Back Isolators and Circulators...
Compiled and HTML'd by Mike Morris WA6ILQ


A single stage tunable isolator made by dB Products

The information below was compiled and summarized from personal experience plus a number of postings on the repeater-builder mailing list.

A ferrite circulator is a magnetic three-port device that acts like a one-way-valve for RF energy, and the third port usually has a dummy load attached. An isolator is just a circulator with a built-in or bolted-on load (usually fairly low in power handling capability), and so as long as you stay within the power handling limits of the load then all comments below on isolators also apply to circulators and vice versa.
Just remember that isolator = the combination of a circulator + an externally attached load.

The most common use of an isolator/circulator is between the high power RF source (i.e. the transmitter) and the duplexer, or if the source is made up of a transmitter and an external amplifier then between the amplifier and the duplexer. In most cases the isolator is mounted on or near the RF source (transmitter or power amplifier). Since the isolator presents a 50 ohm load to the source the cable length between the source and the isolator is not critical. A single stage circulator / isolator will provide about 35 db of isolation and a dual stage unit about 70 to 75 db.

Ferrite isolators/circulators are physically fragile devices - and for this reason you NEVER want to pay full price for any isolator that has any visible damage (especailly a dent in the housing that could have been created by dropping it or something hitting it). Once you get your second-hand isolator/circulator on the bench you may discover that it is as useful as a doorstop (and I've seen a low band (42 MHz) isolator that externally looked perfectly good being used as a doorstop because it was cracked internally).

Circulators also require tuning unless manufactured as broadband fixed units. Those with tuning adjustments are a balance of Insertion Loss versus Return Loss. A tad more Insertion Loss favoring better Return Loss is my usual preference. Keep the transmitter happy. This tuning requires either a VNA or at minimum a Tracking Generator with Return Loss Bridge.

If you see a "4-port isolator" or "4-port circulator" advertised it's probably a dual stage device with two units in one box with the output of the first fed into the input of the second, and with two dummy load ports on the housing. Triple circulators also exist, and in fact were a factory option on the Motorola MSF5000 stations. Note that the load nearest the transmitter port will usually be much larger than the output port. The load on the output port disipates the reflected power from the antenna, the load on the input / first stage only has to disippate any power reflected from the second stage (which, by design, is very small).

IMPORTANT: The frequency "range" shown on the spec sheet for a ferrite isolator / circulator is not the field tuning range of the device, it is in reality an indicator of the manufacturer's capability to build a device to any one particular frequency. For example, once an isolator / circulator is made for a particular frequency, it can only be tuned about two to three percent either side of that frequency and retain optimum performance. Once outside of that "window" (which is 4 to 6 MHz wide at 2M / high band), it will cost between $200 and $300 (about 1/3 of the new price) to have it remanufactured from (for example) a 167 MHz frequency to your specific 146 MHz frequency (or from a commercial UHF frequency to your 440 MHz frequency) because it involves a considerable amount of precision bench work using RF test equipment, special mechanical tools and professional skills that very few shops or people have: copper or bronze vises (nonmagnetic) to position and clamp the isolator "just right" and nonmagnetic tools to do the actual assembly or repair, plus a network analyzer to get the RF performance "just right". Besides, having your unit rebuilt by a person who does it 8 hours a day, and every day guarantees that it is done right the first time, it comes with a factory guarantee and lands in your hand already precision tuned on your frequency and with some nice proof-of-performance plots. And it is much cheaper than buying a new device!
Edit: Prices quoted are 2009 prices)

From an email to repeater-builder from Eric Lemon WB6FLY:

The tuning procedures for isolators/circulators do not vary much among brands. To do the tuning properly, you really need a network analyzer, or a spectrum analyzer with a tracking generator and a return loss bridge. You have to be able measure return loss. Be aware that some isolators have a very narrow tuning range, and cannot be tuned away from the delivered frequencies more than a few MHz.

In one specific case a local radio club had a donated Sinclair dual isolator that was originally manufactured for 163.5625 MHz and it could not be tuned down to the 2 meter band. It was shipped to Sinclair for re-manufacturing to the desired 147 MHz amateur frequency, and it has worked perfectly ever since. It cost the club US$250 for the rework, but that's about 1/3 the cost of a new unit of that model.

And just because you have access to a network analyzer, or to a spectrum analyzer with a tracking generator and a return loss bridge does not mean that you can tune a circulator.

Be forewarned: Once you open an isolator/circulator housing the manufacturer may refuse to perform the rework - it's way too easy to crack a magnet or something and then all you have is an expensive precision doorstop or paperweight...

If you plan on using your own dummy load on a new circulator, you should send that dummy load with the circulator to the vendor so that the circulator can be properly tuned. Due to the variation in impedance and return loss between individual dummy loads (even if outwardly identical), it is never a good idea to change the load(s) after an isolator is properly tuned. Hint: put a stripe of nail polish across the connectors of the isolator and the load, and use an obnoxious color (like fluorescent pink, bright orange or purple). If anybody swaps the loads then it will be obvious (the broken nail polish will reveal it even if it was borrowed and replaced).

