Just curious, how many people tune up the duplexer with a spectrum analyzer/tracking generator plus accurate 50 ohm terminations on unused ports, then put them in service? (I have)
How many do all that and then fine-tune them on a weak signal while keying up the transmitter (hopefully having disconnected all the test equipment!)
duplexer tuning
Moderator: Queue Moderator
-
MO_TRASHER
- Posts: 205
- Joined: Thu Apr 18, 2002 11:13 am
Howdy, I use a spectrum analyzer on the otput and a signal generator on the input when tuning notch type duplexers. It seems to work quite well. The shop I'm at doesn't have a TG. I had a Cushman 5110 at a shop I was at "in my younger days", it had a built-in TG and was FANTASTIC! I could set up duplexers in minutes. Worked nice with combiners, too.
I haven't tried to do a pass/notch type with the regular generator and spec.analyzer, yet.
I haven't tried to do a pass/notch type with the regular generator and spec.analyzer, yet.
The 'Trasher
N3UHD
PCIA #38104
...If it ain't broke, don't fix it....
"Motorola hasn't built a radio so well that it can't be broken"
Check Out: http://www.lowerytech.net/forum
Join the fun!
Someday this'll be a REAL Site!.......
http://www.busrad.com
N3UHD
PCIA #38104
...If it ain't broke, don't fix it....
"Motorola hasn't built a radio so well that it can't be broken"
Check Out: http://www.lowerytech.net/forum
Join the fun!
Someday this'll be a REAL Site!.......
http://www.busrad.com
HI:
I can offer to you that duplexer tunning
is a real art. Requires Special Cables, the
proper service equipment, and Spectrum
Analizer is a plus.
Spectrum Displays can " reveal " hidden
attributes about how the duplexer will
perform, and there are " many " duplexers
out in the world today.
Although most will work, many maybe less
than perfect, and are not always set up
properly which will cause horrible performance.
I have tuned many [ Band Pass ] and [ Band Pass / Band Reject ]
Duplexers and have had to toss a number of them away do to
excerssive loss, instablity, or just plain bad.
And yes there are many that will work well one day, and
as a temp change is seen, they will not work well the next.
One reason to avoid some of the 25.00 ~ 50.00 types, as one
may spend 100.00 ~ 150.00 in labor just to discover the unit
is bad.
Best to look for duplexers that are not banged up, bent, damaged,
and perhaps only been tuned once or twice.
Duplexers that have alot of damage to the top of the tunning
screws, I would be a little suspicious.
On a more postive note....Most good mfg;s like Celwave, ect, will
be happy to alighn your duplexer for a fee, or there are persons
out there ( including us ) that can due the same thing.
One thing for sure, select the Duplexer for the proper application
for its use. In other words, don't expect a Band Reject Duplexer
to work well next to a 60 Kilowatt UHF TV station transmitter.
Monty
I can offer to you that duplexer tunning
is a real art. Requires Special Cables, the
proper service equipment, and Spectrum
Analizer is a plus.
Spectrum Displays can " reveal " hidden
attributes about how the duplexer will
perform, and there are " many " duplexers
out in the world today.
Although most will work, many maybe less
than perfect, and are not always set up
properly which will cause horrible performance.
I have tuned many [ Band Pass ] and [ Band Pass / Band Reject ]
Duplexers and have had to toss a number of them away do to
excerssive loss, instablity, or just plain bad.
And yes there are many that will work well one day, and
as a temp change is seen, they will not work well the next.
One reason to avoid some of the 25.00 ~ 50.00 types, as one
may spend 100.00 ~ 150.00 in labor just to discover the unit
is bad.
Best to look for duplexers that are not banged up, bent, damaged,
and perhaps only been tuned once or twice.
Duplexers that have alot of damage to the top of the tunning
screws, I would be a little suspicious.
On a more postive note....Most good mfg;s like Celwave, ect, will
be happy to alighn your duplexer for a fee, or there are persons
out there ( including us ) that can due the same thing.
One thing for sure, select the Duplexer for the proper application
for its use. In other words, don't expect a Band Reject Duplexer
to work well next to a 60 Kilowatt UHF TV station transmitter.
Monty
Need help on a Program Cable or (/\/\) Rib Box ?
Contact me anytime/All availible / Free Support Info !
mail to: MSisco9939@aol.com
PLease avoid PM's whenever possible. Box gets too full.
