frequency 866 MHz.Take off the 1mm and voila. Ultracheap LORA(TM) antennas are ready. Final appreciation of V(f)=0,816.

Beware one of my mistakes - I soldered cable inward into resonator keeper, thus the gnd was just 1 mm off from central copper at solder point. When I resoldered it to outside, thus the distance was about 10mm, then 25 Ohm changed to 75 Ohms. Seems too close distance interferes too much. Next mail I shall told about success on the J-Pole.

 

So, now I shall describe a sizing procedure for J-Pole. Moukd gave a 212mm and 71 mm and 7 mm for V(f)=0,816 (=with insulation), I got a frequency 912 MHz with 50 Ohm yet slight reactance thus the SWR=1,26. At resonances R is clearly wrong. So, I recalced antenna to 912/868 and got 223mm, 75mm and 7 mm (all measured by inside of bottom part), means V(f)=0,86. Now it have 868 surrounding what have a demanded ohms, but those are ultra-sensitive about distance between both horns, short one and long. Plus minus 0,2 mm and R swings fro 200 Ohm to 50 and 10 Ohm. So, I need to rethink a better fixture to keep this distance constant. Monday shall work again. Buye buye.