What is the Impedance of a Cross Yagi Designed for RHCP?

[userdnl]

I'm designing a cross yagi as one component of a capstone project. The design I have right now is 3 elements per polarization, counting driven elements (total of 6 elements). I ran a quick 4nec2 simulation of a single polarization (3 elements) using center excitation on the driven element and I'm seeing an impedance between 11 and 12 ohms at the feedpoint. I would like to configure the antenna to be fed for RHCP as described here. I'm trying to figure out what impedance I will observe so that I can design the matching circuit. I'm not sure how to change to a balanced excitation in my nec file or how to attach two 1/4 wavelength long conductors at 90 degrees with each other. Actually, I'm not even sure if it's possible since you usually give a single section of an element to apply excitation to in a nec file. If anyone can comment on how to configure my nec file so I can simulate the impedance or how to find the impedance theoretically I would appreciate it. Thank you

 

[Baluncore]

I'm not sure how you inject multiple different phase currents in nec, but you should be able to model orthogonal yagis independently. The impedance is therefore not significantly influenced by the opposite polarisation yagi.

CP is generated by a 90° phase shift network or transmission line difference between the two feed points. You might model that in nec.

There will be a very slight reduction in optimum parasitic element length in the presence of the second yagi, but it will not be significant.

 

[userdnl]

Thanks Baluncore. I assumed that independent orthogonal Yagis would not effect each other's impedance much. However, that article I linked describes a non-traditional way of generating RHCP where you feed perpendicular components of the driven elements in parallel. I'm not sure how this will affect impedance. If I can't figure it out I'll just do it the usual way with a 90 degree phase shift but this is supposed to be more flexible (less matching).

 

[Baluncore]

Yes, the article is correct, go ahead.
If you use the 75 ohm L/4 lines and move one of the yagis along the boom by L/4, you will generate CP.
Reversing the driven element connection on one of the arrays will reverse the CP direction.
The boom length will be slightly longer to handle the staggered arrays.

 

[userdnl]

I'm still a bit confused about what the antenna's impedance will be though. The article assumes a 50 ohm impedance where I simulated my Yagi at more like 12 ohms. I would also think that putting my antennas in the configuration described in the article is going to affect their impedances.

 

[davenn]

are the driven elements folded or straight dipoles

if folded dipoles the feedpoint will be ~ 200 Ohms when in a Yagi ( 300 Ohms if stand alone)
I have ALWAYS fed folded dipole Yagis with 50 Ohm coax and a 4 to 1 BALUN its the normal practice

A straight dipole will be ~ 75 Ohms stand alone and can be as low as your 12 Ohms when in a Yagi

Straight dipoles are often (very commonly) fed with a Gamma Match arrangement for impedance matching
when they are incorporated into a Yagi

cheers
Dave

 

[Baluncore]

The article assumes a 50 ohm impedance where I simulated my Yagi at more like 12 ohms.

Juggling the Gain, FB ratio, side-lobe amplitude, resistance and reactance can get quite involved. Unless you are watching all the important parameters you can end up with a problem in a neglected dimension. Sometimes it all looks good until you check the band edges and realize it is incredibly sensitive to wavelength. It is 20 years since I designed my last yagi with nec, I am getting a bit rusty.

I would not be happy with a 12 ohm feedpoint. You might try to increase your yagi impedance by using a non-conductive boom or by changing the diameter or position on the boom of the driven element. Changing the DE length will tend to neutralise the reactive component.

What is your centre frequency and bandwidth requirement?

I would also think that putting my antennas in the configuration described in the article is going to affect their impedances.

I agree that at first glance that may seem to be the case. But the array elements are all perpendicular so there should be very little electrical or magnetic coupling between the two arrays, even though they share the same 3D space. The biggest array interaction will be due to proximity of the support structures and the physical arrangement of the feed cables.

 

[userdnl]

Dave, these would be straight dipoles, but the article I linked in my original post describes an unorthodox method of generating circular polarization where you use perpendicular elements of the cross polarized Yagis.

Baluncore, I wouldn't say that I'm happy with 12 ohm impedance, but I assume it could be matched if done carefully. I'm only simulating the elements. I plan to make the boom out of wood or fiberglass and I'm hoping it's affect will be negligible. This is only a 3 element Yagi (size and weight constraints). Is that likely the reason for the low impedance?

The center frequency is 145.825 MHz and it is intended to receive packet radio from ARISS. FM voice would be a nice touch too.

12ohms is the real impedance. I'm also seeing a 58 ohm reactive component. I believe that the reactive component should be 0 at resonance. I ran the optimizer on the driven element length of a single Yagi to reduce the reactive component and ended with a 9 ohm real component. Very low.

I found the network feature of 4nec which will allow you to attach two segments. I did this with the center segment of each driven element of the cross Yagi and, using the optimized parameters from above, I got 7ohms, which makes sense since the two are now in parallel. I'm still not sure that this is an accurate picture of the impedance I will see in practice though because both elements are still being center fed.

 

[davenn]

Dave, these would be straight dipoles, but the article I linked in my original post describes an unorthodox method of generating circular polarization where you use perpendicular elements of the cross polarized Yagis.....