Contact me anytime/All availible / Free Support Info !
mail to: MSisco9939@aol.com
PLease avoid PM's whenever possible. Box gets too full.
Not wishing to defame any company or get into a P match with anyone on here, there are a couple of things to look for in used duplexers.
Like was said, be careful of bent and damaged cans. Some dings are OK, but any major hits will take some careful looking and maybe even some testing.
Look at the tunning rods. Are they in good shape or bent? Are the threads all banged up or look clean? Do all the cavities seem to have the same type of material used for the tunning rods? These rods are a special material that has a much better temperature stability than regular steel and brass. Are the locking nuts on the tuning rods?
Look at the connectors. Make sure that the center conductor has all the pins. If someone wasn't careful while installing a cable, it is easy to bend and break the center pin connection fingers.
Look at the cables with the duplexer. They should be a double shielded cable or "Heliax" type coax cable. Use of single shielded coax cable will cause too much signal leakage and probably desense to a repeater. Make sure the connectors are tight on the coax cables. A loose connector will drive you crazy.
The real hidden issue comes in how the cavities are constructed. One way of making the center conductor tie to the top plate where all the connections are is to use pop rivits. I have re machined many cavities where the center rod would move and have a poor electrical connection as you tuned it. The problem is that pop rivits can not keep the center rod tight. They have a tendancy to allow the center rod to move around and electrically make a poor connection. This causes the tuning to become very erratic and change drastically with temperature changes.
The solution is to drill and tap the mounting point. Then installing a good quality machine screw and tighten the center rod down. Unless your real good with a drill and tap, let someone with the propper tools do the work for you. It takes drilling the hole at a slight angle. Then tapping the hole for the size screw your going to use.
I also install a 4 screw point connection. The pop rivits are only a 3 point connection. It gives a more stable support for the center tuning rod.
If you end up with unstable tuning, it also could be due to a poor connection of the finger stock on the center rod. Over time the connection area will develope a poor connection. If your lucky, you will find a silver plated connection area on the center tuning rod. Do not polish this area very hard. The silver plating is very thinn and will just disapear with very little rubbing.
Have fun with your dulpexers.
Jim
Jim
Like was said, be careful of bent and damaged cans. Some dings are OK, but any major hits will take some careful looking and maybe even some testing.
Look at the tunning rods. Are they in good shape or bent? Are the threads all banged up or look clean? Do all the cavities seem to have the same type of material used for the tunning rods? These rods are a special material that has a much better temperature stability than regular steel and brass. Are the locking nuts on the tuning rods?
Look at the connectors. Make sure that the center conductor has all the pins. If someone wasn't careful while installing a cable, it is easy to bend and break the center pin connection fingers.
Look at the cables with the duplexer. They should be a double shielded cable or "Heliax" type coax cable. Use of single shielded coax cable will cause too much signal leakage and probably desense to a repeater. Make sure the connectors are tight on the coax cables. A loose connector will drive you crazy.
The real hidden issue comes in how the cavities are constructed. One way of making the center conductor tie to the top plate where all the connections are is to use pop rivits. I have re machined many cavities where the center rod would move and have a poor electrical connection as you tuned it. The problem is that pop rivits can not keep the center rod tight. They have a tendancy to allow the center rod to move around and electrically make a poor connection. This causes the tuning to become very erratic and change drastically with temperature changes.
The solution is to drill and tap the mounting point. Then installing a good quality machine screw and tighten the center rod down. Unless your real good with a drill and tap, let someone with the propper tools do the work for you. It takes drilling the hole at a slight angle. Then tapping the hole for the size screw your going to use.
I also install a 4 screw point connection. The pop rivits are only a 3 point connection. It gives a more stable support for the center tuning rod.
If you end up with unstable tuning, it also could be due to a poor connection of the finger stock on the center rod. Over time the connection area will develope a poor connection. If your lucky, you will find a silver plated connection area on the center tuning rod. Do not polish this area very hard. The silver plating is very thinn and will just disapear with very little rubbing.
Have fun with your dulpexers.
