Does coax length really matter when tuning?

[W5LZ]

The thing is, it's all in getting the most power -to- the antenna. That means that there should be little or no losses in the feed line. To make sure of that the feed line should have the same impedance as the output of the transmitter. That means that there is little or no losses (other than what's normal from the resistance of the feed line length) out to the end of the feed line. Then there should be an impedance match between the feed line and the antenna so there sill be little or no losses generated there. In other words, feed line impedance and antenna feed point impedance should be the same. Nothing new there, right?
The problem is that antennas very seldom have an input impedance that's 50 ohms, or 75 ohms, or whatever the radio "wants" to see, without some means of "matching" (the coils etc. AT the feed point, not the 'loading' coils which make the antenna electrically 'longer' than it really is.). The impedance matching circuit is the 'key'. That circuit can be a coil, capacitor, certain length of coax (which works because of it's self inductance and/or capacitance), or a gamma match, and so on and so on. A "tuned" ciruit. Thats just fine and dandy until you realize that the thingy is only tuned for a narrow range of frequencies. It just ain't gonna work from daylight to dark, nothing 'fixed' tuned will.
The next thing to think about is that while your radio wants to 'see' a particular impedance, it's still going to be fairly 'happy' is the impedance it's looking at is close to what it wants to see. There's some leeway there. Not much with solid state radios, but some.
The one 'catch' to all this is that 'radiated power' is not developed in anything except a resistance. Inductance and capacitance (not really resistance) does not generate power. So if you look at the feed point's impedance [R value and X value] the "X" value should be as low as possible. (Readings from an antenna analyzer, don't you know, the "R" and "X" value thingys.)
Just like you feel the bumps in a car, your radio feels the impedance 'bumps'. If those 'bumps' are small, no big deal (but they DO add up!). If they are like the typical Oklahoma road pot-holes, then you got problems, and you should 'smooth' those impedance 'bumps' out so your radio doesn't feel them. Not exactly the most 'expert' way of expressing it, but you see what I'm saying (I hope!).
All of this is the same for ~all~ antennas no matter what it is or who makes it. And all we're talking about here is impedance matching so that the radio puts out all the power, doesn't 'cut back' because it doesn't like what it 'sees'. Doesn't say anything about the antenna's radiation pattern (where the signal goes once it leaves the antenna).

So, theres all of that stuff that you really don't care much about or want to know about, but which makes a big difference (sort of) in your signal...
- 'Doc

PS - If I'm charging for all this 'stuff', then I'm an expert and you're gonna pay for it. If I'm not charging for it, then it just means that I'm 'average', certainly not an expert (just remember some of what I read at one time or another). Feel free to mail your 'blank check' to my callbook address. Please notify me ahead of time though! At my age shock can cause heart attacks...

 

[freecell]

cb antennas aren't built with matching networks or tunable
roller inductors. for the most part they are able to be adjusted
for best match at resonance and then WYSIWYG.

a loading coil is nothing more than an inserted amount of inductance that cancels the capacitance introduced by antennas that are shorter than necessary for operation at the desired frequency. coaxial cable by its very nature is a series of lumped parallel capacitance and series inductance that represents combinations of L & C that can be used within certain limits to mitigate losses introduced into the feedline and specifically at the point where the transmitter represents a fixed impedance output (unlike amateur equipment) that is coupled to the feedline and can be used to minimize mismatch loss and restore as close to full transmitter output as possible when reflected power from a mismatched load exists on the line.

all power generated at the transmitter is ultimately absorbed and radiated by the load, regardless of the feedpoint impedance. this does not however preclude the fact that as reflected power on the feedline increases the amount of power available to the load from the transmitter is reduced because of the increasing mismatch between the transmitter and the feedline.

there are a few reasons why this is not seen or experienced in certain cases. #1, cb radios are designed at a fixed output impedance of 50 ohms while most (not all) hf and amateur equipment comes equipped with either an adjustable, wide-ranging pi-network or a built-in antenna tuner able to match the transmitter to mismatched impedances presented to it by the feedline connection varying anywhere from 16 to 150 ohms or better. #2, in the case of screwdriver antennas, we have an inductor that is able to be tapped for a wide range of inductances as the antenna is moved up and down the frequency range or from band to band to compensate for a number of factors, some not present and others not directly addressable in the typical cb antenna configuration due to the nature of both the equipment and the antenna type in use.

the ability to manipulate both the output impedance at the transmitter and input impedance at the feedpoint of the antenna precludes the possibility of ever seeing or experiencing what others do as they manipulate feedline lengths using antennas (and transmitters) without the previously mentioned features to accomplish basically the same thing, if only to a lesser degree.

