Amp coax jumper length

[eagle1911]

Personally I think a tuning function on both the input and output is a great idea.. As was just said there is no "one size fits all" in real world designs, it's always about finding a point with the least negative aspects and the most efficiency. Also, there is always variation in actual components. You can calculate on paper, but the result of paper calculations, while close, are not going to be exact. Once you build the circuit you're then dealing with variations in the manufacturing processes of the various components, and with VHF and higher you're dealing with effects of the PCB and it's traces. Having a tuning function means you can have many of those variables and still tune for the best performance. When you acquire a system with fixed matching networks what you get is what you get, unless you want to start tracing down the components causing the variations and replace them with components you have measured to have more exact values.

 

[AudioShockwav]

Eagle, you are 100% correct, but they are dealing with builders that do not even want to add negative feedback on a Amp that has 2 output transistors in it.....

Real bare bones stuff, as few parts as possible.


73
Jeff

 

[eagle1911]

Unfortunately, with too much simplicity comes a profound lack of compensation for all of the myriad of variables at work in an amplifier, and that lack of compensation manifests itself as problematic output and other issues. If one were to sit down and really calculate all of the stuff going on in an amplifier on every level it's really pretty staggering. To achieve solid performance requires some extra care and extra circuitry to deal with keeping all those variables under control. Amplifiers are fairly simple in terms of analog circuitry, but there's a lot more to them than is paid attention to with such simple designs. Maybe in time these folks building them will learn more about what's really going on inside what they're building.

 

[W5LZ]

That last "maybe" is a very, very "iffy" one. I honestly don't expect to see it in my life time. Why should they, all things considered??
- 'Doc

 

[5150]

true true on most accounts , but however
"meterbating is meterbating"even if the source doesn't match the load and all points in between feed coaxial length
is at 50 and it is set at an interval of a multiple of a 1/3 in relation to a given multiple to find that sweet spot that the meter "see's " is only "fooling " your ego as a swr bridge doesnt see impedence mismatches , but see reflected wattage for a improperly set load

simply put , if the load is 50 ohm and the source is as well , but you put a mismatched "box in line "that meter isnt gonna tell ya its the box , its just gonna say theres a mismatch,.....

who's on first? comes to mind as this is one of the oldest arguements

 

[W5LZ]

You're right, but if the mismatch only occurs wiyh that 'box' in-line, then it's a pretty good indication that the 'box' is at fault. Right?
The catch to that is that an SWR meter can tell you anything conclusively...
- 'Doc

 

[wavrider]

A half wl with velocity factor figured into the physical length should allow an VSWR meter to display accurately, (more or less) what the input impedance is to that "whatever name brand" amplifier the exciter is transmitting into.

No different than tuning an antenna.

CK posted about this a while back, a 1/2wl length of coax basically shows the true impedance of what ever load it is connected to.


Cut your coax to 1/2 wl with the velocity factor figured into the equation, then connect it directly to your antenna and read the VSWR, remember whatever they are.

Now connect Brand X amplifier into the equation using the 1/2wl of coax from the radio to the brand X amplifier, read your VSWR.

Found the problem? Now it is ready to be fixed, or replaced.

 

[jazzsinger]

and there are no antennas that have a normally occurring 50 ohm input impedance without an impedance matching device being used in/on that antenna. Not a 1/4 wave, nor a 1/2 wave, a 5/8 wave, or pick any 'wave' you want, none of them are 50 ohms without help.
- 'Doc

Starduster ,m400 don't do too bad at being 50 ohms by design without matching network help.

as for that one size fits all, i doubt you'll find many coax's that are 50 ohm by design either, many are 51,52,53.5 etc ohm by design.

 

[W5LZ]

That Starduster and M400 may not do to badly, but how was a 50 ohm input impedance achieved? It wasn't done without 'help'. That 'help' doesn't have to me a coil or capacitor of whatever shape/form by the way. But they both do use an imedance matching device. Do you know what that/those devices are?
48 to 54 ohms isn't unusual for '50 ohm' coax. That's why it's called a 'characteristic impedance', right?
You also don't do much 'impedance transforming' by using the same impedance cable as you want to end up with. Bet you've heard of conjugate matching! If you want to end up with 50 ohms impedance (actually 50R+0J) you can't use 50 ohm coax to do it with. You can dang sure 'fool' that SWR meter though, it doesn't know what that 'J' thingy is on it's best day. (That 'J' is the same thing as the 'X' on atypical antenna analyzer. Except the 'X' can be either (+) or (-). It turns into a 'J' when you know whether it's a (+) or (-) or '0'.) I wish I had a nickle for every possible 'R' and 'J' where the 'R' isn't 50 and the 'J' isn't '0', and an SWR meter still thinks it's 50 ohms!
- 'Doc

 

[BOOTY MONSTER]

That Starduster and M400 may not do to badly, but how was a 50 ohm input impedance achieved?
- 'Doc

Antenna Notes

"As shown in the following graph, when the "radials" are at 0 degrees, the impedance is about 70 ohms, since we are really dealing with a vertical dipole. As the radials are raised to an angle of 90 degrees, the impedance drops to around 20 ohms. Notice that when the radials are at an angle of about 45 degrees, the impedance is very close to 50 ohms, which is similar to the inverted Vee."


but the star duster doesn't us a 45 degree angle .

i'm still wondering why 18 ft was recommended if the amps input is off ???