Jim
Jim
Re: duplexer tuning
That's pretty much how I fine-tune any duplexers I work on. Even with the fancy gear, I can almost always improve upon it by using my R-2670 in Duplex mode & running the repeater & duplexer on it, trying to meet the spec'd receive signal strength. Cleans up that last little bit of desense.DAL-COM wrote:Just curious, how many people tune up the duplexer with a spectrum analyzer/tracking generator plus accurate 50 ohm terminations on unused ports, then put them in service? (I have)
How many do all that and then fine-tune them on a weak signal while keying up the transmitter (hopefully having disconnected all the test equipment!)
Todd
-
Microwave Mike
- Posts: 259
- Joined: Fri Sep 14, 2001 4:00 pm
Duplexer tunning
If you do a lot of duplexer tuning, get a real network analyzer. The toys can get you close, but most people ignore the RX match to 50 ohms. The TX match will follow the path lose through the devise.
RX match is VERY important. If you don't get a good match, 20 Db min , then the low level signal will be reflected back out the antenna. Just like reflected Tx.
The hand held " Ham " type get you in the picture but have poor preformance at return losses lower than 15 Db.
A good HP will set you back around 3K less test set. That can run from 500-1500. This is all used prices. Try to get a 8753 series or 8711-14 series.
I have been running a 8753B for 12 years. Monocrome display but accurate measurments. One of these days I will upgrade to a D series color that works to 6 Ghz.
The secret to good accurate measurments is a Good high directivty testset and a newer analizer that will auto null your cables and test fixture.
mm
RX match is VERY important. If you don't get a good match, 20 Db min , then the low level signal will be reflected back out the antenna. Just like reflected Tx.
The hand held " Ham " type get you in the picture but have poor preformance at return losses lower than 15 Db.
A good HP will set you back around 3K less test set. That can run from 500-1500. This is all used prices. Try to get a 8753 series or 8711-14 series.
I have been running a 8753B for 12 years. Monocrome display but accurate measurments. One of these days I will upgrade to a D series color that works to 6 Ghz.
The secret to good accurate measurments is a Good high directivty testset and a newer analizer that will auto null your cables and test fixture.
mm
It looks like Todd is the only one who actually answered your question.
In the past, I have been tempted to do as you suggest, but mostly don’t believe it does any good. If the duplexer’s specs are well above the minimum that you need for the application, then there is no need to squeeze any more out of it anyway, provided that it is even possible to do so.
Touching up a duplexer in the actual system would make sense if the system impedances were not close to 50 Ohms. But this means that when the antenna or feed line gets changed, the duplexer would need to be optimized again as well.
This amounts to that I tune duplexers with a tracking generator for a 50-Ohm resistive load and do not touch them up in the system. I there is a desense problem, I either screwed up or I will find the problem elsewhere and correct it there.
Nand.
In the past, I have been tempted to do as you suggest, but mostly don’t believe it does any good. If the duplexer’s specs are well above the minimum that you need for the application, then there is no need to squeeze any more out of it anyway, provided that it is even possible to do so.
Touching up a duplexer in the actual system would make sense if the system impedances were not close to 50 Ohms. But this means that when the antenna or feed line gets changed, the duplexer would need to be optimized again as well.
This amounts to that I tune duplexers with a tracking generator for a 50-Ohm resistive load and do not touch them up in the system. I there is a desense problem, I either screwed up or I will find the problem elsewhere and correct it there.
Nand.
I don't have a problem with the idea that a knowledgeable tech could touch up a bandpass only device under power, but I would never recommend touching the adjustment of a notch filter that has been properly set up with a swept alignment.
You can get sort of close with a 2600 but it is limited to a 60dB on screen display range, has no normalization ability, and natters at you to "INCREASE ATTENUATION" if you try to run the analyzer (monitor) at a lower reference level than the generator output.
The HP8920 does a fine job, has 80 dB range, and has trace normalization. You can also use a real spectrum analyzer / tracking generator like a Tek 490 series + TR503, an HP 8590 series or one of several others from Agilent, Anritsu and Advantest.
All that aside, as Microwave Mike said: "get a real network analyzer"
I couldn't agree more.
You can't believe the results when you do a full OSL [Open, Short, Load] calibration at the measurement plane.
You can get sort of close with a 2600 but it is limited to a 60dB on screen display range, has no normalization ability, and natters at you to "INCREASE ATTENUATION" if you try to run the analyzer (monitor) at a lower reference level than the generator output.
The HP8920 does a fine job, has 80 dB range, and has trace normalization. You can also use a real spectrum analyzer / tracking generator like a Tek 490 series + TR503, an HP 8590 series or one of several others from Agilent, Anritsu and Advantest.
