random wire length survey

[wenzeslaus]

I tested 4 lengths of random wire antenna in the house.

feet, length x4 (meters), MHz
12, 14.63, 20.5
15, 18.29, 16.40
25, 30.48, 9.84
35.5, 43.28, 6.93

spacing was 2-3" apart. time was 1/2 hour before to 1 hour after sunset, starting with the high frequencies first.

results:

20 MHz, no difference between all antennas.

15.30 MHz, no difference between antennas. you can test to see if there is any difference at all by touching the other disconnected antenna to the connected one. 12-15, no noise. 25-any, made noise. hard to tell any difference. 35-any, made a little noise. 12-25/35, made a little noise.

12.90 MHz, 35 noticeably faded compared to the others.

10.10 MHz, 35' was useless. the other 3 all seemed to get the same reception. touch test: 12-25, no noise. 25-any, no noise. any others made a noise.

7.80 MHz, all had the same reception. touch test: all made noise.

6.50 MHz, all had the same reception. touch test: 12-15, no noise. 25-35, little noise. 25-12/15, no noise. 15-35, loud noise. 15-25 made noise.

4.75 MHz, all had the same reception, except 35 was sort of faded and had higher noise floor.

4.00 MHz, 25 received signal, the others could not. background noise definitely changed volume between antennas. 12<15/25<35.

3.35 MHz, 25 received a strong signal, not so strong on 12/15, absent on 35. volume/noise definitely 12<15<25<35.

observations:

1. the 35' antenna should have been the strongest at the longer wavelengths, but instead it was the worst below 6 MHz for some odd reason, with a higher noise floor yet insensitive to signal.

2. touch testing showed there often were differences but could not correlate with observed performance. not much of a difference anyway.

3. the 25' length consistently pulled in signal at every frequency, the others were inconsistent (12', 15', 35') or dropped out entirely (35').

4. shorter than 35', the antennas had mostly the same performance, with 25' being the most consistent and never the weakest.

5. longer than 35' is not expected to improve reception, only raise the volume and noise floor.

conclusions:

1. 35' and longer has problems.

2. any wire from 12-25' will work.

3. 25 feet is the best length.

summary:

use a 25' wire.

 

[Dirk_SDR]

Your findings are compliant with other tests, e.g. here:

Random Wire Antennas - Best Lengths To Use For Random Wire

You can see a table with wire lenghts to avoid and another line in green beginning with "REVISED:" for good random wire lengths.
It is also known, that very long random wires do not always give better results.
As you can see, a 29' or 35.5' random wire is best.

 

[WA8ZTZ]

Sometimes longer isn't better.

 

[wenzeslaus]

it's a receive-only antenna so half wavelength doesn't matter.

but for the sake of our TX brethren, we can discuss.

the author used a calculator and C programming to do what anybody could do with a spreadsheet. he felt a need to calcualte to the 32nd multiple of a half wave on 10 meters. so with a spreadsheet, populate the first column (start at A2) with frequencies of interest, then make the row above that (start at B1) 1,2,3...32, then put =(491.78*B$1)/($A2) in B2, which takes into account half wavelengths, conversion to feet, and multiples thereof. format to 0 fixed places. copy that to each cell in the table.

you can go wading through the spreadsheet or you can copy all numbers to a txt and edit by sorting in order.

it's easier to just do the calculation given what you have.

L = (c*a*n)/(f*2)
where
L= length in feet
c= speed of light, 2.99792485 E8
a= 3.2808, convert meters to feet
f= frequency in 1/s
2= convert to half-wavelength
n= multiples of half-wavelength

the equation becomes

L= (491.78*n)/f
where
L= length in feet
f= frequency im MHz
n= multiples of half-wavelength

given the frequency you want to use, try a few multiples to see what length to avoid. calculate with n+0.5 to put the length between nodes. note that n=0.5 is 1/4 wavelength.

or you could use a spreadsheet with that equation to rapidly test n.

(A1) enter n
(A2) enter f in MHz
(A3) =491.78*A1/A2

easiest of all would be to calculate L at the highest frequency you plan to use and n=1, and don't use an antenna longer than that because any lower frequency contains a half-wavelength somewhere and any longer wire at this highest frequency also has half-wavelength multiples somewhere. cut it a little short to avoid the 1/2 wavelength exactly at the highest frequency.

for example, say the highest frequency you plan to use is the top of the 15 meter band. that's 19.10 MHz.

L= (491.78*1)/19.10= 25.7 feet

so use a 25 foot antenna.

speaking of, the values in the author's table are off, and the 35.5' random wire I tested was not optimal compared to 25' and the worst of all antennas under 6 MHz, which should have had no problems according to his calculations.

but he was addressing TX/RX antennas not RX only. while multiples of half-wavelengths are bad for TX, there must be other factors at play that make a good RX antenna or not.

 

[dlwtrunked]

Indoors, a longer antenna may be more likely to pass closer to indoor wiring and other noise source.
Such a test is very subject to local surroundings as well as the impedance of the receiver connection. As a result, there are no hard fast reliable rule.

