profile | register | search | main site
|Forums | Fireside Crystal Radio Clubhouse | Dipole antenna||Post Reply||Send Topic To a Friend|
|gzimmer||Posted - 12 October 2012 1:28 |
I always wanted to try a dipole on a crystal set.
I had taken down my long wire before I left for holidays, so now that I'm back I thought it might be a good time to try a dipole.
Found a couple of likely trees and found my antenna launcher (catapult).
I couldn't find any TV ribbon, so hooked up some old twin-lead as a feed line. The dipole itself is about 50M each side.
At first was very disappointed as it was a deaf as a post. After thinking a bit, I decided the twin-lead might be soaking up the signal, so split the twin-lead down the middle. That made a huge difference, suddenly signals were strong.
The signals are stronger with the dipole connected as a single wire fed against earth, but connected as a dipole a whole different set of stations seem to be audible.
I have a double change-over switch arranged to switch between the two configurations and the difference is amazing.
Definitely worth trying
|homebrew||Posted - 12 October 2012 7:1 |
Looking into different antennas and their patterns is always interesting. You usually lose a couple and gain a couple stations.
I've mapped most of the normal antenna design patterns but have never had a good dipole up. I'd be interested in a map of the stations on both antenna set ups for comparison.
On my counterpoise antenna with the counterpoise 8 ft in the air directly under the longwire I noticed that the higher frequency stations were louder. Do you get the same effect on the dipole?
|gzimmer||Posted - 12 October 2012 7:13 |
Most local stations are quite weak (as you would expect with the cross polarisation) and the strongest signals seem to be more distant DX (eg sky waves).
And yes, it does seem to be much hotter at the top end. I was hoping for this as my previous long wire was a 1/2 wave at the top-end and hard to match. Each arm of the dipole is a 1/4 wave at the top end, so it is low impedance and much easier to match.
|Garry Nichols||Posted - 12 October 2012 7:22 |
I'm not picturing what the antenna/lead-in looks like after you split the feeder down the middle.
Are the two feeder wires still parallel?
What is their spacing?
John has commented a number of times that skywaves close to earth (such as at the height of a dipole) have their horizontal polarization component either attenuated or changed to vertical (can't remember exactly what he said).
So his feeling is that an antenna that does not have a part of its run vertically polarized will not be very sensitive. Apparently vertical polarization in the skywave is not altered.
Comments on this?
|gzimmer||Posted - 12 October 2012 7:53 |
The two feed wires just run parallel a few inches apart (coming down from the center insulator). If I get keen I might install some plastic spreaders.
According to the references, very low angle DX signals are vertical because the horizontal component has been absorbed as the signals skim the earth.
For crystal sets however, DX is merely "interstate" so these signals would be much higher angle (skywave).
And of course, even a perfect horizontal dipole has vertical radiation off the ends....
And yes, signals are mostly weaker on the dipole. With some surprising exceptions.
Edited by - gzimmer on 10/12/2012 7:55:00 AM
|homebrew||Posted - 12 October 2012 9:49 |
Running parallel wires connected at the lead in on my attic antenna, a 18 inch spacing seemed optimal.
Doing the same with my 150 ft longwire I found that 50 ft additional wires worked as well as another 150 ft parallel wire.
With your wires close together I would expect a strong reception lobe off the ends.
A couple short wires run off the far ends angled away and down towards ground should broaden the pattern and give better local reception.
But first I'd try a 500 ohm composition resistor across the far end. That should broaden the strong far end lobe to cover a 180 degree arc.
|gzimmer||Posted - 12 October 2012 9:56 |
To put it another way, if BC signals are vertically polarised it is because they are either local ground-wave (from up to 150 miles away) or very distant DX coming in at a low angle. For a crystal set you can ignore the distant low angle signals as they will be too weak to hear.
The ionosphere acts as a horizontal mirror, so sky-wave signals tend to be horizontally polarised.
Height plays a big part in this as well. A relatively low horizontal dipole (less than one wavelength high), responds only to high angle signals.
So the bottom line is that a horizontal dipole tends to reject local stations, but favours mid-distance (eg interstate) DX.
|gzimmer||Posted - 12 October 2012 9:57 |
> With your wires close together I would expect a strong reception lobe off the ends.
No, the feed-in wires are close together, but the individual dipole legs are 180 deg apart.
The spacing of the feed-lines really doesn't matter. These are known as "tuned feeders". Because they are not terminated in any particular impedance, they will have standing waves. But being open wires they still have low loss.
Edited by - gzimmer on 10/12/2012 10:06:29 AM
|_J_||Posted - 12 October 2012 11:16 |
Well, so here's what I think...
Boundary conditions will modify polarization. There is a space – ground-plane boundary at the surface of the earth. As the wave approaches a wavelength (~hundreds of feet for BCB) from the dirt, this polarization becomes increasingly vertical. That’s why they can’t use dipoles to transmit MW.
There is a minor horizontal component in the direction of travel of the wave though, because the earth is not a pure conductor. This can be sensed with a dipole near the ground (by near, I mean in terms of a wavelength, not a few inches, or feet) .