All that aside, as Microwave Mike said: "get a real network analyzer"
I couldn't agree more.
You can't believe the results when you do a full OSL [Open, Short, Load] calibration at the measurement plane.
My two bits worth:
The simple notch type duplexers are quite easy to tune with or with out all of the fancy test gear, as are the pass only units.
I presume if there is no measureable desense, the insertion loss is as specified, that the duplexer is adequately tuned.
I see no reason to calibrate yard sticks with a micrometer.
Many manufacturers forbid aligning cavaties under high power.
the danger is arcking the finger stock on the plunger. Choke type cavaties, such as Cellwave's 800 meg combiners are not so affected.
The simple notch type duplexers are quite easy to tune with or with out all of the fancy test gear, as are the pass only units.
I presume if there is no measureable desense, the insertion loss is as specified, that the duplexer is adequately tuned.
I see no reason to calibrate yard sticks with a micrometer.
Many manufacturers forbid aligning cavaties under high power.
the danger is arcking the finger stock on the plunger. Choke type cavaties, such as Cellwave's 800 meg combiners are not so affected.
Aloha, Bernie
So the answer to the original question is that normally you don't tune an in-service duplexer. Generally you will find that the manufacturer recommends agains tuning under power.
You tune on the bench using whatever process you have worked out that you are successful with, then you put it in service. You test for desense and proper power out (verifying no excessive insertion loss). If all is well - you are done. As NAND said if it doesn't work right - recheck your alignment or look for another problem.
As far as Bernie's comments about calibrating a yardstick with a micrometer - I don't find that to be an accurate analogy. We are not saying that you must use a network analyzer - only that it makes the job much easier. It's fast, powerful and accurate. Newer ones have 100 dB on screen allowing you to see deep notches. Try tuning a DB4062 with a 2600 - no way! The notches are over 120 dB down!
You use whatever tools you have or are comfortable with. This is more a case of the difference between Roy and Norm (The Woodright's Shop vs. The New Yankee Workshop). You can make furniture with a hand saw, a plane, a chisel, and a pocket knife if you want Bernie, but give me a Powermatic or Delta Unisaw any day(and a long bed jointer, and a mortising machine and....well you can't have too many toys)
BTW - the documentation for the new Smartzone 6 systems explains how to tune the transmitter combiners in the STR3000 base station racks using an Agilent 8714ES VNA. They tell you to do a one port OSL calibration plus through calibration or better yet - a full two port calibration. Then you hook up to the equipment and get started aligning.
The days of fixing radios are coming to an end - the future of two-way is systems support - infrastructure configuration and repair - and being really good at RF.
You tune on the bench using whatever process you have worked out that you are successful with, then you put it in service. You test for desense and proper power out (verifying no excessive insertion loss). If all is well - you are done. As NAND said if it doesn't work right - recheck your alignment or look for another problem.
As far as Bernie's comments about calibrating a yardstick with a micrometer - I don't find that to be an accurate analogy. We are not saying that you must use a network analyzer - only that it makes the job much easier. It's fast, powerful and accurate. Newer ones have 100 dB on screen allowing you to see deep notches. Try tuning a DB4062 with a 2600 - no way! The notches are over 120 dB down!
You use whatever tools you have or are comfortable with. This is more a case of the difference between Roy and Norm (The Woodright's Shop vs. The New Yankee Workshop). You can make furniture with a hand saw, a plane, a chisel, and a pocket knife if you want Bernie, but give me a Powermatic or Delta Unisaw any day(and a long bed jointer, and a mortising machine and....well you can't have too many toys)
BTW - the documentation for the new Smartzone 6 systems explains how to tune the transmitter combiners in the STR3000 base station racks using an Agilent 8714ES VNA. They tell you to do a one port OSL calibration plus through calibration or better yet - a full two port calibration. Then you hook up to the equipment and get started aligning.
The days of fixing radios are coming to an end - the future of two-way is systems support - infrastructure configuration and repair - and being really good at RF.
