Sleeve dipoles

I don’t think I follow your description, sounds like using the coax outer has the lower vertical element?

The kind I know was made from a bunch of baked bean tins as the sleeve/lower vertical element. Upper vertical radiator plugged into connector on topmost can lid. Coax feed hung down inside the sleeve. Support was insulating pole going up inside cans, so high voltage node at the rim of lowest can was in the clear.

It is possible to achieve the same thing by stripping off the outer cover of the co-ax and turning the braid back over the outside of the sheath for about a quarter of a wavelength - with the inner protruding on a further quarter wavelength from the turn-back point.

That's more like a conventional dipole with the centre feed point 'cold'. In the design I had in mind, the co-ax ran up inside the lower tube of the dipole, which was earthed at the bottom. The dipole got progressively 'hotter' as you went up. The feed point in the middle was 'half-hot', so the quarter wavelength of cable feeding it acted as an isolator and there was no difference in potential between the tube and the braid.

All the sleeve dipole designs I've personally seen, have been of the sort that piglet describes, where the downward-facing radiator of the dipole is a cylinder that's open at the bottom and is well away from ground. In these sorts of sleeve dipoles, the lower radiator is connected to the coax shield at the center feedpoint. I've read that these are easier to get working, if the diameter of the lower radiator is quite a bit larger than the diameter of the coax which runs up through it, so there's minimal capacitive coupling between the coax shield and the radiator around it, and the bottom of the radiator presents the high impedance that's expected at the ends of a dipole.

I haven't seen the sort of sleeve dipole you describe, where the very bottom is grounded. For there to be no potential anywhere between a point on the lower radiator tube, and the closest point on the coax braid, it seems to me that the tube and coax would have to be shorted together at both ends... and electrically this would behave a lot as if the tube wasn't there at all, and the braid was acting as the bottom half of the dipole (grounded at the bottom). You'd have the upper (tuned) half of a dipole sticking up into the air, and the lower half would be the (untuned) entire length of the feedline and/or mast. Not terribly predictable... and I think this is the focus of your concern?

Can you point me to a picture or article which shows this sort of construction?

For what I think of as a "standard" sleeve dipole, there's definitely the possibility of current flow on the outside of the feedline braid or shield, just as there is with any sort of coax-fed dipole which doesn't have a balun or isolator (J-pole antennas suffer from the same sort of thing). This can definitely disturb the pattern, often in ways which are difficult to predict due because every installation's feedline is going to be of a different length and its braid will present a different radiation resistance at the feedpoint.

The old "Isopole" sleeve-dipole design tries to work around this by having one or more decoupling skirts attached to the mast (or the outside of the feedline - same thing) below the bottom of the lower coaxial radiator. As I understand it (possibly wrongly) each skirt presents a low impedance to RF flowing down the outside of the feedline (so the current flows out along the skirt) and a high impedance to RF trying to "flow back up" the inside of the skirt. I don't know how well this works... the Isopole design seems to have fallen out of favor over the years.

My reasoning is as follows:

The end result could be regarded as a quarter-wave whip above a folded-back ground plane of indeterminate size.

For sufficiently different values of "about the same potential", I think.

There will be capacitive coupling between the coax sleeve, and the radiating element outside it. How much, depends on the distance between the two, and the dielectric constant of whatever's in between them (PVC and air, in this case).

I don't think it's justified to say that they will actually be at the same potential, since they aren't actually shorted together anywhere other than up at the central feedpoint.

Agreed, and it shouldn't be.

I'd say that the length of the coax is going to be relevant, but perhaps not critical.

That, I believe, is not a logical conclusion, based on what the author writes. Also, it doesn't jibe with what I've read elsewhere about RF-on-the-coax situations.

At DC, sure. At RF, I truly do not believe so.

I fear that you're chaining together a series of somewhat unsupported assumptions to come to this conclusion... too many "therefores" which aren't really justified.

I think you'd actually need to build one, and perform some tests and measurements, to see if those conclusions and arguments actually hold up.