One characteristic of ferrite isolators and circulators is that they generate harmonics. This property is inherent to the device and cannot be "tuned out". For this reason, every isolator / circulator is always followed by a bandpass or low pass filter in any application where good engineering practice is followed. This same situation is why folks say that isolators cause intermod. They do! NEVER, EVER have an isolator or circulator looking directly into an antenna !

Example: Years ago I heard about a new 52.525 MHz remote base that was added to an existing UHF system... it was "done right" with a a circulator but they forgot the low pass filter on the 6m transmitter. It was added to a UHF repeater that had a point-to-point link on 420.200 MHz to another system. So? You ask... Well, the 8th harmonic of 52.525 MHz is dead on 420.200 MHz. Whenever the remote base transmitted it interfered with the FM station on 105.05 MHz (actually the station on 105.1) in the FM band, it interfered with 157.575 MHz (with a signal 15 kHz wide), and since 5 KHz deviation at 52 MHz is 40 KHz at 420 MHz it trashed 420.175 MHz, 420.200 MHz and 420.225 MHz... (and since a "5 KHz" deviation signal actually occupies about 16 KHz of spectrum the interference was actually quite a bit wider than the single VHF channel or the three UHF channels). Ooops. Somebody hadn't done their system planning homework.
Always remember, ferrite isolators / circulators are non-linear devices, and as such, are capable of generating harmonic content on their own. Generally speaking, a low-pass filter should be used on the output of the isolator, though in some cases a bandpass cavity filter can also be used, but you need to watch for odd-order harmonic responses if using a conventional bandpass cavity filter. Yes, (ignoring insertion loss for a moment) a 150 MHz cavity will pass 450 MHz (x3), 750 MHz (x5), 1050 MHz (x7), 1350 MHz (x9), etc equally well. A pass / notch duplexer generally will not provide the required harmonic attenuation.

When you rent space at a 2-way site and are handed the contract terms do not be surprised to see the requirement of an "intermod panel" on each transmitter. This "panel" consists of either (a) an isolator followed by a pass cavity, or (b) a circulator, an external dummy load, and a pass cavity. Many radio communications equipment manufacturers make isolators and circulators: Andrew, Connecticut Microwave, EMR, Celwave, Decibel Products, MECA, REC-USA, Raditek, Telewave, TX-RX, and many more. There may or may not be a amateur radio club discount at any given manufacturer. Always ask - if you don't ask they can't say "Yes".

Intermod panels are not cheap because they are made to order on your frequency. Expect to pay about US$900 for a single isolator (with a 50 watt load), about US$1,400 for a dual isolator (same), and about twice that for one with a 100 watt load. Your author has seen ones rated at 500 watts with a 100 watt load at one site and with a 250 watt load at another site. 900 MHz panels are about 10% higher in price. Telewave offers turn-key "Repeater Panels" that include a duplexer, a receiver preamplifier, a receiver preselector, a dual isolator and a low pass filter for about US$4,500 for high band, and US$4,300 for UHF or 900 MHz. (All prices in this paragraph are as of mid-2009.)

Remember that a ferrite isolator/circulator is a MAGNETIC device, and don't mount it on a steel rack panel (and other circulators)... you WILL detune it... use aluminum panels only, using brass screws and aluminum standoffs if you need standoffs (or use a block of wood). I'm only mentioning it because I forgot when I was cleaning up a repeater cabinet - the UHF circulator was floating in the breeze, supported by the 7/8 heliax jumpers, so I rearranged things and mounted the circulator to the same rack panel as the UHF pass cavity... then the system transmitter performance dropped like a rock. A group member that lived in the fringe area of the system (he had a beam antenna on his tower fixed pointed at the repeater) reported that the system went from its normal 90% quieting at his site to about 1/3 quieting. It was only after I measured the SWR between the isolator and the transmitter did I realize that the isolator was detuned... I switched the mounting back to "in the breeze" and the system performance went back to normal... The temporary mounting the following day was a piece of 1x8 inch pine shelving mounted to the rack rails with drywall screws through the holes in the rail into the wood. The permanent fix a couple of weeks later was relocating the pass cavity and the isolator to an aluminum panel and using brass screws to mount the isolator to the panel. The pass cavity didn't care, I used regular screws for that. Brass hardware is available at most marine/boat shops.

If you are going to make any adjustments to a tunable isolator / circulator (like in the photo at the top of this page) then you need to do it with low power - 2 to 5 watts, maybe 10 watts max, and EACH port must be terminated with a 50 ohm load. You don't want any off-resonance circulating currents or voltages during tuneup... at high levels they could demagnetize (or change the permanent fields of) the internal magnets and therefore damage the isolator/circulator. Also when tuning, in an ideal environment you would have available another (already tuned) isolator/circulator to put between the RF source and the isolator/circulator being tuned to make sure the unit under adjustment sees a consistent 50 ohm source... in other words, the first one isolates the transmitter from the one being adjusted...

If circulators and isolators are a new topic to you, I suggest that you read the first four articles on the previous page, especially the one by William F. Lieske. Yes, there is some repeated material between the articles, but if you're new to the topic, it won't hurt. When you have multiple people writing on the same topic some repetition is inevitable.

Contact Information:

The author can be contacted at: his-callsign // at // repeater-builder // dot // com.

This web page was split from the main page 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.