-
richyradio
- Posts: 83
- Joined: Fri Dec 13, 2002 1:52 pm
....Duplexer tuning
Hey, anybody remember the old Motorola "Green Box" MTS/IMTS
mobile phone? the big M's tuneup proceedure was to reverse the rx/tx ports and tune for min pwr/sens...found it a little humorous that they suggested you do it "quickly" (lest the varactor tripler starts hissing at you like a pissed off cat)....I tell you, I've used everything from a bearcat scanner L.O. (add 21.6 to desired frq) back in the dim time (high school days) up to the most expensive Hp ("High Price") network analyzers and pretty much you guys summed it up...me thinks your brain is the best piece of equip. on the bench...if you are acutely aware of what you wish to achieve all will be fine....I.E. don't sacrifice isolation for 1/2 of a db loss....a crappy repeater (power/sens. wise) at 500ft haat will always outperform the best setup 50 feet above the ground...not saying you should willy nilly disregard losses, but there is a balance to everything...
the one thing spec. analyzer/tracking generators are really handy for is when you are tuning one of those little 4 cavity preselectors...I always "skirt tune" those devices- they are so wide as to be practically useless if you center tune them...run your desired freq right at the skirt
(low side or high depending on what you might wish to reject) accept a tad more loss and it'll become a useful filter...I also like using 10 db pads right at the duplexer ports during initial tuning of untuned cans- speeds up things a bit so you don't have to go back and forth a second time to fine tune...
mobile phone? the big M's tuneup proceedure was to reverse the rx/tx ports and tune for min pwr/sens...found it a little humorous that they suggested you do it "quickly" (lest the varactor tripler starts hissing at you like a pissed off cat)....I tell you, I've used everything from a bearcat scanner L.O. (add 21.6 to desired frq) back in the dim time (high school days) up to the most expensive Hp ("High Price") network analyzers and pretty much you guys summed it up...me thinks your brain is the best piece of equip. on the bench...if you are acutely aware of what you wish to achieve all will be fine....I.E. don't sacrifice isolation for 1/2 of a db loss....a crappy repeater (power/sens. wise) at 500ft haat will always outperform the best setup 50 feet above the ground...not saying you should willy nilly disregard losses, but there is a balance to everything...
the one thing spec. analyzer/tracking generators are really handy for is when you are tuning one of those little 4 cavity preselectors...I always "skirt tune" those devices- they are so wide as to be practically useless if you center tune them...run your desired freq right at the skirt
(low side or high depending on what you might wish to reject) accept a tad more loss and it'll become a useful filter...I also like using 10 db pads right at the duplexer ports during initial tuning of untuned cans- speeds up things a bit so you don't have to go back and forth a second time to fine tune...
While on the subject, can anyone point me in the direction for some good directions on tunning the duplexers with a Tek 496 Spectrum Analyzer and the associated tracking generator? I've looked around but not found anything good on the subject. I need to retune one of the duplexers in our quantar repeater for a new frequency pair. A book, pdf file, text instructions, anything would be nice. Thanks!!! Everything I've found as far as instructions is how to do it without the proper test equipment and nothing of how to do it if you have the equipment. Everything says if you have the equipment you should know how to do it. Doh! Thanks!
Keith Dobbins N8KLD
IT Mainframe Network Engineer
W8TAP Repeater Group
Repeater Technician
Parkersburg, WV
IT Mainframe Network Engineer
W8TAP Repeater Group
Repeater Technician
Parkersburg, WV
"..can anyone point me in the direction for some good directions on tunning the duplexers with a Tek 496 Spectrum Analyzer and the associated tracking generator?..."
------------------------------------------------------------------------------------
Your 496 and tracking generator make a fine setup for aligning duplexers. At one time Tektronics used to publish application notes that would explain how these things work. Sometimes these notes show up on ebay, like this one:
http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... gory=25397
Basically the hook up is straightforward and since the 496 has digital storage with A-B trace capability, you can normalize out cable loss and see an accurate response on screen. The combination of spectrum analyzer and tracking generator comprise what is sometimes called a scalar measurement system [scalar meaning amplitude only] or Scalar Network Analyzer. That being true, material explaining network analyzers can help you understand the application of your equipment as well.
