System return loss performance

[dthayden]

Hi all, I have a work related question I am hoping someone can help with. Andrew, a major antenna and waveguide manufacturer offers a system design guide to design antenna/waveguide systems with guaranteed return loss performance. The method, on page 228 of the catalog here (warning ~36MB file) : http://plymouthcolony.net/starcityeng/files/AndrewCatalog38.pdf

The method generally makes sense in that the return loss, or VSWR, performance of each element in a transmission system additively degrades the system return loss. I am wondering if anyone has practical experience with this or has knowledge of any software or online tools which performs a similar function?

At some of our radio sites we are measuring a system return loss of 14dB with as few as 5 rigid waveguide sections between our radio and elliptical waveguide which terminates in a high quality antenna. At other sites we have 6 to 7 sections of rigid and the system return loss measures higher than 20dB. Per the Andrews design guide, using only 5 sections of rigid waveguide with a VSWR spec of 1.05, it could be this bad. It could be better, but the guaranteed number is 14dB. Switching to narrowband rigid waveguide with VSWR spec of 1.02 could solve this, if Andrews is correct. However, coworkers are reluctant to believe the results and I am looking for other authoritative sources to collaborate (or debunk) what we are seeing.


Thanks,

Todd

 

[zz0468]

...The method generally makes sense in that the return loss, or VSWR, performance of each element in a transmission system additively degrades the system return loss. I am wondering if anyone has practical experience with this or has knowledge of any software or online tools which performs a similar function?

I am quite familiar with this. I don't have any fantastic software tools to share, but the methodology Andrew uses in their catalog is simple enough. It'll come up with a worst case number to work with. You should be able to meet, or exceed that.

At some of our radio sites we are measuring a system return loss of 14dB with as few as 5 rigid waveguide sections between our radio and elliptical waveguide which terminates in a high quality antenna.

How are you measuring this? I will assume, for now, that you're making swept measurements with a VNA. Has the entire test setup been properly calibrated prior to making the measurements? Calibrating into coax, and then putting a transition on to measure the waveguide will cause some HUGE measurement errors, but I see people doing it all the time.

Assuming you can plot the swept return loss measurements, what does the plot look like? Your sweep should be wider than the full waveguide bandwidth, so you can see if there are any anomalies within the waveguide. Generally, when I've seen these sorts of problems, it shows up with a wide sweep. You should be able to observe the upper and lower cutoff frequencies in the waveguide, with no odd modes, notches, or peaks in between. Disregard a nominal amount of ripple in the bottom of the passband.

At other sites we have 6 to 7 sections of rigid and the system return loss measures higher than 20dB. Per the Andrews design guide, using only 5 sections of rigid waveguide with a VSWR spec of 1.05, it could be this bad.

"This bad" being higher than 20 db? Yes, possibly.

It could be better, but the guaranteed number is 14dB.

I'd be interested in seeing the math. Post the RL on the waveguide sections, along with the reflection coefficients that you're using to calculate. If the calculations are matching the measurements, well, there you go...

But that seems pretty awful to me. I've never seen a waveguide run that bad, unless it was damaged. More on this in a minute...

Switching to narrowband rigid waveguide with VSWR spec of 1.02 could solve this, if Andrews is correct. However, coworkers are reluctant to believe the results and I am looking for other authoritative sources to collaborate (or debunk) what we are seeing.

I'm a bit reluctant to believe it as well. A RL of 20 db is a VSWR of 1.22, and a reflection coefficient of 0.1 . I can imagine a bunch of waveguide bends and twists doing that. Let's see the math from the calculations.

I mentioned damage. Assuming that the measurements don't meet or exceed the calculations, you might look for damage to the waveguide. The SLIGHTEST little dimple on the inner wall will cause reflections. That would show up on your RL sweep. Sweep each individual piece of waveguide and see if there's one piece causing a problem. Check the RL right at the end of the elliptical and see how that looks. The problem could be further up. Are you using tuned connectors on the elliptical? If so, have you tuned them?

If you're too close in frequency to either cutoff or a higher mode, adding bends and twists could cause the RL to go out of line.

