Help needed with transmitter combining project.

[WPXS472]

I am not sure this is the correct place to post this. If not, I apologize.
Anyway, I have been asked to provide technical help in a project involving combining several VHF repeaters to use one transmit, and one receive antenna. This involves a 300 foot cellular type tower hosting mostly public safety and local government users. There are currently 45 antennas mounted on the tower, and everyone seems happy with their level of performance. There are others who want to move onto this tower, so they want to look at combining several repeater transmitters to use one antenna, with another antenna with 50 feet of vertical separation, through a multicoupler for receive. They got a quote from a local radio shop to do this, but when that shop said they would use LMR400 for all the jumpers, it was decided to not use that shop. They have some used, donated equipment they want to use. What I am looking for is documentation on the ground rules for doing this kind of thing. I have a radio background and the proper test equipment for doing this at my disposal. What I am looking for is something that will tell me how much loss I can expect given certain frequency spacings. I assume this will involve using two cavities and a dual circulator per transmitter. These will feed into a star combiner. I assume that the coax length going into the combiner is somewhat critical, but I don't know how to determine the correct length. So, can anyone point me toward some documentation that explains the basics of doing something like this?
Thanks

 

[iMONITOR]

a project involving combining several VHF repeaters to use one transmit, and one receive antenna. This involves a 300 foot cellular type tower hosting mostly public safety and local government users. There are currently 45 antennas mounted on the tower, and everyone seems happy with their level of performance. There are others who want to move onto this tower, so they want to look at combining several repeater transmitters to use one antenna, with another antenna with 50 feet of vertical separation, through a multicoupler for receive.

What a quagmire! No offense do you honestly feel a project of this magnitude should be trusted in the hands of anything less than experienced professionals that specialized in this business? Especially for public safety and local government use? I think you'd be opening yourself up to mega-liabilities if anything should go wrong under any circumstances. I think the local government is crazy to let anyone else handle it!

 

[fwradio]

iMONITOR is 100% correct. Even with a lot of experience designing antenna and filter systems I still send requests like this to the filter manufacturer. No one is going to home make a combiner that is going to work and be as reliable and efficient as a factory-made combiner. We use EMR for all of our projects. They are built in the USA and every system goes through an engineer for design and specifications. And VHF even makes the issue harder. More possibilities for intermod especially with the mixed TX and RX channels with no set spacing.

 

[RFI-EMI-GUY]

"There are currently 45 antennas mounted on the tower, and everyone seems happy with their level of performance. "

Boy; I would hate to step into "fixing" that! Let sleeping dogs lie!

There are several manufacturers of TX combiner and RX multicoupler equipment. TX-RX (Bird) and EMR Corp, Come to mind.

(Bird agreed to sell its Radio Infrastructure Products Division (RFIP), formerly known as TX RX Systems, to the Combilent Group. Combilent USA will operate under the TX RX Systems name from its U.S. headquarters in Angola, New York.)

You will need to create an inventory of all of the equipment on the site, transmitters and receivers as well as the associated frequencies and station model number. By all the equipment, you should include frequencies in UHF and VHF, VHF-Lo, 800/700, etc for purposes of running a site wide inter-modulation study. The equipment vendor can recommend the proper configuration to attain the best isolation and minimal loss. If you have proposed new stations, this is the time to identify them as well.


Caveats:
You might not be able to combine all of the VHF equipment into two, TX, RX antennas. If some of these are simplex stations, they may need their own antenna. Also VHF band was never ideal for duplex pairing (The FCC dropped the ball on this) .

Some of these 45 antennas might be licensed for a particular elevation. So if you have a master TX antenna, you may need to re-coordinate the TX for a station that is at a lower elevation.

Putting all your TX and RX each on one antenna will create a single point of failure should an antenna, line or multicoupler fail. Make a contingency or redundancy plan.

You are correct, the coaxial cable and antennas will have to be low passive intermodulation (PIM) rated or the interference potential may be worse than your 45 separate antennas experience.

 

[WPXS472]

When the tower was originally built and antennas were about to go up, they contacted TxRx systems for a quote. They quoted $150,000 for a very complex system of cavities, and even then, wanted to mute certain receivers if certain other transmitters were up. The fellow who had the tower built just used common sense in trying to isolate systems on the tower that might interfere. Things, as built, seem to work well. I think funding is an issue due to there being local governments involved. I will have a list of frequencies and one rack of equipment next week to look at. The people doing this don't really seem to see any difficulties. They mostly just want me to do the tune-up. If it doesn't work, then it doesn't work, and I think they will revert to the current set-up.