Agilent has a paper known as "Network Analyzer Basics" which you can download for free. Now I have always considered it somewhat of an oxymoron to use the word "basic" in the same sentence with "network analyzer", but if you have a high speed connection go to:
http://cp.home.agilent.com/upload/cmc_u ... BASICS.pdf
------------------------------------------------------------------------------------
Your 496 and tracking generator make a fine setup for aligning duplexers. At one time Tektronics used to publish application notes that would explain how these things work. Sometimes these notes show up on ebay, like this one:
http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... gory=25397
Basically the hook up is straightforward and since the 496 has digital storage with A-B trace capability, you can normalize out cable loss and see an accurate response on screen. The combination of spectrum analyzer and tracking generator comprise what is sometimes called a scalar measurement system [scalar meaning amplitude only] or Scalar Network Analyzer. That being true, material explaining network analyzers can help you understand the application of your equipment as well.
Agilent has a paper known as "Network Analyzer Basics" which you can download for free. Now I have always considered it somewhat of an oxymoron to use the word "basic" in the same sentence with "network analyzer", but if you have a high speed connection go to:
http://cp.home.agilent.com/upload/cmc_u ... BASICS.pdf
-
richyradio
- Posts: 83
- Joined: Fri Dec 13, 2002 1:52 pm
hmmm...I understand your ? , but then I'm not sure....meaning: do you know what style duplexer is in there? pass/reject/pass-reject? how far does it have to move? if its less 100khz away or so you probably wouldn't have to touch it (probably not that lucky 'tho) I'm assuming you're borrowing the tek 496 and the tr503 tracking generator? I figure if you owned it you would have played with it enough to figure out it's use and limitations, which is my next point...I own a number of tek 492's which is the 21 ghz version of the 496, which tops out at 1.8 gig (anyone ever notice in the old days all teks and hp's either topped out at 1.8 or 18 gig, then came 21 gig, and now up to 26 before you have to go to external mixers? think it was sales dep't wars) ok, im DRIFTING off the subject, but not really....I like that vintage of tek's (cheap to fix) they work well
EXcept the frequency accuracy is abysmal in those things....in the old days
none of the analyzers were very accurate, (freq wise)they were'nt supposed to be- their main pupose in life was to compare amplitudes relative to other signals and maybe checking bandwidth now and then (the spans seemed to be relatively accurate) ..thats why there is that cal button next to the tuning dial!...notice you can only adjust it in 1 meg steps...and that's wishful thinking...SO..I guess what I'm trying to say is looks like you have a few things working against you from the get-go...if you have a service monitor you can iso-tee it into the spec. analyzer input and use it as a marker which works well...it all depends how much complications you're willing to put up with...
EXcept the frequency accuracy is abysmal in those things....in the old days
none of the analyzers were very accurate, (freq wise)they were'nt supposed to be- their main pupose in life was to compare amplitudes relative to other signals and maybe checking bandwidth now and then (the spans seemed to be relatively accurate) ..thats why there is that cal button next to the tuning dial!...notice you can only adjust it in 1 meg steps...and that's wishful thinking...SO..I guess what I'm trying to say is looks like you have a few things working against you from the get-go...if you have a service monitor you can iso-tee it into the spec. analyzer input and use it as a marker which works well...it all depends how much complications you're willing to put up with...
"thats why there is that cal button next to the tuning dial!..."
----------------------------------------------------------------------
That is also why there is a counter output on the front of the tracking generator. The app notes and the manuals explain this, but basically you connect a counter, then you put the analyzer in manual sweep. Now lower the peak/average line to where it intersects with the point of interest on your trace. Then turn the manual sweep control - a dot will move across the peak/average line. When the dot reaches the desired intersecting point - bingo - the frequency counter reads out the exact frequency - more or less a tuneable marker!
This works whether you are viewing a tracking generator trace or one 'off the air'. The ideal counter for this application is a DC508 - it goes to a gig, has a 9 digit readout, and is only one TM slot wide. You put this and your TG in a TM503 and you have a small accessory package that greatly enhances your analyzer.
Also the 494 series are coming down in price. Same form factor and user interface but have stored traces and setups, synthesized frequency accuracy, etc. The book says accurate to 41 Hz at 4 GHz after a 30 minute warm up.
----------------------------------------------------------------------
That is also why there is a counter output on the front of the tracking generator. The app notes and the manuals explain this, but basically you connect a counter, then you put the analyzer in manual sweep. Now lower the peak/average line to where it intersects with the point of interest on your trace. Then turn the manual sweep control - a dot will move across the peak/average line. When the dot reaches the desired intersecting point - bingo - the frequency counter reads out the exact frequency - more or less a tuneable marker!