The last thing I would comment on is why are you using so many rigid sections in the installations? These days, a typical waveguide installation is elliptical from the antenna to the top of the stack, and a relatively short piece of flexiguide to the radio. Bends are smooth, and twists are minimized by planning the installation so that the waveguide comes off the tower oriented (E plane or H plane) so that no actual twists need to take place.

 

[dthayden]

I am quite familiar with this. I don't have any fantastic software tools to share, but the methodology Andrew uses in their catalog is simple enough. It'll come up with a worst case number to work with. You should be able to meet, or exceed that. .

That methodology is what is in dispute. I was hoping for another authoritative source to back it up. I think the fact that some sites measure fine and some don't is the root of the suspicion. They don't seem to understand the meaning of "worst case" performance, or just do not want to believe the results.

"This bad" being higher than 20 db? Yes, possibly.

"bad" is 14dB..

But that seems pretty awful to me. I've never seen a waveguide run that bad, unless it was damaged. More on this in a minute...

We were surprised as well, until we found the Andrew design guide which backs up the result. I think the fact that some sites measure fine and some don't is the root of the suspicion. They don't seem to understand the meaning of "worst case".

If you're too close in frequency to either cutoff or a higher mode, adding bends and twists could cause the RL to go out of line.

Can you elaborate on this?


Thanks for the response. We are using non-tunable elliptical connectors which measure great (see below). We are calibrating out the waveguide transitions during measurements. The tester (Anritsu 820D) we are using has measurement uncertainty on the order of 3dB based on the directivity of the coupler in the instrument.

The numbers I used in the Andrew calculation:

Antenna vswr 1.08 (.038)
Elliptical vswr 1.06 (.0295)
Presssure Window vswr: 1.01 (.005)
Rigid Power sampler vswr: 1.10 (.047)
5 Sections Rigid vswr: 1.06 each (.0295)
Twist Flex vswr: 1.06 (.0295)


Measuring at the elliptical connector furthest from the antenna, we get 26dB over 5.9-7.1Ghz swept range.
Measuing at the rigid waveguide furthest from the antenna, we get as low as 14dB over 5.9-7.1Ghz swept range.
For our measurements we replace the antenna with a 50 ohm termination (vswr 1.10).
We did not sweep outside the rated bandwidth of the waveguide.

According to the Andrew design guide, switching our rigid sections to 1.02 VSWR (i.e. narrowband components) should provide a system worst case of 18dB return loss.

There are various reasons that drove us to using the rigid waveguide. The system degredation was unexpected and we are trying to ensure we understand why now, before spending more money to address the issue. Thanks again for any ideas.

Todd

 

[mikewazowski]

We typically bring our elliptical just inside the shelter where we transition to rigid waveguide right to the radio. Flex is rarely used and only a very short length is permitted where the rigid can't be aligned directly with the radio due to obstacles.

It's been years since I've swept an elliptical line. Don't have many problems with it and its not that often that we don't see the radio tower mounted.

I'd go over each piece and look for any dents, out of square areas or compressed areas. Can you sweep the individual components to see if one stands out from the rest?

 

[prcguy]

I have not used Eliptiguide in years but always had tunable connectors. We have hundreds of feet of C through Ka rigid guide and everything is swept and dimple tuned on site. When working with rigid and eliptical in the same system we tuned each section separately then measured the entire run. 14dB RL is not very good, especially in a high power system. We also use very high directivity broad wall couplers for measurements.
prcguy

 

[zz0468]

That methodology is what is in dispute. I was hoping for another authoritative source to back it up.

I've seen that formula for adding reflection coefficients and converting to return loss before, outside the Andrew catalog. It shows up in some application notes from HP and, of all places, Aeroflex. But I don't have time to search and dig up links.

I think the fact that some sites measure fine and some don't is the root of the suspicion. They don't seem to understand the meaning of "worst case" performance, or just do not want to believe the results.

The sites that measure ok... are they all identical?

"bad" is 14dB..

You left out the insertion loss of the waveguide feedline, but I made an educated guess, then ran the numbers, per the formula on page 228 of the Andrew Catalog. The number I arrived at is... wait for it... 14 db.