 

[prcguy]

The length of coax between the star combiner and cavity filter is critical and there are many variables to consider including how the loops in the cavity filters interact. Most commercially made units ordered from a factory have a guy with bins of jumpers of different lengths and they try several to get the best isolation. I've discussed this with several mfrs and there is no fool proof formula for determining the coax lengths using random cavities. Insertion loss for a cavity/isolator combiner can be reasonable and I'm guessing 1 to 1.5dB would be typical.

Using a star combiner is also for frequencies with wide spacing like a MHz or more at UHF where the cavity filters will start to give decent isolation with whatever skirt response they have. If the frequencies are closer than a certain amount, then a much more lossy hybrid combiner is used. These have a 3dB hybrid and usually dual isolators to get the needed isolation but you can combine two transmitters on the exact same frequency if needed. Problem is the loss can be very high and a two channel version is close to 4dB loss, a three channel about 6dB loss and a four channel is around 7dB loss. That's giving up a lot of power but if frequencies are too close for a cavity filter/isolater combiner then you have no choice.

You can have a hybrid system with a 3dB hybrid/isolator combiner used for some close spaced frequencies, then tap that into to a cavity filter/star combiner if all the frequencies can play together.

Your main tool for tuning a cavity/isloator type is a scaler network analyzer or spectrum analyzer with tracking generator. You tune for least amount of insertion loss, best return loss and highest isolation between channels, which is probably the most important spec. Getting the cable to its critical length is what gives you the needed isolation and low VSWR.

Its best to use RG-142B/U dual shielded coax with a known working length from another system, then you cut a jumper a little shorter than a known good one, install an SMA male connector, then use SMA "plug savers" to lengthen the cable about 1/2" at a time until it plays nice in the system. Once you find the right length you can then make a permanent cable the same length, as the SMA plug savers with Teflon dielectric have the same velocity factor as RG-142 coax.

Installing and tuning a factory ordered combiner system is one thing and fairly straight forward as its already been designed and cables are made. Starting from scratch with a bunch of cavities, isolators and a spool of coax is a first class ticket to the nut house, you will go crazy before getting it to work right unless you have some secret knowledge that few possess. I have the proper test equipment and a little knowledge and building one from scratch years ago almost put me in a mental ward.

 

[RFI-EMI-GUY]

When the tower was originally built and antennas were about to go up, they contacted TxRx systems for a quote. They quoted $150,000 for a very complex system of cavities, and even then, wanted to mute certain receivers if certain other transmitters were up. The fellow who had the tower built just used common sense in trying to isolate systems on the tower that might interfere. Things, as built, seem to work well. I think funding is an issue due to there being local governments involved. I will have a list of frequencies and one rack of equipment next week to look at. The people doing this don't really seem to see any difficulties. They mostly just want me to do the tune-up. If it doesn't work, then it doesn't work, and I think they will revert to the current set-up.

So you have two options:

1) Status Quo, 45 antennas, everybody is happy, but sorry, no more room at the inn! Let sleeping dogs lie!

2) A $150 - $200 K (assuming inflation and new stations) ballpark price to "do it right". Indeed it will be very complex. VHF is a mess to multiplex.

Before going to the exercise of getting a new quote, the powers to be should know that option 2 will require capital investment in that range.

 

[iMONITOR]

If it doesn't work, then it doesn't work, and I think they will revert to the current set-up.

"We don't mind paying for the failure"...said absolutely no one!
I wonder if there are any incentives or tax write-offs from the Federal Government for public safety improvements like this?

 

[RFI-EMI-GUY]

If it were me, and management were reluctant to do this correctly (IE: spend the BIG bucks), I would be checking my IRA, my Pension Benefits, Social Security etc, then duck out early on a high note. If you are a younger guy, there is always someone retiring in the comm bureau in the next county over!

