Back
System Engineering, and a few commments...
Compiled from discussions with several knowledgeable people and written, HTML'd and maintained by Mike Morris WA6ILQ.

Comments / critiques / suggestions on this page (actually any page at this web site) are welcome.

Rule number 1: The antenna system can make or break your repeater. If you are building a system on a budget, do the initial scrimping on the stuff inside the building. It's a lot easier and cheaper to update / upgrade later on. You need a high performing antenna system before you even power the repeater on.
a $1,000 repeater with a $30 antenna system is a $30 repeater system. See the paragraph below that starts with "Simply put..."

Rule number 2: This goes hand in hand with Rule number 1: Maximize your received signal. Always favor your system receiver and the antenna system. A system's coverage area is primarily determined by how well it receives stations in the field, not by how much transmitter power it has. You can always swap transmitters (or amplifiers) for more power, but if you can't hear the users the repeater is useless.

Rule number 3: Concentrate on minimizing the insertion loss on the receive side of the system. Test for Effective Sensitivity frequently (at least twice a year) - this includes tests with the transmitter on (but keep in mind the effective sensitivity can be a dynamic value unless the environment is a broadcast site with a constant RF environment).

Rule number 4: This goes hand in hand with Rule number 1: Maximize your received signal by maximizing the isolation from receiver to transmitter.

Rule number 5: Keep good records - a site logbook is important, ESPECIALLY if you can get the other site tenants to add entries to it. If something changes between visits you need to find out what changed and why. It may have been something that someone else did!
Example: At one multi-tenant commercial site the landlord (ATC) installed LED tubes in place of the old, tired and dead ceiling flourescent tubes. The noise floor on 220 MHz went up - but only when the lights were on! And the lights were only on while any tech was at the site! So there was an intermittent problem, but nobody had made any changes to any of the systems there! Switching back to flourescent tubes fixed the problem.

Rule number 6: With repeaters, the best you can afford may be barely good enough.

Rule number 7: You can't "balance" a system unless you can define your user demographics - i.e. who you are balancing it for. Are you catering to the base stations, to the mobiles, or to the handhelds?
A repeater system that has the transmitter power to reach out to the base stations and mobiles that have 60‑75 watts ERP could be accused of being an alligator system (all mouth and minimal ears) by a user with a 2‑watt handheld that has a ‑3 dB rubber duck antenna. And if he's inside a car driving down the road then there's an additional ‑3 to ‑9 dB due to the steel car body... and then you subtract the mobile flutter...
On the other hand, a system that is designed from the outset to support hand‑helds and has multiple voting receiver sites and a single centralized transmitter (like some paramedic systems) could be accused of being an elephant system (all ears and a small mouth).

Both the receiver and the transmitter depend on the antenna system - no matter if it is a duplexed single antenna or two separate antennas. Both the antenna(s) and the feedline(s) are the most critical components in any amateur radio station, be it a home station, a mobile, a repeater, a remote base or a point‑to‑point link. Perversely they are usually are the most difficult to install, maintain, or troubleshoot... and it seems like all of the troubleshooting usually has to be done during extreme climate conditions like in freezing temperatures and / or high winds while up on a tower. If you are installing a repeater that will be primary in any type of emergency services situations it does not make any sense to go "cheap" on the antenna system. Creative (legal) aquisition of good quality equipment, yes. Low or poor quality equipment, no... that decision will always come back and haunt you.

Simply put, a good quality professional land mobile grade antenna and feedline will hear and talk better than a poor antenna and do so for a lot longer time. If you go with quality hardware and proper installation the first time you won't have to do it over and over (like my late father used to say about both hand and power tools: "Buying quality only hurts once"). A good Phelps‑Dodge, DB Products or Sinclair antenna connected to good Heliax™ feedline will last 20 to 30 years of seriously hostile winters. How many Comets, Diamonds, Hustlers, homebrew antennas or replacement lengths of RG‑series or LMR‑series feedline will you buy and install in the same 20-30 years?