We were surprised as well, until we found the Andrew design guide which backs up the result. I think the fact that some sites measure fine and some don't is the root of the suspicion. They don't seem to understand the meaning of "worst case".

What you need to do is convert the formula to a spread sheet, and run some "what if" scenarios. Remove a couple of sections of rigid, or the sampling port. That right there is a sizable hit to RL. See what you can do to redesign the waveguide system to reduce the number of bumps and mismatches. More on that later... *

Can you elaborate on this?

You mention that you swept 5.9-7.1 GHz, but you didn't specify your actual operating frequency, or the specific type of waveguide. So, I'll assume you're using EW63 eliptical, and some rigid equivalent. If your operating frequency is actually 5.8 or 5.9, for example, and you're close to cutoff, any bends or joints or twists can, for a short distance, increase the cutoff frequency to something above your operating frequency. You'd see that as a lower RL, and additional insertion loss. The same can happen if you're too close to the upper frequency limit. Discontinuities inside the waveguide can excite higher order modes, and have the same end result.

So, make sure that frequency you're operating at is as close to the center of the waveguide's frequency range as you can get.

Thanks for the response. We are using non-tunable elliptical connectors which measure great (see below). We are calibrating out the waveguide transitions during measurements. The tester (Anritsu 820D) we are using has measurement uncertainty on the order of 3dB based on the directivity of the coupler in the instrument.

Sounds like you're doing everything right, although my first recommendation would be to only used tuned connectors.

According to the Andrew design guide, switching our rigid sections to 1.02 VSWR (i.e. narrowband components) should provide a system worst case of 18dB return loss.

That's still pretty dismal for a waveguide system. I guess the system engineer should be able to specify an acceptable figure. My worst case spec was always 23 db RL for a 30 MHz bw digital radio. Analog was even more stringent. Multiline routes with 1:n protection need a much better RL than that.

There are various reasons that drove us to using the rigid waveguide.

For one, it takes up less space inside the shelter than elliptical, and you can use tighter bends. What they don't tell you in the catalog is that those tighter bends cost you in terms RL. So do all the flanges bolting those rigid section together.

The system degredation was unexpected and we are trying to ensure we understand why now, before spending more money to address the issue. Thanks again for any ideas.

I'm pretty confident that, by now, we have identified the cause of the degradation. It's all those little things adding up.

* The "more" from above...

What to do? You have 5 sections of rigid bolted together, along with a sample port, a pressure window, and a piece of flexiguide. Every flange bolted together is a major cause of reflections inside the waveguide. It wants to be glassy smooth, and every joint is a physical bump. You have two choices... reduce the bumps, or tune them out. Reducing the bumps would entail doing something other than bolting 5 chunks of waveguide together. One long one? Are there any bends?

You could try loosening the joints one at a time, and aligning them, and retightening, and see of you can tune the bumps out that way. Or, you could put a tuning section (like in the tuned connectors) and tune the bumps out that way. See Andrew or Pasternack for waveguide matching sections. Or... got a drill press? Make your own. A 3 or 5 screw tuner is easy to make. Threaded bolt holes in the center of the waveguide spaced 1/4 guide wavelength apart. Tune with a network analyzer until you have that eureka moment when all the bumps go away, and your return loss hits 30 db. They make those tuned elliptical waveguide connectors for a reason!

More than anything, though, make the formula into a spread sheet, and start playing with it and see how all those little things add up.

 

[zz0468]

I have not used Eliptiguide in years but always had tunable connectors. We have hundreds of feet of C through Ka rigid guide and everything is swept and dimple tuned on site. When working with rigid and eliptical in the same system we tuned each section separately then measured the entire run. 14dB RL is not very good, especially in a high power system. We also use very high directivity broad wall couplers for measurements.
prcguy

Do you use any special alignment tools when bolting together the wg sections? AT&T techs used a modified vicegrip that held the sections together, precisely aligned until the bolts were all tightened. They did that to avoid the same problems that the OP is discussing.