 

[C W Morse]

O

Freecell,

That's the best answer to this "challenge" I have ever gotten!
I do believe that most of the solid-state radios are now 50 ohms unlike the old Swans and Heaths of yore who DID have pi-network or other matching methods. Yer still pretty much on your own WRT matching the antenna at the feedpoint. Also matching is not as critical as you go higher in frequency--more critical down below 40 and 75 Meters because a high Q antenna will generally exhibit a feedpoint impedance of around 17 ohms. It must, then, be increased to create a TOTAL system impedance of 50 ohms or thereabouts. This can be done with a small coil, capacitor, toroid, even a roller inductor if you want to get fancy. I favor the small coil at the feedpoint as it can be tapped while reading at 75 Meters and, for the screwdriver antenna, remain at around 50 ohms thru out the antenna's tuning range.

Still in the old days of the "Bugcatcher" antennas we used for years (and sometimes still do), I never found that fiddling with coax had any significant affect of the SWR or the performance of the antenna, and we often tuned these for multi-band ops using taps on the loading coil from 75 up to 10 Meters. Tuning was also very narrow (contrary to the CB gurus who think broadband antennas are "better") IOW, and generally speaking,
the broader the coverage, the sloppier the performance. Again, I don't believe the solid-state rigs are anything but 50 ohm output and rely on "foldback" to protect from high SWR.

The higher you go in frequency, the less matching is needed for the antenna it-SELF. Beyond 8 MHZ one usually doesn't need to use the normal matching devices and can rely on setting the length of the whip to bring the system to resonance. So I still can't understand why people cling to the "coax length" thing so adamantly. Oh well, whatever works. Some can keep trimming coax to a "certain" length, and I'll keep trimming the antenna.


CWM

 

[C2]

The impedance matching circuit is the 'key'. That circuit can be a... certain length of coax (which works because of it's self inductance and/or capacitance)...

The one 'catch' to all this is that 'radiated power' is not developed in anything except a resistance...

I guess a question once talked about was what really happens to the signal when you have an impedance mismatch with the load? I read someplace, maybe in a reflectionary article, that a matching circuit, such as an antenna tuner, does not really match the load to the source. Then I consider the signal that travels to the mismatched load; does it reflect back to the impedance transformer and again bounce back to the mismatched load, back and forth until all has been radiated? Or does the matching circuit actually transform the source/load match such that there are no reflections at all?

If the latter is true, then using the coax as the impedance transformer does indeed lower the SWR at the antenna feedpoint and result in a matched system, albeit less than ideal since optimum power is not developed.

Since, for a mismatched load, the coax itself can influence the circuit, you would then conclude that a specific length of coax would be called for, but only in the context of understanding transmission line characteristics.

When the transmission line's electrical length is some exact 1/2 wave multiple, the impedance at the terminated end is reproduced exactly as it is. A short will look like a short, and an open will look like an open.

When the transmission line's electrical length is some exact odd 1/4 wave multiple, the impedance at the terminated end is reproduced exactly opposite as it is. A short will look like an open, and an open will look like a short.

Some lengths in between will have their own unique effect.

For 1/4 wave line lengths, the relationship of the line impedance to the source and load impedance can be shown to be:

Zline^2 = Zsource x Zload

We can use this to our advantage, for example, to match a full wave loop antenna. Here, a typical antenna impedance is 100 ohms, the source is 50 ohms, so the line impedance should be roughly 70 ohms and 1/4 wave long to match the antenna and source. We typically use 75 ohm line for this purpose.

The point here then would seem to be, that for our systems where we are trying to match a 50 ohm load to a 50 ohm source, we would want to use line lengths of exact 1/2 wave multiples so that the load impedance is duplicated at the source (or meter). Some other random length might otherwise influence the measurement and "fool" your meter, as odd as that would seem to be.

Of course, after the adjustments have been made to your antenna when using a line length of exactly a 1/2 wave multiple and you have acquired a "perfect" match, line length is purely academic and will, thusly, be of no consequence, assuming of course that it is capable of reaching from the source to the load.

Yet, we again have something else to consider. And that is loss. The incident signal will experience some line loss on its way to the load. Given a mismatch, the reflected signal will again experience loss on its way back to the source. As most of us do, our SWR meter will be located at the source. The result will then be an apparent better match than what is actually true. This is because the meter will see the full forward power, but will see much less of the reflected signal than is actually the case. This is a minor issue, but coupled with the discussion of line lengths, it seems appropriate to dismiss the entire argument and simply put the meter at the antenna's feedpoint, make the adjustment for ideal match, and worrly no longer about how the length of your transmission line affects your SWR meter.