 

[RFI-EMI-GUY]

There is something you can do as a DIY and that is to model the site you have. The software I would recommend is called Wireless Site - RFI WS-RFI Radio Frequency Interference Analysis Software. It will cost you several thousand dollars, will require you to carefully input many parameters about your stations, antennas, frequencies and existing filters. You will have to manually create filter bandpass profiles for any VHF filter that are not in the software filter library. It will be tedious. The result will be a computer simulation of the intermodulation and Transmitter noise receiver degradation (IM/TNRD) environment on your site. The isolation of the antenna seperation, isolators and filters will be used to determine if a transmitter(s) at a certain power level will interfere with receiver. The result will be the ability to simulate changing antennas positions or adding new stations and frequencies. It is a simulation, so your results are based on "GIGO" principal and may vary.

Riverview Software Solutions - Software Products

 

[WPXS472]

Thanks for the info "prcguy". I am already retired, and just do a little consulting on the side to supplement my Social Security. I don't know if there is a date when all this has to be done. Actually, I know very little except for the fact that they want to do it. If things don't want to work, I will just tell them I have done all that I can and they need to bite the bullet and hire a pro. I don't know what kind of pressure is being put on the guy that contacted me, but as far as I am concerned, I am more or less anonymous. I will be paid for my time at my going rate by him. I don't expect any actual contact with the folks with the radios. I expect that when I see the list of frequencies, there will be some that will be a no go from the start. I expect that some additional equipment will have to be purchased, as well as a lot of hard line jumpers of various lengths. He already has a UHF system, not at this tower, that combines 4 transmitters. It was purchased ready to go. He decided to change one transmit frequency and all I did was re-tune a couple of cavities. Too easy, I know.

 

[RFI-EMI-GUY]

Thanks for the info "prcguy". I am already retired, and just do a little consulting on the side to supplement my Social Security. I don't know if there is a date when all this has to be done. Actually, I know very little except for the fact that they want to do it. If things don't want to work, I will just tell them I have done all that I can and they need to bite the bullet and hire a pro. I don't know what kind of pressure is being put on the guy that contacted me, but as far as I am concerned, I am more or less anonymous. I will be paid for my time at my going rate by him. I don't expect any actual contact with the folks with the radios. I expect that when I see the list of frequencies, there will be some that will be a no go from the start. I expect that some additional equipment will have to be purchased, as well as a lot of hard line jumpers of various lengths. He already has a UHF system, not at this tower, that combines 4 transmitters. It was purchased ready to go. He decided to change one transmit frequency and all I did was re-tune a couple of cavities. Too easy, I know.

I have done quite a bit of consulting work since 1994, independently since 2010. At times I get referrals to work for a company that already has a prime contract with a client and wants me to pass through my hours via them (they mark up).

Most times it works out fine. I just closed out one where I was sub to the prime. However there was data presented to me that was lacking detail because those who did site surveys a year or two prior did not collect some data that was needed to conclude the report. I pointed this out immediately, and my concerns assuaged by the prime. But it became critical weeks later when the prime presented me with a new task which was no longer simply getting budgetary quotes but producing some more detailed design drawings. We worked through it in the end getting that data from another source. You can lose sleep over the details.

I passed on another project, ($35K) much like the one you are embarking on because I was asked to work through a sub to the prime (for billing purposes) and subsequently I was asked by the Prime to radically change the scope of how the work would be done (and to start work without a contract!) , even though the proper SOW had been presented to the end user. It was a high risk contract (airport radios). Basically the Prime wanted me to take shortcuts to accelerate the project . I saw risk to me and the end user and advised the Prime that unless I was working directly for the end user I was backing out. And I did.

I try to operate like the buck stops with me because in the end it will be my neck on the line if it all goes sideways. I don't assume somebody else will carry the water (without my paying) and I really don't like walking away from business, usually it comes to me as a referral. Sometimes you have to think ahead as to how it could go wrong.

It helps to watch Peoples Court at my lunchtime! A great lesson in contract law. Never become a "litigant".

 

[Project25_MASTR]

VHF is a pain. I'll build combiner systems for UHF and 800 MHz systems all day long from circulators and notch filters I have laying around but VHF is a whole different animal (I've done quite a bit of VHF trunking in Texas). The 5 and 8 channel cavity combiners (TXRX) I've worked with were were in the $15,000-$20,000 range. The hybrids where $7,000 per 4 channel brick. One the receive side, you basically have to notch out all of your transmitters which means you'll likely need a range of pre-selectors to pass 47 inputs but notch out the 47 outputs.