 

[dthayden]

We typically bring our elliptical just inside the shelter where we transition to rigid waveguide right to the radio. Flex is rarely used and only a very short length is permitted where the rigid can't be aligned directly with the radio due to obstacles.

This is exactly what we had planned for our new installations. On on retrofit installations, the elliptical typically extends quite a ways into the shelter. The plan there was to build rigid connections from the existing elliptical connector to the new radio rack flange.

On our new installs done to date, some sites measure well and some bad. I became concerned when it looked like our planned retrofit sites were going to require 8-12 rigid sections to plump from the radio flange to the existing elliptical connector. Getting up to 3 waveguides (TX and RX different polarity, plus diversity RX) up and aligned with 3 existing elliptical runs can require quite a few sections if constructed using rigid.

 

[prcguy]

No, just the typical SS captive screws and there seems to be no electrically measurable alignment problem. Some of our systems are also tested up to 5KW at Ku band and 6KW at C band.
prcguy

Do you use any special alignment tools when bolting together the wg sections? AT&T techs used a modified vicegrip that held the sections together, precisely aligned until the bolts were all tightened. They did that to avoid the same problems that the OP is discussing.

 

[dthayden]

I've seen that formula for adding reflection coefficients and converting to return loss before, outside the Andrew catalog. It shows up in some application notes from HP and, of all places, Aeroflex. But I don't have time to search and dig up links.

...

That's still pretty dismal for a waveguide system. I guess the system engineer should be able to specify an acceptable figure.

...

I'm pretty confident that, by now, we have identified the cause of the degradation. It's all those little things adding up.

I have searched HP/Agilent and Cobham (Continental Microwave) for suporting docs with no luck. I will search Aeroflex also.

Our radio supplier has specified 15dB minimum.

So you have come to the same conclusion I have.

You could try loosening the joints one at a time, and aligning them, and retightening, and see of you can tune the bumps out that way. .

We are using EW63 elliptical and WR137 rigid, operating at lower and upper 6GHz frequencies, 30MHz bandwidth packet radios. No two sites are the same. As I posted above, we have new and retrofit sites with 1 to 3 waveguides per radio, 1 to 3 radios per site, each shelter is unique, radio positions unique, cable trays run differently, etc. The rigid runs are "designed on site" from a collection of purchased sections.

I believe our connectors are fine given the results of the elliptical sweeps. We used to used tunable connectors but our preferred supplier no longer offers them.

A coworker is heading out to one of the poor measuring sites tomorrow to look further, including experimenting with flange alignment. Another coworker found reference to wavguide alighment pins, used primarily at high freqs and with waveguide from multiple suppliers. No one here has experience with dent tuning ... we would end up with a pile of scape bronze in no time.

Thanks again for your inputs.

 

[zz0468]

Getting up to 3 waveguides (TX and RX different polarity, plus diversity RX) up and aligned with 3 existing elliptical runs can require quite a few sections if constructed using rigid.

That definitely aggravates the problem of running elliptical straight to the top of the radio. I've had to deal with as many as 4 waveguides going to a single radio, and rigid would have made the job cleaner looking, but in the end, it was cost and ease of installation that won out. I tend to think that performance won out, as well.

I have searched HP/Agilent and Cobham (Continental Microwave) for suporting docs with no luck. I will search Aeroflex also.

The Andrew catalog has always been an authoritative source, as needed when it comes to system planning. It's always nice to find another source, but in this case, I wouldn't waste too much time on it.

Our radio supplier has specified 15dB minimum.

What's the manufacturer?

So you have come to the same conclusion I have.

Yes, and I think it's a pretty solid conclusion. When the calculations match so closely with measured reality, it's time to stop looking at why it's happening, and time to start looking at what you can do about it.

I will admit that I'm surprised an how poorly those 5 waveguide sections and a flexiguide are performing, but I have to admit, I've never bolted so many sections together without something to tune out the bumps.

We are using EW63 elliptical and WR137 rigid, operating at lower and upper 6GHz frequencies, 30MHz bandwidth packet radios. No two sites are the same. As I posted above, we have new and retrofit sites with 1 to 3 waveguides per radio, 1 to 3 radios per site, each shelter is unique, radio positions unique, cable trays run differently, etc.

