Question About Length of Radials

[Foofaraw]

I've read that when building a basic 1/4-wave antenna on VHF, the radials should be about 1/4-wavelength, or about 19.5", depending upon the exact frequency you're using. If we think of the radials as playing the role of a dipole's bottom-half, that makes sense to me.

But I've also read that if you're adding radials to a dual-band antenna (VHF and UHF), you can just use one set of radials, which are trimmed for the length of the larger/longer wavelength. So in this case the 19.5" radials would work fine for the UHF performance of the dual-band antenna.

Is this true? Does that mean that radials have a minimum length, but not maximum length? Or is it just a trade-off, and you get better performance when the radials are trimmed to match the length of whatever wavelength you're transmitting?

 

[BC Coyote]

Does that mean that radials have a minimum length, but not maximum length

Your close. The radials should be a quarter wavelength at the lowest frequency of operation. Making them any longer will not be an improvement, and can affect the angle of radiation of the antenna in an adverse manner.

 

[brandon7861]

Take four or more 1/4 wave radials, bend them down a little for a decent 50Ω impedance match, then imagine the radiating element was made the same. It would look like a biconical antenna, just made out of wire. I think the ground plane, even if paired with a single straight radiator, still behaves as a conical antenna (which are known for their bandwidth). The wikipedia page for biconical antenna talks a little about the BW. I think the reason a VHF ground plane works on UHF will be found somewhere in there.

 

[The DB]

Is it something like 2 meter and 70 cm ham bands where the frequencies in question are harmonically related (or at least close)?

Is there some form of matching on the antenna?

Is there some form of inductive/capacitive loading on the antenna?

How did they achieve said multi-banding?

Although there are specific times when it is necessary, don't assume that all radials have to be 1/4 wavelength in length, there are things that can affect this. Sometimes antennas are intended for use with either longer or shorter radials, and sometimes just having radials where they are not needed can cause problems.


The DB

 

[HarryUSA]

First of all, how are you getting a 1/4λ ground-plane antenna to give you the proper input impedance on both 2M & 70cm? Unless you are using a 70cm trap, this is not possible as your input impedance at 70cm, for a g-p antenna, designed for 2M, would probably be over 400 Ohms, not 50 Ohms.

Anyway, let's say you had a trap in the antenna and made a dual-band 1/4λ ground-plane antenna for 2M & 70cm, then the answer is yes, you should have a set of 1/4λ counterpoises, cut to their respective lengths for each band.
WHY?
Because a 1/4λ ground-plane antenna is still in-effect a 1/2λ dipole antenna, the difference is, the counterpoise is not being fed by an equal but opposite voltage like a true dipole would be. But that doesn't mean the counterpoise doesn't have current on it, even though it's connected to the coax shield, which most folks ground. It does have current flowing on it, through capacitive coupling with the "radiating" element ( I put "radiating" in quotes because your counterpoise actually radiates and reflects energy as well.)

The reason your counterpoise elements should be 1/4λ, is that, at 1/4λ, the voltage at the antenna feed-point is 0 volts (because of phasing of the voltage induced on the counterpoises, reflecting off the ends of the counterpoises and nulling at the feed-point end of the counterpoises), which is what you want if you don't want your coax to become another radiating element; and since you can only get resonance at exactly one frequency for a given antenna dimension, you therefore only have one frequency where the voltage at the feed-point actually "perfectly" nulls itself. This is why, if you use your antenna at various places in a band, you should always use a choke at the feed-point of a coax fed antenna, to choke off any current that wants to use your coax as a counterpoise, instead of your actual radials. (Unless you want to use your coax as part of your radiating system.)

 

[brandon7861]

First of all, how are you getting a 1/4λ ground-plane antenna to give you the proper input impedance on both 2M & 70cm? Unless you are using a 70cm trap, this is not possible as your input impedance at 70cm, for a g-p antenna, designed for 2M, would probably be over 400 Ohms, not 50 Ohms.

Anyway, let's say you had a trap in the antenna and made a dual-band 1/4λ ground-plane antenna for 2M & 70cm, then the answer is yes, you should have a set of 1/4λ counterpoises, cut to their respective lengths for each band.
WHY?
Because a 1/4λ ground-plane antenna is still in-effect a 1/2λ dipole antenna, the difference is, the counterpoise is not being fed by an equal but opposite voltage like a true dipole would be. But that doesn't mean the counterpoise doesn't have current on it, even though it's connected to the coax shield, which most folks ground. It does have current flowing on it, through capacitive coupling with the "radiating" element ( I put "radiating" in quotes because your counterpoise actually radiates and reflects energy as well.)

The reason your counterpoise elements should be 1/4λ, is that, at 1/4λ, the voltage at the antenna feed-point is 0 volts (because of phasing of the voltage induced on the counterpoises, reflecting off the ends of the counterpoises and nulling at the feed-point end of the counterpoises), which is what you want if you don't want your coax to become another radiating element; and since you can only get resonance at exactly one frequency for a given antenna dimension, you therefore only have one frequency where the voltage at the feed-point actually "perfectly" nulls itself. This is why, if you use your antenna at various places in a band, you should always use a choke at the feed-point of a coax fed antenna, to choke off any current that wants to use your coax as a counterpoise, instead of your actual radials. (Unless you want to use your coax as part of your radiating system.)

70cm is an odd harmonic of 2m, so for that reason alone, it should work. The ground plane is not presenting a high impedance at 70cm when it is 3/4 wavelength. Also, ground planes can act as a traveling wave antenna for situations outside of resonances (but nobody checked out that biconical antenna wiki page I mentioned a few posts up). Think of a long wire angled such that the incoming signal hitting it lines up with the induced wave traveling down it. The wave on the wire is a little bit slower due to the velocity factor, so angling it allows the incoming signal and the signal induced in and traveling down the wire to reinforce and increase in strength with increasing length irrespective of wavelength. The radiation angles may not be ideal with the GP angle set for an impedance match, but it's a contributing factor here. It's not as cut and dry as simply saying "its not a quarter wavelength".

 

[HarryUSA]

70cm is an odd harmonic of 2m, so for that reason alone, it should work. The ground plane is not presenting a high impedance at 70cm when it is 3/4 wavelength. Also, ground planes can act as a traveling wave antenna for situations outside of resonances (but nobody checked out that biconical antenna wiki page I mentioned a few posts up). Think of a long wire angled such that the incoming signal hitting it lines up with the induced wave traveling down it. The wave on the wire is a little bit slower due to the velocity factor, so angling it allows the incoming signal and the signal induced in and traveling down the wire to reinforce and increase in strength with increasing length irrespective of wavelength. The radiation angles may not be ideal with the GP angle set for an impedance match, but it's a contributing factor here. It's not as cut and dry as simply saying "its not a quarter wavelength".

Now that I did the arithmetic, I see you are correct, 70cm is very close to 3/4λ, so yes, a 1/4λ antenna for 2M antenna will have a 50 Ohm feed-point, and so will the 70cm band since 3/4λ is also fed at a current loop.