And don't forget that most of the really good commercial 2-way sites are owned by people that have been burned - and badly - by ham radio groups in the past. Some have been burned repeatedly. Some sites require a bonded and / or certified tower climber to do the work. If that's the case you can expect to have to pay some big bucks - both an arrive‑on‑site fee and a hourly rate. Some tower monkeys charge from the time they leave home to the time when they areive back home. Some charge from when they arrive until when they leave. Some charge by the vertical foot - doing something at 175 feet costs more than at 75 feet. So replacing your $120 Diamond antenna at 200 feet up the tower may cost upwards of $900, and in some cases upwards of $1500, and that's EACH TIME. For that much money you could have had a NEW Sinclair, DB Products or a Phelps-Dodge complete with NEW Heliax the first time instead of a Diamond and cheap no-name RG‑ 214‑type cable... and no need to touch it or replace it for 20 years or more. At one of the sites I visit frequently there is a 35+ year old ham antenna (fiberglass, with a top support) with 35+ year old 7/8‑inch Heliax that still works perfectly (yes, the antenna has been re‑gell‑coated at least twice).

A receiver preamp can help some systems, but realize that good preamps are all about the internal noise figure (NF) of the preamp, the amount of gain it provides and the point where it overloads. The lower the NF, the weaker the signal that will be heard. Regarding gain, most modern preamps have gain on the order of 14‑18 dB in a single stage. That's enough to take a signal that's barely above the noise floor and make it very usable on a decent receiver. But what works on the workbench may not be what works at the site... it's all about effective sensitivity. If you're already hearing down to the noise floor without a preamp then adding one isn't going to make any improvement in sensitivity and in fact may create new problems. But if you do decide to add a preamp, don't attempt to use it to make up for a long, lossy feedline - there's a good reason that the professionals use inch‑and‑five‑eights Heliax for a receive antenna: minimal loss). I've also seen three‑and‑one‑eighth and six‑and‑three‑eights Heliax. Likewise don't attempt to use a preamp to make up for a poor receiver - fix or replace the receiver.

Many sites use shared top‑of‑the‑tower "community receive" antennas that feed a preamp / multicoupler panel. These systems are very popular at UHF, and to a lesser extent VHF.

Example: The author of this article visits a 5300 ft mountain top communications site on a semi-regular basis. This site is home to a 120 foot (37m) tower. The top of that tower is dedicated to receive only antennas ‑ one each for VHF (high band), 220 MHz, UHF, 800 MHz, 900 MHz and 1200 MHz. There are also hot spare antennas already in place for each of the aforementioned bands except for 220 and 1200 MHz. The master receive antennas for the VHF and UHF bands are broadband antennas manufactured by Sinclair Technologies. The VHF and 220 MHz antennas are connected to 7/8 inch Heliax, the others feed their own runs of inch‑and‑5/8 inch Heliax. All of the runs of Heliax eventually terminate into a separate high level AngleLinear preamplifier panel for each band.

All of the systems I maintain at that site are UHF so I am familiar with that antenna system. The output of the UHF preamplifier feeds a multicoupler (splitter) and that feeds a bank of "window" filters. Each window is made from 8 cavities in series. The first window is for the UHF ham band. The second window is for the commercial repeaters that receive at 455‑460 MHz. The third window is for the 465‑470 MHz commercial repeaters. The fourth window is for the 473‑476 MHz commercial repeaters. Finally, the fifth window is for the 509‑512 MHz public safety repeaters.
The entire UHF window filter system completely fills a 7 foot rack that has't been touched in 20+ years (no need to). The adjacent rack has the post‑window distribution amplifiers ‑ each window filter feeds an AngleLinear distribution amplifier panel for that range's repeater receivers.

Over on the VHF side there is a single broadband, VHF high band Sinclair antenna. The Heliax feeds it's AngleLinear preamplifier panel and a distribution amplifier that feeds all of the VHF repeater receivers‑ each has a series pass cavity in front of the receiver. There is a similar system for the 220 MHz amateur repeaters. The one 6‑meter repeater at that site has it’s own antenna as well as it’s own 6‑cavity duplexer.