 

[ROC1]

i guess i focused too much on the 18 ft part . im also of the impression coax has to be coiled to form a balun to change impedance . for years ive been on this forum and the big brains have said over and over that coax length doesn't matter (other than loss when comparing a longer run to a shorter run) and that changing jumper lengths with a amp to get a better vswr doesn't really change anything . this is the first time i've herd of a jumper being a transmission line/changing impedance for a amp with poor input tuning . i've read about it with some weird (for a cb'er) antennas way outside of the cb band and figured it was one of them frequency dependent things..........

i still think ROC was wrong about the 18 ft comment , and that's what my initial post disagreeing with him was commenting on . he got pissed and flamed me and i got pissed and rode his azz about it . it seems at this point that i owe him an apology since his "Or, if it's not and you have the means to model the transmission line you can cut it specifically to present a 50 ohm load to the radio - assuming the load at the input isn't waaaay out there, which it shouldn't be." comment has merit . i think i do well when folks explain why im wrong (when i am) without flaming me , but when i get flamed it kinda pisses me off .

ROC , i apologize .

I think I said 1/2 wave length, not 18ft which for most coax - when you adjust it for velocity - is more than half a wave length. When I was reffering to the use of the 1/2 wl jumper it was just to present a "non-transformed" impedance to the radio. Coax that is electrically 180 degrees long (half a wave length, 1/4 wave is 90) will give us the same input impedance (Zi) as the load impedance (ZL) that we see at the antenna mount, or amp etc regardless of the lines actual impedance (Z0). This should be the case for any multiple of half-wave lengths adjusted for line losses - which again in a mobile environment is about nill.

Here's an interesting example.
The wavelength of 27Mhz is roughly 36 feet long. A half-wave length would be 18ft if velocity factor wasn't part of the equation.

Let's assume we've got some dollars to spend, lot's of room and arms that look like popeye and we want to run 1/2 inch hardline in our Geo Metro.
1/2" 50 Ohm Hardline with a velocity factor of .81 works out like this; 36/2*.81 = 14.58 ft. So in this case about 14.5 feet will give us at the radio what we see at the antenna mount or amp. Let's say in this case we've run it to the antenna mount where we have a 40 ohm load (probably a 102 inch whip with some ground losses in there) and a Reactance of 0 (it's on frequency). At the input end of our hard-line we should see 40.12 -j0.25 or very close. The point is, almost identical to what we had at the antenna mount...

Now let's assume you have spent the day arguing with me and you weren't thinking straight when you bought your hard-line and accidentally bought 75 ohm line instead. 1/2 wave-length of this line gives us 40.36 -j 0.75...

Again - when a 1/2wave length of coax is used the input impedance will equal the load impedance regardless of the lines characteristic impedance (minus line losses).

This gets more interesting when you begin to look at the quarter wave transformers and what you can accomplish by using them but that a story for a different day.

 

[BOOTY MONSTER]

i understand about using velocity factor and frequency to determine the measured length of an electrical half wavelength , and that a physical half wavelength is longer than the electrical one . i guess the 18 ft you posted was a typo , no biggie .

"If the amp is presenting a 50 ohm load at the input you can run any length of coax and it will present 50 ohms to the radio. If it's not you're better off cutting the coax at 1/2 wave-length. Or, if it's not and you have the means to model the transmission line you can cut it specifically to present a 50 ohm load to the radio - assuming the load at the input isn't waaaay out there, which it shouldn't be."

you did say 1/2 wave , my bad . but i still fail to see how that would help if the amp isnt showing the radio 50 ohms .

maybe im too dense

 

[ROC1]

i understand about using velocity factor and frequency to determine the measured length of an electrical half wavelength , and that a physical half wavelength is longer than the electrical one . i guess the 18 ft you posted was a typo , no biggie .

"If the amp is presenting a 50 ohm load at the input you can run any length of coax and it will present 50 ohms to the radio. If it's not you're better off cutting the coax at 1/2 wave-length. Or, if it's not and you have the means to model the transmission line you can cut it specifically to present a 50 ohm load to the radio - assuming the load at the input isn't waaaay out there, which it shouldn't be."

you did say 1/2 wave , my bad . but i still fail to see how that would help if the amp isnt showing the radio 50 ohms .

maybe im too dense

It doesn't help... that is a two part statement. If you have the means to model the transmission line (and can do the coax ends etc) cut it at the right length to give you the 50 ohms on the input side - this of course is a "within reason" statement. If the amp is wtf off that isn't going to help and as has been mentioned - the amp input tuning should really be fixed - but that can be hard.

If you can't do that IMHO you're better off just running the 1/2 wave length just to keep it simple... at that length the radio will see what the amp is putting out which if the amp is remotely close will be fine. Additionally it makes is easier to determing the tuning of the amp. If it's way off you'll know it - running some odd length in-between the amp and the radio, you won't know if it's the lenght of coax, the amp or some combination of the two.

I'm going to go drink a dozen beers now and remind myself to never post anything again on this forum...

 

[AudioShockwav]

"If the amp is presenting a 50 ohm load at the input you can run any length of coax and it will present 50 ohms to the radio.

I think you understand that one
If you do use a 1/2 wave it will show you what is REALLY going on at the end.

A 1/2 wave length ( with an antenna analyzer ) will show you what the mis match IS at load.
Then you can Fix what ever is throwing a monkey wrench into the system, such as tuning the input of the offending amp.
You can use the coax as a measurement tool

Or, if it's not and you have the means to model the transmission line you can cut it specifically to present a 50 ohm load to the radio

Or you can , by trial and error, trim the coax ( this time using it as a impedance transformer) to get the SWR down to a acceptable level.
This is where 5150`s post about "meterbating" comes in.
You have not solved the problem of the bad input on the amp, you simply use the coax to allow the SWR meter ( and radio ) see a load that it wants to work into.
The input on the amp is still living in screwiness, mismatch.
Is that any help....or no?

73
Jeff

 

[AudioShockwav]

Looks like Me N ROC where posting at about the same time, my bad.

73
Jeff