So, that would make me ask if you can correlate return loss directly with the number of rigid sections and bends, etc. That might be a useful piece of information, but then again, Murphy likes to stick his nose into places like this, and might make the most complex site beat the calculated RL by 10 db, just because he can. =)

The rigid runs are "designed on site" from a collection of purchased sections.

Would it then be acceptable to make a tuner section out of one or more of the waveguide sections? It's not hard to do, and not hard to tune up. It sounds as if you have the right tools and skills. You could try purchasing a tuning section, but I'm not sure who makes one without it being a custom assembly.

I believe our connectors are fine given the results of the elliptical sweeps. We used to used tunable connectors but our preferred supplier no longer offers them.

They're usually ok, and when you measure a 26 db RL at an untuned connector, there's little compelling reason to do anything about it. Except in your case, where the issue is the little things adding up. That's just one more little thing that you can subtract. If you can get to where you need by doing something else, then by all means.

A coworker is heading out to one of the poor measuring sites tomorrow to look further, including experimenting with flange alignment.

Another thing to look for, make sure you're not trying to mate any choke flanges together. A choke flange MUST mate with a flat flange. Two flat flanges can mate, but you want an RF gasket. Two choke flanges together can put a significant bump on the line. If you have to do this, you could put a pressure window in, or something like that, to give each choke flange a flat plate to press on.

Another coworker found reference to wavguide alighment pins, used primarily at high freqs and with waveguide from multiple suppliers. No one here has experience with dent tuning ... we would end up with a pile of scape bronze in no time.

The alignment pins can be helpful, but they might be hard to find. I have some, but I'd be hard pressed to get any more. They come with the waveguide calibration kits that go with my network analyzer.

 

[zz0468]

No, just the typical SS captive screws and there seems to be no electrically measurable alignment problem. Some of our systems are also tested up to 5KW at Ku band and 6KW at C band.
prcguy

Some types of flanges seem to line up better than others. They're all pretty close, but some time if you're able, hook some up to a VNA, loosen the bolts a bit and slop them around while you watch the VNA. It can get pretty ugly.

 

[prcguy]

When I used to do lab work in the millimeter range of 40-90GHz all the flanges had alignment pins and very precise tolerances.
prcguy

Some types of flanges seem to line up better than others. They're all pretty close, but some time if you're able, hook some up to a VNA, loosen the bolts a bit and slop them around while you watch the VNA. It can get pretty ugly.

 

[zz0468]

When I used to do lab work in the millimeter range of 40-90GHz all the flanges had alignment pins and very precise tolerances.
prcguy

Yes! The components for those frequencies are very precisely made. The flanges have built-in alignment pins, and there's virtually no slop.

The problem with wg flanges in the 6 GHz range is, they depend on the bolts themselves for alignment, and there necessarily needs to be a bit of slop there in order to pass the bolt through one flange to thread into another. It's especially problematic when neither flange is threaded, and a nut is required.

The alignment is pretty good, but when you sit there with a network analyzer watching RL in real time, you'll see it go from 1.02 to 1.1 as you wiggle it around. Now bolt 5 of those pieces together and see what happens.

 

[prcguy]

Standard practice here for C through Ku band is bolt them together and measure in logical sections then dimple tune using some modified channel lock pliers with BBs welded to the jaws.
prcguy

Yes! The components for those frequencies are very precisely made. The flanges have built-in alignment pins, and there's virtually no slop.

The problem with wg flanges in the 6 GHz range is, they depend on the bolts themselves for alignment, and there necessarily needs to be a bit of slop there in order to pass the bolt through one flange to thread into another. It's especially problematic when neither flange is threaded, and a nut is required.

The alignment is pretty good, but when you sit there with a network analyzer watching RL in real time, you'll see it go from 1.02 to 1.1 as you wiggle it around. Now bolt 5 of those pieces together and see what happens.

