Coax to atu to antenna Z & SWR questions.

AM Radio discussion. Directional arrays are FUN!
Post Reply
Fran3
Posts: 59
Joined: Wed Jun 24, 2015 1:30 pm

Coax to atu to antenna Z & SWR questions.

Post by Fran3 » Tue Nov 17, 2015 5:25 pm

I'll probably show how little I know with this question but here goes.

Three tower directional. Two towers day and three towers night.

- Disconnected 7/8" heliax from output of phasor and from input to ATU.
- Terminate heliax at atu end in 50 ohms restive.
- Measure SWR and Z and it is good for all three cables. VSWR in range of 1.1 or less. Z near 50 ohm restive j zero.

- Now connect cable back to atu's and measure again from the disconnected phasor end.
- bad swr and Z but resonant at transmit frequency.

Here are the questions. Since the cable is 50 ohms shouldn't the input to the atu's be 50 ohms as well?
And shouldn't the output of the phasor be 50 ohms as well?

I thought it should but it get's complicated as when you switch from day to night not only do switches switch in the phasor cabinets but switches switch in the ATU's as well.

If the phasor output is always set to 50R j zero why switch the atu?

I've always thought the magic of a phase adjustment circuit was to adjust the phase of the signal while maintaining a 50 ohm j zero input and output impedance... and the job of the ATU was to match the 50 ohm j zero output from the coax/heliax cable to the actual impedance of the antenna.

Thanks for any confirmation of my thoughts or a little education.

User avatar
Deep Thought
Posts: 3364
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: Coax to atu to antenna Z & SWR questions.

Post by Deep Thought » Tue Nov 17, 2015 6:51 pm

Your assumptions are correct in theory but in practice they can be difficult to implement completely.

Very few directional antenna ATUs are tuned to exactly 50 j0 at the ATU...it isn't really necessary. I usually try for a match of under 1.5:1 or better, and if I have time get as close as I can to 50 j0 for best power transfer to the towers but as long as you are close and in spec for the transmission line ratings everything usually works fine without spending what can be a lot of extra time trying to squeeze out that last couple of decimal points. Every change you make at the ATU changes the operating parameters so you have to go back and readjust everything and then see what happened to the ATU match. It can take a LONG time.

However, transmission lines which are not terminated in their characteristic impedance (in this case 50 ohms) act as transformers, changing the impedance you see at the other end in ways depending on the length. They also "slow down" when mismatched, which can be a real issue in a DA since their length is part of the phasing system. If you have a network "looking" at that it won't behave correctly either since it is not working into its design load and that affects both the power transfer and the phase though it.

Another thing you found is the tower impedance changes when it is operating in the DA. This is caused by the mutual coupling between towers and the static self-impedance (even with the others connected) is usually much different than when the array is operating. This causes the ATU match to look incorrect unless the system is in operation. You need an operating impedance bridge to make that measurement.

The tower networks are switched between modes since the operating impedance changes, and therefore the network adjustment has to change to get back to 50 j0 at the input. It is also often necessary to use different network phase shifts for each pattern.

Your transmission line test is good for making sure all the lines are OK, but once you reconnect the ATUs all the loads change.
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

User avatar
kkiddkkidd
Posts: 690
Joined: Mon Dec 03, 2007 11:13 am
Location: Lawrenceburg, TN

Re: Coax to atu to antenna Z & SWR questions.

Post by kkiddkkidd » Tue Nov 17, 2015 7:34 pm

Deep Thought wrote:The tower networks are switched between modes since the operating impedance changes, and therefore the network adjustment has to change to get back to 50 j0 at the input. It is also often necessary to use different network phase shifts for each pattern.

Your transmission line test is good for making sure all the lines are OK, but once you reconnect the ATUs all the loads change.
Not to mention the 2 tower Day to 3 twr Night transition and the mutual coupling changes that inserting/removing the 3rd tower creates.

I had a local client with a 4 tower inline DAN that one of Mulaney Engineering's engineers (Ed something) tried to get the feed lines matched better. We worked on it for at least 3 or 4 days before Ed decided that we were just tormenting ourselves. 2 of the lines could be matched almost perfectly. 1 was just barely acceptable and the 4th (very low power) was horrible but every time we changed something to help the 4th line the whole array went toes up and we had to walk all those changes back. All lines were operating well within their voltage ratings at the power and VSWR, so we just threw up our hands and left it. I maintained that array for a number of years before Cumulus bought it and my services were no longer needed by their superior AM engineering division <g>. It was very stable for years after that tuneup.

The last I noticed, that array had been operating on STA's for over a year.

Later,
--
Kevin C. Kidd CSRE/AMD
WD4RAT
AM Ground Systems Company
http://www.amgroundsystems.com
KK Broadcast Engineering
http://www.kkbc.com

User avatar
Deep Thought
Posts: 3364
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: Coax to atu to antenna Z & SWR questions.