 

[prcguy]

it's a receive-only antenna so half wavelength doesn't matter.

but for the sake of our TX brethren, we can discuss.

the author used a calculator and C programming to do what anybody could do with a spreadsheet. he felt a need to calcualte to the 32nd multiple of a half wave on 10 meters. so with a spreadsheet, populate the first column (start at A2) with frequencies of interest, then make the row above that (start at B1) 1,2,3...32, then put =(491.78*B$1)/($A2) in B2, which takes into account half wavelengths, conversion to feet, and multiples thereof. format to 0 fixed places. copy that to each cell in the table.

you can go wading through the spreadsheet or you can copy all numbers to a txt and edit by sorting in order.

it's easier to just do the calculation given what you have.

L = (c*a*n)/(f*2)
where
L= length in feet
c= speed of light, 2.99792485 E8
a= 3.2808, convert meters to feet
f= frequency in 1/s
2= convert to half-wavelength
n= multiples of half-wavelength

the equation becomes

L= (491.78*n)/f
where
L= length in feet
f= frequency im MHz
n= multiples of half-wavelength

given the frequency you want to use, try a few multiples to see what length to avoid. calculate with n+0.5 to put the length between nodes. note that n=0.5 is 1/4 wavelength.

or you could use a spreadsheet with that equation to rapidly test n.

(A1) enter n
(A2) enter f in MHz
(A3) =491.78*A1/A2

easiest of all would be to calculate L at the highest frequency you plan to use and n=1, and don't use an antenna longer than that because any lower frequency contains a half-wavelength somewhere and any longer wire at this highest frequency also has half-wavelength multiples somewhere. cut it a little short to avoid the 1/2 wavelength exactly at the highest frequency.

for example, say the highest frequency you plan to use is the top of the 15 meter band. that's 19.10 MHz.

L= (491.78*1)/19.10= 25.7 feet

so use a 25 foot antenna.

speaking of, the values in the author's table are off, and the 35.5' random wire I tested was not optimal compared to 25' and the worst of all antennas under 6 MHz, which should have had no problems according to his calculations.

but he was addressing TX/RX antennas not RX only. while multiples of half-wavelengths are bad for TX, there must be other factors at play that make a good RX antenna or not.

Multiples of a half wave are actually very good for transmit and receive. The only problem is they are very high impedance and require appropriate matching. And when you match those exact 1/2 waves or multiples using something like a 49:1 or 64:1 transformer you will get a better performing antenna for size compared to the random non resonant lengths connected direct to coax or even with a 9:1 transformer.

A general statement that longer is not better is a completely wrong way of looking at this. A longer wire with respect to wavelength will have more gain lobes and nulls. If you have a random length of wire that is a fraction of a wavelength or no more than 1/4 wavelength long it will be mostly directional off the sides. Lengthen that to a full wavelength and it will have nulls off to the sides so you may have been originally listening to a station broadside on the shorter wire at or less than 1/4 wavelength long and now your full wavelength wire put a lossy null towards the station and you now think a longer wire is not so good. But if you were to rotate that longer wire a little and land on a gain lobe you would actually have a stronger signal.

Antenna testing of any kind is very complex and you need to run enough variations to answer for directionality of your various samples not to mention what any length of wire offers in impedance to your receiver and how much that will degrade reception. A person on your same block testing the exact same length of wire but at a right angle to your testing may come up with completely different results.

Its been a very long time since I experimented with a wire antenna for SW extending it from maybe 25ft out to 200ft in increments and my recollection was every increase gave me more signals on all bands. The 200ft of wire across our 1acre plot was amazing and pulled it super strong signals everywhere. I believe I was using a Hammarlund SP600 JX17 during that test and it had a somewhat high impedance antenna input and I think it also had a trimmer for peaking the antenna. That can make up for problems you may have using 50 ohm input.

 

[krokus]

Its been a very long time since I experimented with a wire antenna for SW extending it from maybe 25ft out to 200ft in increments and my recollection was every increase gave me more signals on all bands. The 200ft of wire across our 1acre plot was amazing and pulled it super strong signals everywhere. I believe I was using a Hammarlund SP600 JX17 during that test and it had a somewhat high impedance antenna input and I think it also had a trimmer for peaking the antenna. That can make up for problems you may have using 50 ohm input.

What was the configuration of that 200 ft wire?

 

[prcguy]

What was the configuration of that 200 ft wire?

Wire strung horizontal about 20ft off the ground run through a window right to the receiver, no coax or transformer.

 

[wenzeslaus]

guys... guys... here is my antenna input

 

[wenzeslaus]

Sometimes longer isn't better.

it ain't the size of the antenna, it's how you wiggle it

 

[wenzeslaus]

what can be done to improve a long wire connected to a telescopic antenna? is there something you can do with a balun, variable capacitor, grounding? V or L shape? or what

 

[krokus]

what can be done to improve a long wire connected to a telescopic antenna? is there something you can do with a balun, variable capacitor, grounding? V or L shape? or what

Get it away from the ground, and away from occupied buildings. Be sure to provide a static bleed route, if more than a few feet long.

 

[merlin]

Hmm, Whats wrong with that ?

 

[commscanaus]

guys... guys... here is my antenna input

That setup is hard core!
Saved a few $$$ on connectors and adapters there too!

Commscanaus

 

[CanesFan95]

So do you just drape a long piece of wire out the window and clip it to the receiver? The current antenna setup is grounded with lightning arrestors in a box and so do I just run a piece of wire into the window for SWLing?

 

[wenzeslaus]

that's what a lot of people do, more or less

I don't put it outside. I can fit 50' inside the house and find that 25' is a lot more comfortable.