HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Oct 10, 2014, 03:43 AM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
Discussion
Influence of cable on antenna performance

Since I've received my new gadget the miniVNA tiny, I have played around measuring my antennas with all sorts of cables I had lying around.

I am really amazed what effect the cable has on antenna performance.

Have a look at the results of a 3turn helical connected with equal length rg316 and rg174.

I didn't expect such a difference. Could the experts maybe elaborate a little bit about this. That would be so cool :-)

Thanks a lot

suxi

PS: The bandwidth of a 3turn helical is really amazing.
suxi is offline Find More Posts by suxi
Reply With Quote
Sign up now
to remove ads between posts
Old Oct 10, 2014, 03:46 AM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
That's the antenna:
suxi is offline Find More Posts by suxi
Reply With Quote
Old Oct 10, 2014, 04:06 AM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
Spooky, without a cable it performs much worse. Any ideas?
suxi is offline Find More Posts by suxi
Reply With Quote
Old Oct 11, 2014, 04:06 AM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
OK, those are my noob thoughts:

In cables with length unequal n*lambda/2 the reactance (imaginary part of the impedance) will not be zero, thus displaying maybe better measurement results then actually present.

So it should be best to measure antennas without a cable, by for example calibrating the VNA at the end of the cable. If the antenna shows an Z of 50 Ohm with no imaginary part it should also tolerate different cables better.

Any thoughts on this would be highly appreciated.

Thanks a lot

suxi
suxi is offline Find More Posts by suxi
Reply With Quote
Old Oct 11, 2014, 04:12 AM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
Concerning cables: RG174 has a higher damping effect then RG316, which means rg174 cables will show a better SWR, because some of the reflected power is absorbed by the cable. Is this already noticeable at 30cm cable length?
suxi is offline Find More Posts by suxi
Reply With Quote
Old Oct 11, 2014, 12:24 PM
Registered User
Mr.RC-CAM's Avatar
USA
Joined Oct 2001
5,213 Posts
Your observations are discussed here:
http://jeff.appliedsyst.com/9V1AS/TL...eder_tunes.pdf

- Thomas
Mr.RC-CAM is offline Find More Posts by Mr.RC-CAM
Reply With Quote
Old Oct 13, 2014, 05:10 AM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
Thank you Thomas, this is a great article. With my current knowledge I don't comprehend all of it but the winter hasn't begun yet ;-).

So in terms of antenna tuning, one should measure and tune the antenna with no cable at all, if possible. If the antenna shows an impedance of 50 Ohm with no reactance then any length of 50 Ohm Coax should work.

Is this correct?
suxi is offline Find More Posts by suxi
Reply With Quote
Old Oct 13, 2014, 09:37 PM
Registered User
Mr.RC-CAM's Avatar
USA
Joined Oct 2001
5,213 Posts
I don't like to comment in detail on the "cable length antenna tuning" discussion because it sometimes ends up in a nasty purse fight. But I follow these concepts:

1. The coax cable is not tuning the antenna (load). The observed measurement is of the Antenna System, which involves any impedance mis-match between the antenna, the coax cable, and anything else in between. Furthermore, the input impedance of the cable will be affected by it's length and the RF frequency because of the mismatched load. These things are the basis of the white paper in the previous link.

2. RG316 and RG174 have different velocity factors. So identical physical lengths would have different electrical lengths.

3. Ignoring any serious resistive losses, if the coax test cable, source, and load are all 50 ohms Z, then your cable would have no significant affect on the measurement. But that is theoretical perfection.

- Thomas
Mr.RC-CAM is offline Find More Posts by Mr.RC-CAM
Reply With Quote
Old Oct 14, 2014, 04:16 PM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
After reading the great article posted by Mr.RC-CAM several times I think I see clearer now ;-)

We are using untuned transmission lines (our simple coax cables) and our antennas must show a pure 50 Ohm resistance without any imaginary part (reactance) in order to perform at its best and measure reliable results regardless of the cable length.

My new understanding is: If we build and measure antennas with an attached coax of random length, the results shown by a SWR Meter will very likely be wrong as an antenna with a resistance unequal 50 Ohm together with a coax of a random electrical length will portrait an unpredictable result not in line with the actual antenna performance, which could be worse.