As you can see, you'll need to know every frequency in use by the system. You'll also need to take antenna limitations into play. Your average transmit combiner has 2 dB of loss per channel. You average hybrid combiner has 8 dB of loss per channel. What that translates to...100W into a cavity combiner will get you about 60W out (30W out for 50W in) or 16W out on a hybrid (8W if 50W is your input). Of course, we are negating cable losses at the moment. So even then, 60W out by 47 stations gives you 2.8 kW of combined output to the antenna...is that within the antenna power rating? Also keep in mind, a hybrid combiner can allow for adjacent channels (12.5 kHz of separation or even 6.25 kHz for ultra-narrow channels) where a cavity combiner needs 100-200 kHz of separation between transmitters typically.

Another thing to consider, I'm rather fond of Motorola's GTR8000 ESS setup. In an ideal world (VHF is not ideal) they have one receive antenna for the site, shared across up to five 6-channel racks. Each 6-channel rack has it's own 6-channel combiner and transmit antenna. While it seems like a lot, it adds a level of redundancy and provides a better maintenance experience. I say that because you don't have to take down an entire system to work on the transmit antenna system...only a single rack and you pickup some horizontal and/or vertical separation allowing you to space some transmitters more closely together (for example, you can fit 50 kHz separation or even 25 kHz separation on cavity combiners which would traditionally require much more separation).

 

[speedway_navigator]

I was at a combiner mfg shop last week, at lunch they were telling me about a client who had nothing but problems with a new VHF system the mfg had shipped. During the client's rant he happened to mention 'you guys waste a lot of coax, I cut all your jumpers down to 5" to make the combiner fit into the cabinet'.

 

[WPXS472]

That is one of the things that is worrying me about this. The existing jumper lengths might not work for the new frequencies. I am trying to choose frequencies that are close to the old ones in hopes that I can just tweak the tuning and it will work. TxRx used a combining method I have not seen before. They use the coupling loops on the cavities and just loop thru. It must have worked, because the rack came from a working system. So far, I have been looking at it trying to learn as much as possible before touching anything. I am supposed to go in tomorrow and begin the tuning process. It was used to combine 4 transmitters. There are 18 cavities, best I remember, with as many on receive as transmit. I believe only 1 was a pass/reject type.

 

[TampaTyron]

As stated above. Bring in a professional filter manufacturer for this. It is the most likely way to success. Taking it on yourself is inviting a bad monster into your life. TT

 

[prcguy]

The combined systems I've built and worked on always had a separate receive antenna with a receive band pass filter that had good rejection of all transmit freqs. I'm about to assemble a two channel hybrid combiner for two repeaters on the exact same frequency. One repeater is Yaesu Fusion and the other is P25 and of course only one can be used at a time otherwise they will interfere. The combiner will feed the transmit side of a PD524 duplexer with an additional cavity off the receive side feeding an Angle Linear preamp and a two way divider for the receivers.

 

[speedway_navigator]

This looks interesting, have not had a chance to dive into the tech specs.

Auto Tunable Combiner

sinctech-canada

www.sinctech.com

The new Sinclair INTELLITUNE™ Series of Auto Tunable Combiners features an integrated control unit for fully autonomous frequency tuning to keep return loss to a minimum and optimize overall performance. This allows the combiner to be used in unconventional ways to satisfy customers’ unique needs.

 

[RFI-EMI-GUY]

That is one of the things that is worrying me about this. The existing jumper lengths might not work for the new frequencies. I am trying to choose frequencies that are close to the old ones in hopes that I can just tweak the tuning and it will work. TxRx used a combining method I have not seen before. They use the coupling loops on the cavities and just loop thru. It must have worked, because the rack came from a working system. So far, I have been looking at it trying to learn as much as possible before touching anything. I am supposed to go in tomorrow and begin the tuning process. It was used to combine 4 transmitters. There are 18 cavities, best I remember, with as many on receive as transmit. I believe only 1 was a pass/reject type.

TX RX will sell you a set of new jumpers and instructions to go along with the new frequencies in an old "T PASS" system. However, you still may require new hardware, additional cavities, new or retuned isolators. This is not for the faint hearted..

 

[RFI-EMI-GUY]

This looks interesting, have not had a chance to dive into the tech specs.

Auto Tunable Combiner

sinctech-canada

www.sinctech.com

The new Sinclair INTELLITUNE™ Series of Auto Tunable Combiners features an integrated control unit for fully autonomous frequency tuning to keep return loss to a minimum and optimize overall performance. This allows the combiner to be used in unconventional ways to satisfy customers’ unique needs.

UHF only...