.

not really unorthodox, in fact its the common standard particularly for amateur radio installations as your one appears to be on 145 MHz

both Yagis on same boom and perpendicular, they don't need to be separated along the boom to any extent as Baluncore suggested earlier. Neither of my 2 satellite Yagis (2m and 70cm) have the 2 Yagis widely separated along the boom, The elements are mounted side by side but at 90 degrees. And the usual practice is to use a phasing harness to produce the LHCP or RHCP polarisation and relay switching to switch between left and right.

From a purely practical and experience point of view, I have never seen any adverse interaction between the 2 Yagis when they are perpendicular to each other ... ie ... no great changes in feedpoint impedance when measured with an antenna analyser.

OK on the straight dipoles ... gamma match will be your best choice, plenty of info in the ARRL Antenna Handbook for that. Folded dipoles are easier as you don't have to fiddle around with gamma match tuning

Cheers
Dave
VK2TDN

 

[the_emi_guy]

I'm not sure how to change to a balanced excitation in my nec file or how to attach two 1/4 wavelength long conductors at 90 degrees with each other. Actually, I'm not even sure if it's possible since you usually give a single section of an element to apply excitation to in a nec file. If anyone can comment on how to configure my nec file so I can simulate the impedance or how to find the impedance theoretically I would appreciate it. Thank you

Maybe you should post your NEC file.

I'm not sure I understand your NEC issues.

There is no such thing as "balanced excitation" in NEC. If your antenna is symmetrical with respect to the feed (which I assume it is), it will be balanced. Hooking it up to an unbalanced cable via. matching network/balun should be outside of the domain of NEC.

You attach the two 1/4 wavelength conductors at 90 degrees by just using the appropriate geometry cards. You can pretty much create any arbitrary arrangement of wires in 3 dimensions with NEC.

Make sure you sweep across your bandwidth rather than just modeling at your center frequency.

You should go through a good NEC tutorial before starting out. There a number of "best practices" when developing the model that, if not followed, will cause bogus results.

 

[userdnl]

Some NEC code

Single Yagi Simulation

SY DIAM=.0174845							'Wire Diameter
SY R=.991474								'Reflector Length
SY RO=.4114									'Reflector Offset
SY DR=.956505								'Driven Element Length
SY D=.909194								'Director Length
SY DO=.4114									'Director Offset

GW 2 9 -RO -R/2 0 -RO R/2 0 DIAM			'Reflector
GW 1 9 0 -DR/2 0 0 DR/2 0 DIAM				'Driven Element
GW 3 9 DO -D/2 0 DO D/2 0 DIAM				'Director
GE 0

LD 5 1 0 0 3.5E7							'Copper conductivity

EX 0 1 5 0 1 0								'Excitation at center of Driven Element
FR 0 1 0 0 145.825 0						'Design Frequency of 145.825MHz
EN

Cross Yagi Simulation

SY DIAM=.0174845							'Wire Diameter
SY R=.991474								'Reflector Length
SY RO=.4114									'Reflector Offset
SY DR=.956505								'Driven Element Length
SY D=.909194								'Director Length
SY DO=.4114									'Director Offset
SY OF=.03									'Offset between cross-polarized pairs

GW 2 9 -RO -R/2 0 -RO R/2 0 DIAM			'Reflector
GW 1 9 0 -DR/2 0 0 DR/2 0 DIAM				'Driven Element
GW 3 9 DO -D/2 0 DO D/2 0 DIAM				'Director
GW 4 9 -RO-OF 0 -R/2 -RO-OF 0 R/2 DIAM		'Reflector Cross
GW 5 9 -OF 0 -DR/2 -OF 0 DR/2 DIAM			'Driven Element Cross
GW 6 9 DO-OF 0 -D/2 DO-OF 0 D/2 DIAM		'Director Cross
GE 0

NT 1 9 5 9									'Connect the two Driven Elements

LD 5 1 0 0 3.5E7							'Copper conductivity

EX 0 1 5 0 1 0								'Excitation at center of Driven Element
FR 0 1 0 0 145.825 0						'Design Frequency of 145.825MHz
EN

I don't think that the way I am exciting the elements in this file is correct. I would either need to use a transmission line to generate a phase shift (like Dave was discussing) or find a way to excite the 1/4 wave elements individually and unbalanced to do things like the article I linked. At the moment, I don't see how I can find the impedance the way the antenna is configured in that article. If I can't figure it out I may just do the 90 degree phase shift.

I run optimization and found that DR=0.862199 will eliminate the reactive component at my center frequency. I haven't updated this yet in those files.

 

[userdnl]

I've been thinking about the matching network. I am now leaning against the method of generating CP in the article I linked as it is going to make it very difficult for me to match the impedances. I plan to go with the 90 degree phase shift method and feed them as two separate Yagis. To re-iterate, this antenna will be receive only (not licensed yet :( ). I will need to match the two Yagis to a 50 ohm coax cable which will interface with our test equipment and radio. My primary concern is minimizing loss.

I had originally planned on using a balun as I expected to use an unbalanced feed to generate CP as the article I linked showed. However, I'm now thinking maybe the gamma match would be my best choice. If one of the other matching method would generate less loss, I would reconsider. Is there an optimal point to connect the center conductor to? Or do I just choose a point and use the variable capacitor to match impedance? I'm thinking that the cap should only cancel the reactive component. I'm thinking a 1/4 wave transformer with 75ohm coax to match the two antennas (which I intend to tune to 50 ohms) and a 1/4 wave piece of 50 ohm coax to generate the phase shift. I have access to a network analyzer and definitely welcome any input.