The tests I'd do would be two-fold:

- Test to see if there's actually RF flowing on the coax when transmitting.

- Measure the SWR (ideally, the actual complex impedance) as seen at the beginning of the feed-line at the base of the mast. Then, try adding a quarter-wavelength of additional feedline, and see if that changes things. Try cross-connecting the bottom of the lower tube to the outside of the braid, and see if that changes things (it won't _if_ the two are actually at the same potential, but I believe that it probably will). Try connecting the bottom of the lower tube to a metallic mast which is grounded at the bottom.

In both of those cases, I think your initial "if" isn't valid. It seems to be based on the assumption that the entire outside of the coax shield is necessarily at earth potential. That's going to be true at DC, but I don't believe it's valid at RF.

Rather: consider the current flows. Inside the coax, you're going to have balanced current flow... up the center conductor, and back down the inside of the shield (or vice versa depending on which half of the RF wave cycle you're in).

When the coax hits the feedpoints, the current coming up the center conductor is going to go into the upper-half-of-the-antenna radiating element (the upper foil tape) since that's the only place it can go.

The current coming up the inside of the braid has two possible places to go - out and down the lower foil tube (balancing the current flow into the upper foil tube), or down the outside of the coax braid (the "RF on the coax" situation, leading to "RF in the shack"). How the current divides itself, is going to depend on the RF impedances the current "sees" presented by these two conductive paths.

The outer tube will present an impedance of about half of the antenna total... somewhere in the 25-35 ohm range, most likely, at resonance.

The outside of the coax will present an impedance of... well, it's going to depend on the installation. It will depend on the length of the coax (back to the transmitter or another ground point) as this will influence its radiation resistance. It'll depend on whether it's close to an odd or even number of quarter-wavelengths, or somewhere inbetween (and it's not likely to be purely resistive). It is _not_ going to be either zero (a "pure short to ground") or infinitely high (and thus unable to carry any current).

So, you're going to end up with some amount of RF flowing on the coax braid. How much, and how much it affects the pattern, is going to depend on the specifics of the installation. If you're lucky, the braid impedance is going to be high and/or reactive enough that it won't allow much real current to flow, the coax won't radiate much, and

Kinda agree, but I think there are actually three elements in play here:

- The quarter-wave upper whip.

- The slightly-less-than-quarter-wave lower element, which is behaving like a normal not-grounded-at-the-end dipole element.

- The feedline braid, and any metallic mast to which it's attached. This is connected to the "down" side of the dipole, at the feedpoint, but not elsewhere. It acts as a secondary pathway for currents on that side of the antenna, it _will_ radiate, and it will affect the SWR/impedance/pattern in somewhat- unpredictable ways.

My overall conclusion is that the feedline coming down the inside of the lower tube isn't really an "isolator". Its braid serves as an auxiliary (rather-unwanted) radiator, which one hopes doesn't radiate enough to mess up the antenna pattern too greatly.

Thanks for those thoughts and I see your points. My theory would work as long as the feed point correctly terminates the co-ax - which it may not. Also, in theory, the co-ax iside the lower element ought to behave like a quarter-wave 'metallic insulator', but it sounds as though you have your doubts about that too.

The propagation velocities of the co-ax and the lower element may differ. Perhaps this isthe real purpose of the offset feed point, it keeps the length of co-ax inside the lower element at a quarter-wavelength whilst maintaining the correct overall length of the dipole.

I can certainly see that the theory may not hold good over the entire band - but how much would that matter in practice?

snipped-for-privacy@poppyrecords.invalid.invalid (Liz Tuddenham) Wrote in message:r

The bottom dipole half is never grounded. Another name is coaxial antenna. Usually you find them in marine application because of the counter poise.