Post by Deep Thought » Tue Nov 17, 2015 10:32 pm

Yeah, most four tower inlines have a nasty habit of the major lobe end tower running negative...which makes matching the ATU input futile. The power flows *out* of the ATU and the load is actually the phasor power divider which recycles it into the system. For many systems it is easier to terminate the ATU in a 50 ohm dummy load, set up the rest of the towers, reconnect the ATU to the system and then adjust the phasor controls for that tower to get the parameters where they need to be. It can be a real challenge.
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

ChuckG
Posts: 971
Joined: Tue Mar 11, 2008 11:12 pm
Location: Moo
Contact:

Re: Coax to atu to antenna Z & SWR questions.

Post by ChuckG » Wed Nov 18, 2015 12:56 am

Fran3 wrote: If the phasor output is always set to 50R j zero why switch the atu?
.
It startles the mice in the tuning hut so they move out. :lol:

As DT said, In addition to matching the line, the ATU also contributes a specific amount of phase shift. Those jobs will require different settings for different patterns. Matching the lines is often secondary to getting a working, stable pattern.
One thing you'll find quickly while working on DA's is that everything you do affects everything else. It can be maddening.

You may find these useful:
http://www.crawfordbroadcasting.com/Eng ... hasing.pdf
http://www.jmu.edu/wmra-eng/archive/Mul ... tional.pdf
There is also a very good book by Jack Layton that others here may have in PDF form.
Last edited by ChuckG on Wed Nov 18, 2015 1:05 am, edited 1 time in total.
<><><><><><><><><>
Chuck Gennaro
Central Wisconsin

Kelly
Posts: 490
Joined: Thu Aug 21, 2008 5:32 pm
Location: Washington D.C. Area

Re: Coax to atu to antenna Z & SWR questions.

Post by Kelly » Wed Nov 18, 2015 8:25 am

I reminded myself the hard way one late night, that essentially the phasor is nothing more than a tightly coupled power divider. When something throws the balance out, you need to really think through what is happening and how to troubleshoot it. I had recently completed a diplexed array with a 6 tower 50kW DA1 at 950kHz and one at 850kHZ 10KW Day, 1kW night. Two of the six 950kHZ-only towers were negative.

I got called out late one night because according to the antenna monitor, the 50kW station array parameters went way out. Drove the hour to the site and confirmed something was up with one of the negative towers which returned current back to the phasor. Walked out to the dog house and confirmed there was no base current. So stupidly I thought; "I'll just take the tower with no current out of the circuit so I can troubleshoot what's going on." ""I mean hey, it's negative right?"

Opened the back of the phasor with HV gloves and proceeded to pull the J-plug connecting to the dead tower. I then realized maybe that wasn't the best idea when a giant continuous arched yellow flame of RF started jumping between the tower feed to the J-plug in my hand and a 3" copper transmission line about three feet from the phasor. I quickly slammed the J-plug back into the clips, smelling burning hair. In this case, it was my right eyebrow that was now missing. After gathering my wits and now much more awake, since it was after midnight I switched the station to ND on the reference tower and was able to troubleshoot the failed capacitor in the Tower 6 ATU.
Skype:kellyalford Twitter: @KellyAlford

User avatar
kkiddkkidd
Posts: 690
Joined: Mon Dec 03, 2007 11:13 am
Location: Lawrenceburg, TN

Re: Coax to atu to antenna Z & SWR questions.

Post by kkiddkkidd » Wed Nov 18, 2015 8:30 pm

Deep Thought wrote:Yeah, most four tower inlines have a nasty habit of the major lobe end tower running negative...which makes matching the ATU input futile. The power flows *out* of the ATU and the load is actually the phasor power divider which recycles it into the system. For many systems it is easier to terminate the ATU in a 50 ohm dummy load, set up the rest of the towers, reconnect the ATU to the system and then adjust the phasor controls for that tower to get the parameters where they need to be. It can be a real challenge.
I don't recall if it was the front or 2nd tower that was negative but the negative tower wasn't the main problem. The problem child was the back tower which was very low power and as I recall just barely positive Z itself. Ed (I wish I could recall his last name) kept saying that it shouldn't work this way but it worked that way time after time. We could get 3 towers reasonably well matched and start working on the low power tower and it turned to crap. We tried both starting and ending with the back tower and every time it caused the array to depart to the point that the other towers had to be retuned which caused the line Z on the back tower to go nuts again.

We finally gave up since it was fed with 7/8 carrying something like 25 watts. Ed even thought that it would work as a 3 tower and for S & G's, we tried to tune it up that way (with the back tower detuned). Nope... never could get one of the back MP's below max just using 3 towers. The MP's were on very steep skirts of the pattern which made every adjustment very sensitive. This site was originally licensed as a "critical" array.

We also discovered that the array had been built to magnetic north. MD here is only about 3 degrees but it was enough to make it a pain getting the MP's on one side below max.
--
Kevin C. Kidd CSRE/AMD
WD4RAT
AM Ground Systems Company
http://www.amgroundsystems.com
KK Broadcast Engineering
http://www.kkbc.com

Post Reply