 

[dthayden]

... you could put a tuning section (like in the tuned connectors) and tune the bumps out that way. See Andrew or Pasternack for waveguide matching sections. Or... got a drill press? Make your own. A 3 or 5 screw tuner is easy to make. Threaded bolt holes in the center of the waveguide spaced 1/4 guide wavelength apart. Tune with a network analyzer until you have that eureka moment when all the bumps go away, and your return loss hits 30 db. They make those tuned elliptical waveguide connectors for a reason! .

We may try this, I have a spare section of waveguide we can sacrifice. Does the 1/4 wavelength apply to the midpoint of the RX and TX frequencies at the site? Would be still be looking for high return loss across a wide band (5.8 - 7.2GHz) or just at our operating frequencies?

 

[zz0468]

Just shoot for something in the middle. Since you'd probably need a micrometer and a magnifying glass to see the difference on wavelength between the TX and RX frequencies, that aspect of it isn't critical. Placement of the screws in the overall scheme of things can be quite critical, so if you put maybe 5 screws 1/4 wavelength apart, you should be able to find some combination that changes the wg impedance in the right place at the right amount. Attempting to duplicate the configuration on the tuned EW connectors wouldn't be a bad place to start if you've never doen this before.

Once you've measured, drill and tap the holes in the center of waveguide. I've found it doesn't really matter what the screw size is. But for wr137, 10-32 would work. Use brass screws. When you tune it, just fiddle with whatever individual screws seem to make it better. You'll probably find some screws do nothing, or can only make it worse. Don't forget to factor in the wg velocity factor when calculating 1/4 wave.

In the end, you will end up with the functional equivalent of a tuned connector, with the ability to shape the RL response. The reflections are still there inside the wavegude, unless you tune each piece, but the radio won't know it. Unless you're trying to run 18,000 analog mux channels, no one else will, either.

 

[beachmark]

This is a week old, but I'll back up the results that you are learning with the Andrew info. The methodology is simple and sound, and is computing worst case RL, when reflections from each individual junction add up in phase all along the lines.

If you are doing a very narrow band system, this will usually result in a large variation of measured results from system to system, as the phasing will be randomly spread for all the junctions for that frequency. For a wide band system, when you sweep across a very wide set of frequencies, then you will get a large variation in how the reflections phase together across all the frequencies for that individual system, and you will invariably see the worst case phasing combination at one or a few frequencies in that wide band.

You will always tend to have these problems with non-rigid waveguides in wide band sweeps. You can align the flanges down to the micrometer but that will help only so much. The remaining problem that you can't solve in the field is in the way the flex guides are cut and terminated in the flanges and the basic non-uniform cross-section of flex guide itself; this will always be somewhat 'messy' and irregular compared to a clean straight end cuts and cross-section of rigid waveguide. Think of elliptic WG as being like a randomly crushed coax cable with crappy end cuts; we all know that will end up with poor VSWR/High RL. The basic spec of ER63 elliptic waveguide in the Andrew catalog is 1.15 (23 dB RL) BEFORE you even put on the flanges, which is really quite poor for any sort of unterminated transmission line. The basic spec for a WR137 90 degree H-plane elbow WITH flanges is 1.02 or 40 dB RL. So you are starting in a hole already with elliptic WG.

If you are field terminating the flanges onto the any of the waveguides, that is another big source of problems. Bottom line: don't be surprised if these poor RL's are showing up and concentrate on the fixes.....which I think has been said. You became committed when the elliptic was used in such proliferation.

 

[dthayden]

This is a week old, but I'll back up the results that you are learning with the Andrew info. The methodology is simple and sound, and is computing worst case RL, when reflections from each individual junction add up in phase all along the lines.

I appreciate you chiming in, the whole purpose of the post was to gether info from people more experienced in these matters than we are. Unfortunately there are still folks here who disbelieve the results and the methodology. Their attampt at a solution is to remove components in the path and disregard the measurement uncertainty of the test instrument until they see the number they want to see. We have a couple dozen sites we are upgrading and the measured results will speak for themselves. I was trying to head off a possible schedule impacting issue, but working for a monopolistic power utility there is really no need. My background is R&D of products in manufacturing environments in competitive markets. I am finding my old habits die hard.