If it is of interest, for CB channel 19 (center band), the 1/4 wavelength multiples of RG58/U coax with velocity factor of 0.66 are as follows:

1/4 wave: 5.68'
1/2 wave: 11.36'
3/4 wave: 17.04'
1 wave: 22.72'

 

[jazzsinger]

For 1/4 wave line lengths, the relationship of the line impedance to the source and load impedance can be shown to be:

Zline^2 = Zsource x Zload

We can use this to our advantage, for example, to match a full wave loop antenna. Here, a typical antenna impedance is 100 ohms, the source is 50 ohms, so the line impedance should be roughly 70 ohms and 1/4 wave long to match the antenna and source. We typically use 75 ohm line for this purpose.

your theory here is sound,but as generally you would know the source and the load impedance but want to find the 1/4 wave line impedance to match them, a better way (albeit it means exactly the same as you say) of writing it would be Z line = sqrt of Z load x Z source.

which for your example is exactly what you say,sqrt of 50 x 100 = sqrt of 5000 = 70.71 ohms.

If it is of interest, for CB channel 19 (center band), the 1/4 wavelength multiples of RG58/U coax with velocity factor of 0.66 are as follows:

1/4 wave: 5.68'
1/2 wave: 11.36'
3/4 wave: 17.04'
1 wave: 22.72'

i hope you don't mind me pointing out this minor error C2,as i think it may benefit lots of people who picked up the same mistake from the same source.

in these examples you have the wrong lengths which if you were using them, as 1/4 wave or 3/4 wave inverters you wouldn't get the match as close as you would expect.

the reason you've got the wrong lengths is you used the so called "magic number" approx .95 or 95% of free space for cutting dipoles rather than the actual wavelength in freespace.you used 934 instead of 984 which is 5% less, which leaves all your coax lengths 5% short.

984 comes from 300 (velocity of radio waves) x 3.28 ( to convert meters to feet)
multiply that 984 by .95 and you get 934.8, theres no magic involved if you understand the theory,using magic numbers makes you lazy,and its very easy to get them mixed up.better to learn the mathematics for all situations.

anyway back to my point,

984/27.185 = 36.19 feet (a fullwavelength in free space at 27.185 mhz)

36.19/4 = 9.04 feet (1/4 wave in free space)

now figure in that coax velocity factor of .66 x 9.04 = 5.97 feet so there you have it,a 1/4 wave at 27.185 in .66 vel. factor coax = 5.97 feet or rounded up 6 feet.

if you want to check that i'm telling you the truth multiply your 5.97 feet by .95 and i guarantee you get the figure you quoted of 5.68 feet, infact its actually 5.67379 feet, but as you no doubt rounded up to the nearest hundreth of a foot you get 5.68 feet,as near as dammit.

you just made the exact same mistake lou franklin famously made in his screwdriver experts guide or his understanding and repairing cb radios (can't remember which), when i first read it i noticed his mistake right away as my strongest point is arithmetic, but through reading thousands of posts on many cb and ham forums many people took his word as gospel and never challenged it, incidentally his book was where i first learned that theory many years ago too,

being exceptionally good at arithmetic myself and having the sort of mind that questions everything numerical that don't look right i guess i was lucky in spotting a very very very easy mistake to make. i've spotted a few other glaring mistakes in his books too like claiming a moonraker 4 was quad driven when in fact the quad element is parasitic and its yagi driven, his books are excellent but do have a couple of niggly flaws.

 

[jazzsinger]

Freecell,

That's the best answer to this "challenge" I have ever gotten!
I do believe that most of the solid-state radios are now 50 ohms unlike the old Swans and Heaths of yore who DID have pi-network or other matching methods. Yer still pretty much on your own WRT matching the antenna at the feedpoint. Also matching is not as critical as you go higher in frequency--more critical down below 40 and 75 Meters because a high Q antenna will generally exhibit a feedpoint impedance of around 17 ohms. It must, then, be increased to create a TOTAL system impedance of 50 ohms or thereabouts. This can be done with a small coil, capacitor, toroid, even a roller inductor if you want to get fancy. I favor the small coil at the feedpoint as it can be tapped while reading at 75 Meters and, for the screwdriver antenna, remain at around 50 ohms thru out the antenna's tuning range.