In most repeaters the duplexer provides a certain amount of isolation between the receiver and the transmitter (some systems, like those that use two antennas, or even two sites, don't use duplexers). If the amount of isolation, however it is acquired, is greater than what is required (the excess is sometimes referred to as "headroom"), then the system design is adequate for the job (see the article Some thoughts on Repeater Receiver-to-Transmitter Isolation on the previous page). That situation is fine until they decide to add a preamp to help out the handheld users. Then they discover that the amount of isolation isn't enough. They forgot that you need (at least) the same amount of extra isolation ("headroom") as the amount of gain the preamp provides, since it raises the apparent noise floor as well as the signal of interest. In many cases you will have to fight with desense when you add a preamp (a top‑quality preamp like an AngleLinear will help). Always have enough extra headroom in your receiver, transmitter and duplexer to handle either or both of a couple of situations: First, the site owner adds additional transmitters to the site, or second, that you want to add a preamp later on. If the duplexer is your primary provider of receiver‑to‑transmitter isolation then do not scrimp on the duplexer.

Next to a quality antenna and feedline the duplexer provides the primary receiver‑to‑transmitter isolation. Because it is primary it is the most critical part of a good repeater system ‑ but that is worthless if the repeater receiver and transmitter does not have good isolation built into the radio or radios that you are using. The use of double‑shielded coax between the duplexer and both the transmit and receive antenna connections is a must to limit the problem of the TX signal leaking into the RX input and desensing it (the Motorola GR300, GR400, GR500 and CDM700 low-end repeaters were (in)famous for using cheap coax on their factory duplexer harnesses. Simply swapping the three factory-provided RG58 cables for RG400 usually gets you from 2 db to 6 db reduction in desense).

Good RF bypassing on any wire leaving or going into the repeater and control circuitry is a must. This keeps any RF from causing problems in the audio circuits and control circuits of the radio. Then you need a good sharp front end on the receiver, good intermod rejection, and good clean transmitter signal with no spurs. A large majority of the lower cost repeaters from Yaesu, Motorola, and others are built from modified mobile radios that have a DC‑to‑daylight front end, and that's asking for trouble at a site with high RF levels. A in‑line pass cavity on the receiver will help, but that adds measurable insertion loss. Your author ran into this on a GR300 tabletop repeater (dual GM300 mobile radios) at a shopping mall... The receiver died and the local shop replaced it with a CDM mobile... the GM300 has a untuned front end, the CDM has a varactor tuned front end. The desense was noticeably lower on the CDM.

Back to duplexers... Long ago I gave up on four-cavity duplexers (two cavities on each side) on VHF/2m, 222 MHz and 440‑470 MHz UHF, I use the six cavity pass / reject type exclusively. It's not unusual to see 2 to 3 millivolts of VHF transmitter noise 600 KHz away from your repeater transmitter. Duplexer tuning is very, very critical. A return loss bridge is preferred, a spectrum analyzer with a tracking generator is the second choice. And don't tune the duplexer on the bench, then transport it to the site over a bumpy four‑wheel‑drive road, and expect it to be as precisely tuned when you get there. Likewise don't trust the tuning of ANY duplexer once it's been shipped - always assume that the transportation method will shake things up. Always have the test gear with you at the site to verify final tuning after mounting it in the system rack.

On the transmit side, never forget that SWR is not the only measure of antenna performance. A low SWR only means that the transmitter is "seeing" a reasonably non-reactive load. That is, it is neither capacitive or inductive, it looks like a 50 ohm resistor (a dummy load). The SWR tells you nothing about what is really important, the antenna efficiency, the antenna noise level, its pattern (gain) and decoupling of the RF from the feedline. And don't forget that the feedline loss runs both directions and can dramatically affect the SWR reading !! Your transmitter sends power up the feedline (let's say it's 100 watts), and some gets lost going up (let's say that it's 20% and 80w gets there). The lightning-damaged antenna reflects some power back down (let's say it's 10%, or 8 watts). The 8 watts comes back down, and 20% gets lost, and you see 6.4 watts on the Bird wattmeter. So you see 100 watts going up and 6.4 watts coming back, and you think the SWR is a lot better than it really is at the antenna. Look at this web page on the topic: Power Antenna Manufacturing Inc. SWR Calculator (an offsite link). It removes the "masking effect" of the feedline loss.