Am I on the right track?

Best wishes

suxi
suxi is offline Find More Posts by suxi
Reply With Quote
Old Oct 14, 2014, 11:14 PM
Registered User
Canada, BC, Capital
Joined Jan 2014
94 Posts
My understanding of wire length/diameter is their is a drop in gain with size.
I am far from an expert but have set up many point to multi point 2.4 GHz systems for remote wireless ineternet and the mantra was always the same limit connectors,short run and max gauge for the wires. There was math for everything, like 7 Db over 100 feet for LMR 75? A 6 foot run of LMR 195(used to pass a wall or reach a desk could disrupt service. Each connector was close to a Db.
Losses added up quick. The solution power over cat5. Radio, antenna and enclosure in the same spot. We have this in our flying frames.
The wireless companies now use a device enclosed as big as a small pizza box.
I see a lot of posts that recommend a cable for vtx relief. If it works great! Any problems and I think your gains could drop by almost half.
My teacher taught me simple rules; use the largest wire and least amount of connectors. Most important rule (while diagnosing on his $6000.00 scope) "This is an inexact science."
i have links still in use today over 35km over water.
On a side note I see that they using a modified patch for circular. Very interesting, we thought we were innovative using linear @ 45 degrees
Chad.
thenerd is offline Find More Posts by thenerd
Reply With Quote
Old Yesterday, 01:09 PM
Open Source It!
suxi's Avatar
Joined May 2013
177 Posts
This experiment backs my observations and I really don't know what to make from it.

The antenna with 7cm of rg316 soldered to it measures a perfect SWR of 1 with an impedance of 50 Ohm and no reactance. So it should be a perfect antenna.

But look what happens if I attach 19cm of a high quality aircell5 cable to it, which by itself and terminated with 50 Ohm also shows a SWR of 1: the resonance frequency shifts down considerably and this antenna system shows now a really bad SWR of nearly 2 at our desired frequency of 1280Mhz.

Any further thoughts on this?

Happy winter

Suxi
suxi is offline Find More Posts by suxi
Last edited by suxi; Yesterday at 01:29 PM.
Reply With Quote
Old Yesterday, 10:14 PM
Engineer for Christ
IBCrazy's Avatar
Amherst, VA
Joined Jun 2006
10,706 Posts
Suxi - touch the cable with your hand. Watch what happens. We call the skin effect

I won't go into elaborate detail about it as controlling it is a tightly controlled secret between a few of us manufacturers.

What you were seeing earlier in your helical, however, is a different phenomenon. It has to do mainly with loss in lines. As as wave travels down a line the impedance changes phase (think about a Smith Chart here). Loss in the cable would therefore change properties with cable length. We know that the impedance of a helical is ~110 ohms and tuned with a phasing stub of some sort. The phase of this stub can be additive to the loss value of the cable or it can be subtractive. This all depends on the properties of your cable and what length is used.

-Alex
IBCrazy is offline Find More Posts by IBCrazy
Reply With Quote
Old Yesterday, 10:17 PM
Engineer for Christ
IBCrazy's Avatar
Amherst, VA
Joined Jun 2006
10,706 Posts
Quote:
Originally Posted by Mr.RC-CAM View Post
3. Ignoring any serious resistive losses, if the coax test cable, source, and load are all 50 ohms Z, then your cable would have no significant affect on the measurement. But that is theoretical perfection.
I'm glad someone finally said this. I have found many lines that are supposed to have a characteristic impedance of 50 ohms to have very different impedance values. For example, I find that RG405 from Belden has a characteristic impedance of 62 Ohms @5.8GHz as measured on my VNA.

-Alex
IBCrazy is offline Find More Posts by IBCrazy
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Help! Poor loiter performance and pitching on yaw. JonnyJ Multirotor Electronics 0 Sep 23, 2014 02:29 PM
Discussion does the length of Coaxial Cable of 2.4 GHZ Antenna affect the performance of receive lazy-b Radios 23 Apr 30, 2011 02:19 PM
Discussion Will the diameter of some F3P motor influence F3P flying performance? YXmotor Power Systems 3 May 07, 2010 04:45 PM
Question Factors influencing low lift performance? dst Slope 3 Sep 18, 2006 12:36 AM