Still in the old days of the "Bugcatcher" antennas we used for years (and sometimes still do), I never found that fiddling with coax had any significant affect of the SWR or the performance of the antenna, and we often tuned these for multi-band ops using taps on the loading coil from 75 up to 10 Meters. Tuning was also very narrow (contrary to the CB gurus who think broadband antennas are "better") IOW, and generally speaking,
the broader the coverage, the sloppier the performance. Again, I don't believe the solid-state rigs are anything but 50 ohm output and rely on "foldback" to protect from high SWR.

The higher you go in frequency, the less matching is needed for the antenna it-SELF. Beyond 8 MHZ one usually doesn't need to use the normal matching devices and can rely on setting the length of the whip to bring the system to resonance. So I still can't understand why people cling to the "coax length" thing so adamantly. Oh well, whatever works. Some can keep trimming coax to a "certain" length, and I'll keep trimming the antenna.


CWM

i think you've missed the point on the screwdriver working with any length of coax CW, its a different ballgame as it has its own rem control/auto matching networks built in unlike cb antennas.once its brought the antenna match close too 50 ohms the line length is irrelevant as it is 50 ohms at both ends,and all along it,

very few resonant cb 1/4 waves full length or shortened have a 50 ohm match at resonance,more like 36 ohms or 1.3:1 vswr approx for all those swar devotees out there,therefore there is a difference between the antenna impedance and the line/radio impedance. thats why when tuning a cb antenna for resonance ie, not 1.0: vswr there will be an element of matching bythe coax unless you use a proper multiple of a 1/2 wave repeater. the same would apply to any single band ham antenna not using a matching network too.

too cut a long story short, if you see a 1.0:1 swr on a mobile antenna that isn't coil tapped for a 50 ohm point the same way a base station 1/2 wave or 5/8 wave is), the chances are you either got lucky with the coax length and it matched it for you,or your swr meter is innacurate, or more likely you have tuned the antenna off resonance and therefore it ain't operating too its full potential, i'm sure many people will have noticed a slight improvement in performance when they move to a channel that is resonant instead of the one they thought was resonant because their swr meter showed a 1.0:1 swr.

i hope that helps reinforce the point freecell was trying to put to you in simpler less technical terms.

 

[RatsoW8]

I'm not an engineer by any means but I've picked up a couple things about mobile antenna installs along the way. Maybe these are off but they seem to hold true in practical use.

1- If you are using coax length to match the antenna to the transmitter you're basically causing the coax to become part of the radiator itself. Not good. You want to tune the antenna for resonance not the feedline.

2- Even the best mobile antenna installs are only going yield to 7-10% effieciency if your lucky so do it the right way the first time. Most of your precious mobile power is gone to ground losses.

3-Mount the antenna to the body of the vehicle as high up as possible with the most metal underneath it. No mag mounts, lip mounts, blah, blah , blah..drill some holes and mount and use a real antenna.

4-Bond the metal components of the vehicle together. I.e., hood, fenders, trunk lid,engine block, etc with braided ground strap.

When I've followed these basics things work as they should. I've had to compromise at times when I couldn't drill holes and use big antennas on my vehicle but I also know not to expect much in the way of performance when I do have to compromise.

 

[jazzsinger]

1- If you are using coax length to match the antenna to the transmitter you're basically causing the coax to become part of the radiator itself. Not good. You want to tune the antenna for resonance not the feedline.

the idea isn't to use coax to match the antenna, the idea is to use a 1/2 wave multiple impedance repeater when tuning to ensure you are NOT using the coax to match the antenna (its the only way to be certain), once you have the antenna matched you can use any length of coax you want as the line will be flat.

that single fact is what causes so much confusion, the measured length is for when your matching the antenna to coax and radio which are the same impedance,once its matched any length does, which i believe was a point DOC made earlier in this thread. if you use any other length during matching the coax will transform the measured impedance/vswr to some degree.

2- Even the best mobile antenna installs are only going yield to 7-10% effieciency if your lucky so do it the right way the first time. Most of your precious mobile power is gone to ground losses.

3-Mount the antenna to the body of the vehicle as high up as possible with the most metal underneath it. No mag mounts, lip mounts, blah, blah , blah..drill some holes and mount and use a real antenna.

4-Bond the metal components of the vehicle together. I.e., hood, fenders, trunk lid,engine block, etc with braided ground strap.

When I've followed these basics things work as they should. I've had to compromise at times when I couldn't drill holes and use big antennas on my vehicle but I also know not to expect much in the way of performance when I do have to compromise.

cant argue with any of your other points,they all make for sound practice,especially number 3.