Speaking of "antenna gain", there are a lot of people that don't understand it. Basically it involves directivity. Every db of gain is a db that isn't going somewhere else. Some people actually think that by picking a higher gain antenna they can "regain the loss caused by the coax". Well, that's a total joke. Go read the Antenna Gain writeup by Marc Dekenah ZS6MGD at http://www.marcspages.co.uk/tech/antgain.htm.     (local copy)

Before you sign a site agreement, pay your money, and go to the trouble of installing a complete system you will want to measure the noise floor on your target frequency at the site. Just borrow an appropriate already-installed antenna for fifteen minutes or so (with permission) and make the measurement (i.e. if you are installing a VHF system use a VHF antenna). If site "A" has a 0.8 µV noise floor (due to the broadcast and paging systems there) and site "B" has a 0.1 µV noise floor take a guess as to which site will hear better (and I have seen sites that have noise floors that are above 3 µv on the frequency of interest... Yes, three‑dot‑zero, not zero‑dot‑three). And fifteen minutes may not be enough - you want to get all the offenders, and that may take an hour or two, or even longer... some commercial transmitters only ID every half‑hour, some every hour or so, and some do not have an ID at all, and they get away with it. Many multichannel trunking systems only ID the lowest frequency, the rest do not ID at all. Some public safety systems do not ID at all. Some signals may be from systems that are active only Monday through Friday during business hours. There are modern spectrum analyzers with digital memory and some can do additive recordings - they can hands-off assemble the worst-case scenario, and I have seen one of those parked at a site for two weeks. But there are workarounds: on a site visit 20+ years ago I saw a motor home at a site and parked outside the building, and inside there was a tripod with a video camera tight-focused on a spectrum analyzer... the gentleman volunteered that he had borrowed a good but abandoned-in-place antenna for the analyzer, had his motor home (with a refrigerator, a microwave oven, a TV and DVD player) outside, and that he had a enough videotape to get over 96 straight hours... he said he was going to sleep for 5 hours and 45 minutes at a time (he claimed he had two alarm clocks, just to make sure) then change tapes... He volunteered that when he got home he was going to do frame grabs with his computer, superimpose them and build up a worst-case analyzer image. And he was going to make multiple trips over a month to try and get all the signals.

Once you have a site interference profile mapped, you may chose to install a UHF system at a particular site if the noise floor on VHF is intolerable (or vice versa). And don't overlook 220 MHz, as the 220 noise floor at many sites can be way below that of 2m or UHF. And the atmospheric noise at 220 is lower than 2m or UHF as well. Remember that with repeaters it all depends on how well you hear (see the article on measuring effective sensitivity). Given the limits of antenna power rating, feedline and the local geography / topography, increasing the talk range is easy - how much amplifier, duplexer and antenna can you afford?

When you go attacking a desense problem, don't assume that any on‑channel receiver noise is a function of energy from the transmitter. That's often not the case. The transmitter can be absolutely clean, the cavities and / or duplexer tuned perfectly, but nevertheless, if there is some corroded metal joint in the antenna induction field, it will generate broadband noise, some of which inevitably will be on channel noise. Or the source could be someone's old LMR cable on their simplex dispatch system). You can play with your transmitter, duplexer and cables forever and it won't mitigate this problem. It may take crawling all over the tower with a sniffer to find it. Basically, you find the biggest culprit, solve that, and try again, and keep doing this till you're fed up, happy with it, too tired, or give up. Think "whack-a-mole".

There are several publications that you should look at if you are going to be doing any serious antenna systems work at a site. One is the site owner/manager's requirements, others are below on this page in each section. Another is the Motorola "R56" "Standards and Guidelines for Communication Sites" manual. The printed version is part number 6881089E50 and the CD-ROM version is 9880384V83. A good overview of that manual is here. And for a long time the BLM web site had a PDF of the 2005 ‑B version (outdated) of the print manual for free download. At the time of the last update to this paragraph (2020) the 2017 revision of R56 (the ‑C version) was current and that's the one you want. There were some very significant updates over the previous (2005) ‑B version. Quite a bit of content was added in addition to some content that was changed. One big difference was a change in naming conventions to clearly define different parts of the Bonding / Grounding infrastructure to better define their purpose (Bonding or Grounding). Every so often someone posts the PDF of the ‑C manual (it usually gets taken down rather quickly) but it's worth Googling for it every so often... just look for "6881089E50-C" or "68-81089E50-C".
And if anyone discovers that it's been updated to a ‑D please let the page maintaner know.

Another good (but dated) reference is here (if there is a newer version the author is not aware of it): GE/MA-Com/Tyco Installation Manual LBI-39185C titled "Specifications, Guidelines, and Practices, Tower Requirements and General Specifications".

