eHam Forums => Elmers => Topic started by: N4NOO on September 20, 2008, 11:22:31 AM

Title: OK, Answer this one . . .
Post by: N4NOO on September 20, 2008, 11:22:31 AM
I have a radio output on 445MHz at 20 watts going to a jumper to a watt meter to an antenna. SWR is 1.1:1. I have three perfect, new jumpers all  made from new 9258 with new conectors. All three jumpers have no defects. The jumpers are 8", 13", and 18". One jumper shows 6 watts, one shows 16 watts and one shows 20 watts to the meter input. The jumpers are the only things that are different. The SWR doesn't change. Only the power level to the meter input. Why?

Title: OK, Answer this one . . .
Post by: W5DXP on September 20, 2008, 12:30:38 PM
Is it a directional wattmeter or not?
73, Cecil,

Title: OK, Answer this one . . .
Post by: N4NOO on September 20, 2008, 12:49:31 PM
All readings were taken with the meter in the forward power mode only. Other than changing the jumpers, nothing else was changed during the test.

Title: OK, Answer this one . . .
Post by: WB2WIK on September 20, 2008, 01:12:58 PM
What is the transmitter, exactly?

Sounds like the coax patch cables are creating various impedance loads at the TX output port which result in varying output levels, with the "highest" one being the sweet spot that matches the transmitter tune point best.

Normally if the load SWR was really perfect (1:1) this shouldn't happen, but it's pretty common when the load is less than perfect.

Can you re-tune the output loading adjustments in the transmitter?  If so, you may find that any one of the cables produces "full" output, as long as you tune up into that combination (cable + meter + antenna).


Title: OK, Answer this one . . .
Post by: VK1OD on September 20, 2008, 02:42:44 PM

N4OO said "The jumpers are the only things that are different."

I am assuming that the jumper appears between the transmitter and directional power meter, and that ALL connectors are properly tightened.

Changing the length of the jumper changes the transmission line length and given that there are standing waves present, your transmitter sees a different load impedance if the load impedance at the far end of the coax is unchanged. Given the change in tx load impedance, it may deliver a different power level.

Next issue is that of the load impedance at the far end of the coax, is it unchanged?

Common mode current flows on the outside surface of the outer conductor.

You have changed the length and layout of the common mode current path, which may affect the load impedance that appears across the coax terminals at the far end. A change in this will result in a change in the load at the transmitter terminals and may result in changed power levels.

Imperfections in the directional power meter may contribute to your observations.

These are just a few possible explanations for your observations.

Your experimental results do not prove any hypothesis that certain length jumper is required or always delivers maximum power.


PS: Directional power meters such as a Bird 43 are often built out with a patch cable to an electrical half wave at the frequency of measurement (ie patch cable + Bird = 180°). There is a sound technical reason what this achieves, and the circumstances in which it is of benefit.

Title: OK, Answer this one . . .
Post by: W8JI on September 20, 2008, 04:51:58 PM
My bet is something is causing impedance bumps in the line or the antenna system SWR is not 1:1.

What connectors are you using?
What meter are you using?
What type of coax (brand and part number)?
What is the SWR?

If the power reading changes that means the radio is not seeing a constant impedance. That will happen with a line mismatch.

73 Tom

Title: OK, Answer this one . . .
Post by: AB7E on September 20, 2008, 05:36:58 PM
Put the wattmeter at the output of the transmitter (without a jumper ... use a male-to-male coupler or whatever connector may be appropriate), and then put the jumpers one after another between the meter and the antenna.  I'll bet you see a difference in SWR readings now.

The way you described it, the meter is always going to see the same SWR because its load is always only the antenna.  The transmitter, however, sees the antenna plus the meter plus the jumpers, and the jumpers being different lengths can present different SWRs to the transmitter, especially if your jumpers are not the same impedance as everything else.

Dave   AB7E

Title: OK, Answer this one . . .
Post by: WB2WIK on September 20, 2008, 06:13:06 PM
He already said he's using 50 Ohm coax..."9258" -- that's a Belden stock number for RG8X.

But I'm guessing the load SWR is not 1.1:1 as stated.


Title: OK, Answer this one . . .
Post by: N4NOO on September 20, 2008, 06:40:03 PM
Thank you one and all for your thoughts.
The radio is an ICOM-706MKIIG, the meter is a Yeasua VHF/UHF meter (can't remember the model number), Belden 9258 (50ohm RG-8X) jumper coax, Amphnol PL-259, 100 feet Belden 9913 (50ohm) run to a Diamond diskcone antenna, mast mounted at 30 feet off the ground.

I will try the jumpers on the other side of the meter tommorrow.

Many thanks again!

Title: OK, Answer this one . . .
Post by: W8JI on September 21, 2008, 04:06:05 PM
Wait a  minute people.

The SWR does not need to change to make the radio "like" or "dislike" the load.

The radio only need to see a different load impedance to change output power, not a different SWR.

Let's look at an example. The SWR could be 1.3 : 1 at the INPUT of the meter. Since the Yaesu is NOT a constant impedance line section like a Bird meter, it will add a bump in the line. It isn't unreasonable to cause a 1.5:1 SWR at the meter input even into a perfectly matched load when the meter is a cheap stripline type meter.

Add to the fact there are now multiple SO-239 females in line, and this is UHF.

Now let's assume it is only a 1.3 :1 at the SWR meter input port.

The rig could see as low as 38.5 or as high as 65 ohms even with a constant 1.3 : 1 SWR.

So as he changes the jumper lengths, the IMPEDANCE can go as low as 38 or as high as 65 even without an SWR change, and that is with a very modest 1.3 : 1 SWR.

Now if the factory matched the PA at 60-70 ohms, something that is common to reduce collector voltage on the PA transistor (tight coupling), it is possible the PA sees anything from an ideal load to nearly half the optimum impedance.

Moving the meter to the input of the jumper will measure errors in the line sections and connectors after the meter, or a null error in the meter itself in combination with a line section impedance error.

A cheap meter will never really tell us what is going on, it will only tell us the meter is bad or the jumper is not constant impedance. The only way to know what the error is, between the cheap meter and the jumpers, would be to use known good jumpers and a constant impedance line section meter like a Bird 43 with a good slug or an accurate impedance measuring device.

I fully expect with a strip-line type meter (the Yaesu) and SO239 connectors at UHF and RG8X cables the system has all sorts of impedance bumps affecting line impedance as length is varied.

I would be more shocked if it didn't behave that way knowing the meter, connectors, frequency, and cables used in the experiment.

73 Tom

Title: OK, Answer this one . . .
Post by: K5LXP on September 22, 2008, 06:07:25 AM
Using a termination instead of an antenna would help to answer some of these questions.

Mark K5LXP
Albuquerque, NM

Title: OK, Answer this one . . .
Post by: WA3SKN on September 22, 2008, 07:18:28 AM
Does anyone have a dummy load?
You need a known good reference to start the testing.