 

[hotrod]

well, I just tuned Kale's new 12in dual coil SW series, and tuned it out perfectly with 17 feet of coax and a quad mag mount, and I didn't even have to cut anything (just pulled it out a few inches), and SWRs fell flat!! I then adjusted it to a 1.1 Never had one go that well before in just a few minutes time.

nice job. just curiouswhere you mounted the quad pad and 10k?
most /if not all users ive read about used these on semis on a miror mount
just curious. hope works good 4 ya

 

[Beetle]

Wow! A six-year-old thread rises from the dead!

Don't forget, when making these precise lengths of coax, that it will only be a half-wave (or quarter-wave, or whatever other fraction of a wavelength) at ONE frequency. Move a couple KHz and it's too long or too short.

 

[jazzsinger]

Wow! A six-year-old thread rises from the dead!

Don't forget, when making these precise lengths of coax, that it will only be a half-wave (or quarter-wave, or whatever other fraction of a wavelength) at ONE frequency. Move a couple KHz and it's too long or too short.

lol, my bad, i noticed someone reading the thread when i checked who was online and it got my curiosity, i have to admit i never noticed the date the last post was made, but in all fairness the points i made are still relevant 6 years later. physic laws rarely change much with time.

yep no matter how far you move either side of that centre frequency those precise lengths fall on their arse, but the point is to get the antenna initially peaked for centre frequency, that way performance drops off equally in either direction and to know coax is playing as small as possible part in the readings you get. failing that you will be constantly chasing your arse in the same way people chase their arse trying to get a mythical 1.0:1 vswr.

optimising an antenna system isn't for everyone, guys who only talk to a couple of locals and running 4w probably wouldn't notice the slightest bit of difference anyway, those working long haul dx or running mega power might well notice the difference as might those who live in highly urbanised areas susceptible to rfi,the last thing you want in that situation is coax radiating as part of the antenna system.

at the end of the day this debate/argument will continue long after i've quit radio.

 

[Kamikaze]

Wow! A six-year-old thread rises from the dead!

Don't forget, when making these precise lengths of coax, that it will only be a half-wave (or quarter-wave, or whatever other fraction of a wavelength) at ONE frequency. Move a couple KHz and it's too long or too short.

A couple KHZ will not make much difference in a tuned feeder.

 

[hookedon6]

A couple KHZ will not make much difference in a tuned feeder.

Yeah,...... but,...... what about a couple of CHANNELS?:bdh:

 

[Kamikaze]

Yeah,...... but,...... what about a couple of CHANNELS?:bdh:

I guess you mean 10 khz channels on 27 mhz.

It is still not critical.

 

[KM3F]

Some of the hams gave the correct answers to the question of coax length.
This is always a point of debate because of lack of understanding transmission line and antenna theroy.
Take it in two bites, then put them togather.
First, a coax nominal impedence is 50 ohms +/- a small amount.
The radio is designed to supply RF power into a 50 ohm load at it's end of the coax.
Second, the antenna impedence must be 50 ohms at the other end the coax is connected to.
When you have these conditions the match is said to be flat and resistive but only for a narrow frequency range because the "antenna impedance" changes with frequency.
SO this means a compromise over the channels from one end to the other.
Set antenna match for about the center channel and the match will degrade somewhwat as you move away from this center channel setting either way.
When the match is flat or nearly so, the coax length does not mean anything except the longer the more 'loss' it will add to loss of power at the antenna.
If there is mismatch at the antenna, NOW the coax length becomes part of the overall tunning. THE ABOVE IS THE TWO BITES OF IT.
You can see now that there are two parts to this often asked question.
To review, If the antenna offers a near perfect match to the feedline, the feedline length makes no differance in the short distance on a mobile or base.
If the antenna does not offer a good match, then the feedline length 'does' make a great difference becaue it's length (at some point) causes the radio to see the coax as 50 ohms due to the "reflected" impedence form the antenna end becoming just right at 50 ohms due to a length found to cause it to be so.
.
Additionally, using a watt meter sometimes shows more or less power than you think it should. Why is because the meter is simple and responds to the forward power and reflected power AND adds/subtracts the reflected power to the total reading making it appear there is more or less power. The meter is dumb and can't tell the different about which direction it should read so it integrates power from both directions.
Additionally if the reflected power is to high from mismatch, it may cause the radio to lower it's power, called foldback.
This often is a built in protection for the radio's final transistor or it will see the results of the missmatch from the coax all the way to the final device as possibly too high an RF voltage that can burn open the device junction.
The reason is while the radio works on 13 volts, the rf voltage developed can be several hundred times this value and burn open the device that is not made to withstand much over about 30 volts.
Please review this reply enough times to get the value and print it out to have as a reference.
Good luck.