Some people build their own repeater antennas from scratch, frequently copying a commercial design, sometimes doing their own experimental thing. Others rebuild surplus commercial antennas, others just buy something new. It's worth getting to know someone who knows how to bend and weld aluminum rod and tubing, and someone else that has access to hot‑dip‑galvanizing equipment... and someone that can build a phasing harness... the antenna systems page at this web site has detailed drawings of many good commercial antennas (and their harnesses) - and they are detailed enough to be used for amateur construction / duplication.
And please let the page maintaner know if anyone finds a source of replacement harnesses for the DB‑408, 411, 413 or 420 antennas.

It's also worth getting to know the local two-way techs and building / site owners.
One example: Over a decade ago an acquaintance acquired a very expensive low band antenna (the same model is over $1200 in 2021) after a phone call from a tech employed by another site tenant alerted him that a certain site had a recently abandoned‑in‑place 43 MHz side-mount dipole antenna. The acquaintance contacted the site owner, let him know that he had heard about the antenna, and the owner confirmed that the tenant had moved out and left the antenna on his tower. My acquaintance asked if he could have it if he could arrange to have it removed. The answer was yes. So he rearranged his schedule to follow the site owner to the site (i.e. at the site owners convenience), and after confirming the type of antenna my acquaintance asked if he could pay the site owners preferred tower climber to remove the antenna, but the owner waved that off. The owner said that he was going to have the climber at the site the following week for another reason, he'd have the climber remove the antenna, and the acquaintance was welcome to be there to haul the low band antenna away. And the acquaintance rearranged his schedule again, borrowed a vehicle with a roof rack and hauled the antenna away. That acquaintance rebuilt the antenna to 52 MHz and it is in service at a different site.

The Metric System - learn it!   The U.S. has been slow to adopt the Metric system of measurement, yet in ham radio circles we use metric terminology daily ("80 meters". "2 meters", "23cm" and more). However whenever I'm building an antenna, or a phasing harness, I'm doing conversions back and forth all the time. I often thought that it would be very nice to have a dual-marked metric / fractional shop ruler and tape measure (both Metric and English graduations). After all, there is a good reason that duplexer inter-cavity cables are precisely cut to an eighth of an inch (about 3mm). Well, after some searching I found that Stanley Tools of New Britain, CT has a model 33-726 (8 meter / 26 feet) steel tape, and there's a model 34-827 PVC coated fiberglass 30 meter / 100 foot for doing long runs. And I'm not saying that Stanley is the only one, there are several manufacturers that make metric / fractional steel tape measures. Most of the big "super home centers" don't seem to carry metric / fractional taep measures with any consistency, but you might try one of the smaller, family-owned hardware stores to see if they can special order one for you. My local Ace store was happy to do that... A metric / fractional tape measure can make your next antenna construction project go much more quickly and with more confidence that everything has been measured correctly.

Update July 2011: Harbor Freight has a 50 meter / 163 foot composition tape for under $12. It's metric on one side and feet on the other. However they caution you, on a sticker on the housing (and the sticker frequently goes missing), that temperature and humidity can cause the tape material itself to stretch and shrink as much as a full inch causing that much inaccuracy in the full length. But getting a 150+ foot measuring tape for under $12 is a darn good deal. Just don't use it for making a duplexer harness or making a dipole array - that's what the steel measuring tapes are for.

Update 2021: Amazon is carrying the Stanley 33-726 metal tape (8 meter / 26 foot) for under $30.

Update July 2023: Amazon has been out of stock of the Stanley for over 6 months however they have the Klein Tools 9375 metal tape (7.5 meter / 24 foot) for under $30.

Credits and Thanks:

Alex KJ7KIN made some very useful comments on some of the wording of the material above.

Contact Information:

The author, Mike Morris WA6ILQ, can be contacted here.

This material is © copyright 2003 and date of last update by WA6ILQ. It was at this web site on another page that was created in 2003, and was moved to this page on 12-Nov-2011.


Layout and hand coded HTML © Copyright 1995 and date of last update by Mike Morris WA6ILQ

This web page, this web site, the information presented in and on its pages and in these modifications and conversions is © Copyrighted 1995 and (date of last update) by Kevin Custer W3KKC and multiple originating authors. All Rights Reserved, including that of paper and web publication elsewhere.