Also, a very short dipole (dimensions need to be viewed on a wavelength scale) will not have a feed impedance typical of twinlead. Normal twinlead will be far too capacitive, as Zim found out. A short feedline also does an impedance transform which is what Smith charts show. Sometimes I look at this from an amateur radio point of view by transforming the dimensions in terms of wavelengths to a 10 meter dipole a quarter wave above ground and see what the equivalent at 10 m situation would look like. Surprising.
Edited by - _J_ on 10/12/2012 11:22:48 AM
|gzimmer||Posted - 12 October 2012 12:11 |
> That’s why they can’t use dipoles to transmit MW.
Actually they can. It's called NVIS, but as you imply, most of the signal goes straight up.
I understand that the reason MW transmitters are vertical is to give them good ground-wave (in order to cover their licensed area), and to prevent them having any sky-wave and thus interfering with distant stations on the same frequency. But perhaps that is what you are saying.
Actually my dipole is around a 1/2 wave at the top of the band, where it should be around 75 ohms (or lower, depending on height).
I don't think that capacitance is the issue. Any transmission line will have a nominal impedance decided by it's physical dimensions (and dielectric). If it's terminated in the correct impedance then it should work.
I think the problem was lousy plastic with high dielectric losses, especially when being used as a tuned line (eg high SWR).
The hams have successfully used twin-lead as a cheap substitute for coax for decades. Tradition says that the clear plastic is less lossy than the black. Mine was the black stuff.
|_J_||Posted - 12 October 2012 22:19 |
As far as I know, NVIS or any dipole is rarely used for MW because the drive impedance gets so unusable as a dipole gets a small fractional wavelength from ground plane. Getting them far enough from ground takes too many supports that are too tall and too hard to decouple. You are right it is possible, I should not say can't around here. (VLF goes through the earth no matter how you launch it.)
Another reason MW uses verticals is it is practical to drive them... Note that MW beams even use verticals.
There is nothing wrong with twinlead as a transmission line (T-line,) used right. Any T-line has a nominal Z but a short piece acts as a transformer, with lots of C. I think there is an aggravated mismatch between the low dipole and the short T-line. As you cut and spread the lines in the twinlead apart, the Z would be expected to go up to reduce the mismatch (a little). T-lines don't behave like small circuits, they really have to be viewed as T-lines. Coax will have the same problem + also be unbalanced. There are some cool Smith chart simulators that will show this so you can play with ideas (without the pain!)
Edited by - _J_ on 10/12/2012 10:30:51 PM
|gzimmer||Posted - 12 October 2012 23:16 |
> Any T-line has a nominal Z but a short piece acts as a transformer, with lots of C.
Yes, if it's not terminated properly. But properly terminated it acts as a transmission line, no matter what it's length.
I think you are talking about lines with high SWR, eg "tuned lines". The resultant reactance is not necessarily a problem, it can be tuned out (eg the C is not inherently lossy).
Hams have used Tuned Lines to feed multiband antennas since day one. The classic is the Zepp antenna, its Tuned Line is terminated in an open circuit (the end of a half-wave dipole). Usually the feed line is a 1/4 wave, but it doesn't have to be, if the ATU can handle the transformed impedance.
So what I'm saying is that the capacitance of the twin-line is not the problem. Any capacitance can be tuned out in the ATU.
Edited by - gzimmer on 10/12/2012 11:33:24 PM
|_J_||Posted - 13 October 2012 12:32 |
Agree, note all T-lines and antennas are tuned --to something. Tuning and matching amounts to the same thing.
|golfguru||Posted - 13 October 2012 15:42 |
>>>> Tuning and matching amounts to the same thing. >>>>
Zim says that "just because you have *tuned* (resonated) an A/G circuit, does not mean that you have *matched* to it" (other threads).
The way I see it ( no doubt wrongly) is that when tuned, the only (major) thing left is the ground resistance. ie. at resonance, the impedance would always (approx.) equal R (Zr=~R).
One would think that a "once only" match to the R, would be all that is reqd. for max power transfer at all freqs. in the band of interest.
Will I ever comprehend?
Edited by - golfguru on 10/13/2012 4:14:03 PM
|_J_||Posted - 13 October 2012 16:14 |
You are right GG!!
I meant tuning and "terminating" are the same thing. But that is because tuning connotes matching both real and imag. parts for T-lines. I assumed Zim probably knows all this but your point is correct, I should have used more words.
Once the antenna is tuned where your signal is, X is gone, the R is what appears at the 'terminals' where X is tuned. That R is the ground resistance combined with the antenna's radiation resistance combined with the losses in the leaves on the oak tree. Ideally, ground & loss resistance would be 0 and you see only radiation resistance, but thats a dream.
By the way, the circuit tuning out X can sometimes transform those resistances to something else (amounts to R matching too.)
Edited by - _J_ on 10/13/2012 4:52:07 PM
|golfguru||Posted - 13 October 2012 16:56 |
If I use a fully resonant, MW, A/G circuit eg. Variable ferrite coil resonated against <=100ft antenna capacitance, the impedance across the band will be ~Rg?
Click Here To Close Thread, Administrators & Moderators Only.
Show All Forums | Post Reply