Cheers

Martin Rid <martin snipped-for-privacy@verison.net wrote:

sleevedipoles (in particular for 2-metres wavelength).I understand the principle that the feeder (assumed 75 or 50-ohm co-ax)is threaded up through the bottom quarter-wave element Thequarter-wave piece of feeder ascts as an isolattion stub so that thebottom of the element can be earthed and the feed point is 'half-hot',with the top of the upper element 'fully-hot'.A further refinement is to offset the feed point slightly lower than theexact physical centre of the dipole so as to allow for the differentpropagation velocity of the waves in the feeder from that in the dipoleelements, thus achieveing a better match.If the bottom of the sleeve dipole is standing on the ground or a groundplane, this makes sense - but what if it is mounted on top of aconductive metal pole of unspecified length? Won't the pole act as anumber of other dipoles which, depending on its length, can distort theradiation pattern in various ways?Worse still, what if the bottom element of the sleeve dipole is simply acontinuation of the pole (or eletrically connected to it) and the co-axis continued down inside the supporting pole to the bottom? Does theco-ax need to be bonded to the pole at the point where the bottomelement should end?Is the pole length irerelevant because a pole diectly below a verticaldipole is in the null zone, so anything below an earthed bonding pointwill not be energised?-- ~ Liz Tuddenham ~(Remove the ".invalid"s and add ".co.uk" to reply)

Suppose the feed point had a 1:1 isolation transformer, there would be no reason why the dipole couldn't be grounded at any point alog its length. The same isolatio could be achieved by a quarter-wave line.

The bottom half can not be grounded no matter what except at the feed point maybe. The sleve dipoe can be thought of as a ground plane with the elements just folded all the way down instead of down at a 45 deg angle that makes the impedance near 50 ohms. It is usually mounted with a mast that goes to the feed point and it can be either a conductor or nonconductor. This pipe can then be fastened to a conductor or nonconductor.

While a coax dipole can be made of a coax cable with the braid just folded back over the outer insulation it may not work very well due to the closness of the brade and the quality of the insulation.

I think you may have misunderstood the point I was trying to make. Suppose, for instance, you drove the dipole from a battery-powered transistor oscillator physically located at the centre point, there is no reason why the bottom of the dipole could not be grounded and the oscillator left to 'float' at about half potential.

In that case, driving the dipole from an isolation transformer would have the same effect (if we neglect the inter-winding capacitance of a practical transformer).

You could perhaps get that sort of effect by installing a few Type 43 (or similar) ferrite beads around the coax inside the lower tube, just below the feedpoint. These would act as a choke, preventing RF from flowing down the outside of the coax braid. Smaller and easier than using a transformer per se.

That would require that the quarter-wave line be fully RF-grounded at its bottom end. If the feedline is an unbroken coax, that wouldn't be the case... there's all that coax shield from the quarter- wave point below the antenna, and ground level to be considered.

Now, what one can probably do is create a "virtual RF ground", by installing something low-RF-impedance at this point 1/4 wave below the feedpoint (or, perhaps more successfully, 3/4 wave below, so it's not too close to the bottom tip of the lower radiator tube). I believe that this is what the Isopole antennas may be doing - they use a downwards-facing cone 1/4-wave long, mounted at this point. A set of quarter-wave radials would serve the same purpose. The low RF impedance at this point is transformed back to a high impedance at the feedpoint.

I'm pretty sure I've seen some coaxial dipoles which had a set of small radials mounted 3/4 wavelength down below the feedpoint, for this very reason... the radials "discourage" RF on the coax below that point.

Right. It'd act as a "metallic insulator" and present a high impedance at its top, only if its bottom was RF-grounded (and presenting a low impedance at that point), and I don't believe that's the case. You'd need some radials or other RF "virtual ground" mechanism to make this work.

Probably will.

I think it's more for impedance matching purposes.

Due to some amount of RF flowing on the coax, the impedance seen on the "ground side" of the dipole feedpoint is going to be less than the impedance on the "hot side". This will cause the whole antenna's impedance to be less than 50 ohms, giving a higher SWR. Shortening the lower tube will raise its impedance somewhat, in effect compensating for the impedance caused by RF on the coax shield, and bringing the impedance back closer to 50 ohms.