A better question at this point is what are the possibly implications of not meeting the RL spec of the radio? We have primarily 2W dual channel hot-standby radios operating up to 256QAM, whose waveguide stacking filters employ circulators to cascade the four filters. The radio manufacurer has specifed 15dB RL maximum at the waveguide flange. I can envision possibly detuning the filters and degrading transmission filter skirts. Any radio performance degredation (decreases sensitivity, higher BER, reduced fade margin, etc) to worry about?

If you are doing a very narrow band system, this will usually result in a large variation of measured results from system to system, as the phasing will be randomly spread for all the junctions for that frequency. For a wide band system, when you sweep across a very wide set of frequencies, then you will get a large variation in how the reflections phase together across all the frequencies for that individual system, and you will invariably see the worst case phasing combination at one or a few frequencies in that wide band.

Agreed, and because of that, the point at which you measure, and the order of the waveguide sections matter. Folks here either don't believe that or don't want to believe that.

The basic spec of ER63 elliptic waveguide in the Andrew catalog is 1.15 (23 dB RL) BEFORE you even put on the flanges, which is really quite poor for any sort of unterminated transmission line. The basic spec for a WR137 90 degree H-plane elbow WITH flanges is 1.02 or 40 dB RL. So you are starting in a hole already with elliptic WG.

Our wideband sweeps of the connectorized elliptical look good, 26dB or better. We have speficied EWP63-59W, the P being the screened/tested waveguide with better VSWR specs. The point of having our contractor measure the installation is to guarantee good field termination and waveguide installation. The rigid waveguide and number of sections seems to be our issue.

Thanks again. Todd

 

[beachmark]

Hey Todd,

Well, it does not matter if the others beleive it or not; none of us can refute the laws of physics and that is precisely what you have there LOL! OK on the elliptial section RL; I thought you we using a lot of them; sounds like not.

As for the mfr spec VSWR/RL, that sounds like a pretty general spec that one would see on a lot of radio equipment. This spec could be set for a number of reasons, but is probably set for the fact that you will typically achieve only that level consistently in fielded installations. So that gets adopted as the standard to which the mfr tests their transmitters and receivers in the lab. Probably bad effects of poor VSWR would be transmitter final amplifier instabiliy, and the TX noise floor due to spectrum spreading, and other similar effects on the transmitters which are going to be the most finicky about poor RL.

If the system is good with 15 dB RL, then 14 dB should be OK 99.9% of the time. 256 QAM can be pretty susceptible to amplitude variations, as it does not take much to move a transmitted bit's phase-amplitide point towards an adjacent phase-amplitude point in the receive constellation. But the forward transmission difference between a 15 and 14 dB RL is around 0.8% which would be approx 0.4% change in signal voltage (which gets doubled back to 0.8% worst case if the same variation is seen at the RX end) so that should not be of any significance to the QAM reception. The in-band variations from 14 or 15 dB up to maybe 25-30 dB across the spectrum used should be a lot worse than 14 vs 15 dB.

One issue of real concern IMO becomes in maintenance, in case the waveguide needs to be disconnected for any reason. If this is at midnight to 4 AM on some stormy night to get the system back up and running, and if the waveguides have to be precisely aligned now to get to a decent RL, what do you think will happen 'in the heat of battle' so to speak? No way you will keep good RL results in the long term. The system VSWR will tend to degrade everytime someone touches the WG and a few years down the road, the same guys who don't believe the VSWR degradation due to a lots of cascaded WG sections will be saying that the system is a POC.

So, IMO, the real concern (if you are trying to convince someone) is long term; it becomes a contest between 'get it done and make it pass' for installation acceptance, versus a long robust maintenance situation. I hope I am not stepping on any toes to say this, but, this is a design aspect of the project that perhaps needed to be understood and planned out differently before using so many sections; I recognize that (without being there to see the sites) that it might not have been avoidable. And it also depends on the skill sets and experience of the maintenance personnel; there are a few guys posting here who obviously understand it and can deal with it. But these types are becoming fewer and fewer, or so it seems.