This works whether you are viewing a tracking generator trace or one 'off the air'. The ideal counter for this application is a DC508 - it goes to a gig, has a 9 digit readout, and is only one TM slot wide. You put this and your TG in a TM503 and you have a small accessory package that greatly enhances your analyzer.
Also the 494 series are coming down in price. Same form factor and user interface but have stored traces and setups, synthesized frequency accuracy, etc. The book says accurate to 41 Hz at 4 GHz after a 30 minute warm up.
Here is a link to some duplexer tuning instructions. These are for certain WACOM models but the concepts are applicable to many others. This is on Kevin Kuster's Repeater Builder site:
http://www.repeater-builder.com/pdf/wp6 ... ctions.pdf
http://www.repeater-builder.com/pdf/wp6 ... ctions.pdf
My experiences on UHF
I've tuned a couple of Celwave pass/reject 6-cavity UHF duplexers (526-4 if I recall). I use an HP spectrum analyzer with a built-in tracking generator and a return loss bridge. Here are the steps I've used.
Disconnect/remove the coax jumpers between the cavities.
With the spectrum analyzer, tune each of the TX bandpass filters to pass the TX frequency, and tune the TX reject to null out the RX frequency. I usually get about 35-40 dB of notch this way.
Similarly tune the RX bandpass filters to pass the RX frequency and tune the RX reject to null out the TX frequency.
Connect all the sections together with the supplied coax jumpers.
Connect 50 ohm loads to the Antenna and RX ports. Connect the return loss bridge to the TX port. Adjust the bandpass tuning of the cavity nearest the TX port for least return signal, i.e. a deep null.
Swap the TX and RX connections and tune the passband of the cavity nearest the RX port for the least return signal.
Disconnect the cavities and run through the reject tuning again of the cavities closest the TX and RX ports.
The spectrum analyzer is great for tuning the reject (null) because it shows a very deep notch. It's not too good for getting the broad bandpass centered. The return loss bridge gives a nice deep null when the VSWR is best, i.e. the bandpass tuning is right on the money. You could go crazy and adjust the bandpass tuning of ALL the cavities with the return loss bridge, but in practice only the end cavities really need to be done. The spectrum analyzer is close enough on the intermediate cavities.
I don't worry about tuning it on site. I also gave up trying to calculate or measure the optimum coax length between the radios and duplexer. If everything is a true 50 ohms, the coax should be invisible and length should have no effect. Now the purists will probably tell me that it should be an odd multiple of 1/4 wave, but so far I've had no problem with random lengths under 2 feet.
A reject-only mobile duplexer is easily tuned with just a spectrum analyzer. Terminate the unused port and tune for deepest notch at the rejection frequency.
Bob M.
Disconnect/remove the coax jumpers between the cavities.
With the spectrum analyzer, tune each of the TX bandpass filters to pass the TX frequency, and tune the TX reject to null out the RX frequency. I usually get about 35-40 dB of notch this way.
Similarly tune the RX bandpass filters to pass the RX frequency and tune the RX reject to null out the TX frequency.
Connect all the sections together with the supplied coax jumpers.
Connect 50 ohm loads to the Antenna and RX ports. Connect the return loss bridge to the TX port. Adjust the bandpass tuning of the cavity nearest the TX port for least return signal, i.e. a deep null.
Swap the TX and RX connections and tune the passband of the cavity nearest the RX port for the least return signal.
Disconnect the cavities and run through the reject tuning again of the cavities closest the TX and RX ports.
The spectrum analyzer is great for tuning the reject (null) because it shows a very deep notch. It's not too good for getting the broad bandpass centered. The return loss bridge gives a nice deep null when the VSWR is best, i.e. the bandpass tuning is right on the money. You could go crazy and adjust the bandpass tuning of ALL the cavities with the return loss bridge, but in practice only the end cavities really need to be done. The spectrum analyzer is close enough on the intermediate cavities.
I don't worry about tuning it on site. I also gave up trying to calculate or measure the optimum coax length between the radios and duplexer. If everything is a true 50 ohms, the coax should be invisible and length should have no effect. Now the purists will probably tell me that it should be an odd multiple of 1/4 wave, but so far I've had no problem with random lengths under 2 feet.
A reject-only mobile duplexer is easily tuned with just a spectrum analyzer. Terminate the unused port and tune for deepest notch at the rejection frequency.
Bob M.