If you do any work with analogue signals at frequencies above the most basic audio, it’s probable that somewhere you’ll have a box of coax adaptors. You’ll need them, because the chances are your bench will feature instruments, devices, and modules with a bewildering variety of connectors. In making all these disparate devices talk to each other you probably have a guilty past: at some time you will have created an unholy monster of a coax interface by tying several adaptors together to achieve your desired combination of input and output connector. Don’t worry, your secret is safe with me.
There are a huge variety of factors that lie behind the choice of RF connector on a piece of equipment. You might imagine that the most important would be the physical and electrical specification of the connector itself, but other factors such as company design policy, the accepted norm of a particular field, and the personal preferences of the designer come into play. This article will look at some of the different types of connector and try to explain some of those choices.
Form Follows Function
Before looking at individual connectors it’s worth looking at the much broader picture, of what an RF connector does and why it is designed in a particular way. And for that we need to talk briefly about characteristic impedance without descending too far into the mathematical proofs that detain first-year electronic engineering students.
So, imagine an infinitely long piece of coax with RF flowing though it. If you were to non-invasively measure the RF voltage and current, you could then calculate an impedance in ohms in a similar way as you might a resistance in a DC circuit. In a perfect piece of theoretical coax this impedance figure will be the same no matter what the RF frequency or voltage is; it is dependent on the physical properties of the coax itself, the dimensions of the conductors and the dielectric properties of its non-conducting components.
If you connect anything to the coax in the previous paragraph, it must be designed to present the same impedance as the characteristic impedance of the coax. If a different impedance is presented then some of the RF will be reflected back down the coax, the total impedance of the system will be changed, and signal loss will occur. From the RF signal’s point of view then, anything you connect to it should look electrically like just more coax.
There in a nutshell is the problem faced by designers of RF connectors: to ensure that their product appears throughout all its parts as exactly the same impedance as the coax it is connecting. Electrically it must appear as though the connector is not there: an unbroken piece of coax at the desired frequency of operation.
To achieve this perfect impedance match the designer of an RF connector must ensure that the internal diameter of the outer shell, the diameter of the centre conductor, and the dielectric properties of the insulator are such that at all points during the signal path the characteristic impedance is the same. If you were to saw a typical high-performance RF connector in half lengthways then you would see it in action, great care has been taken to achieve this aim.
It’s worth mentioning though that sometimes this goal of constant impedance has not been pursued. Popular connectors have a habit of staying in use for many decades, so a few that you’ll encounter such as the Belling-Lee antenna socket on European TV sets or the now-inappropriately-named “UHF” you’ll see on HF radios are really old designs conceived before the mechanism of characteristic impedance was fully understood. Or in the case of the “RCA” phono connector, a similarly aged design whose roots lie in audio and whose RF properties were of no interest to its originators.
There are a myriad coax connector styles, both obsolete and still on the market. Some of them have very specific applications while others are intended for general-purpose use. Fortunately of that huge catalogue only a few styles have become popular enough that you might commonly encounter them, so when taking a look at individual connectors in detail we only need to consider a few examples.
Welcome to the Coax Connector Jungle
Starting with the oldest of our selection, the Belling-Lee has its roots in the 1920s. If you are an American you may never encounter this connector, because it survives only as the antenna connector on European and Australian TV sets. Thus you can buy set-top boxes and even RTL-SDR sticks with one fitted. It is used with 75 ohm cables for UHF TV, but it does not present a constant impedance to the cable. Probably the only reason you’d use one would be if you are working with very old equipment, a TV set, or a TV peripheral, so it is presented here largely for information.
The UHF connector by comparison survives in a multitude of places, being used for video as well as with radio transmitters and receivers from HF into the VHF range. Dating from the 1930s, it takes the form of a shielded version of a banana plug with a threaded collar to provide a secure connection. It does not present a constant impedance, so is recommended for use only below 150MHz. “UHF” meant a somewhat lower frequency in the 1930s to what it does today. You’ll also see it referred to as a “PL259”, a wartime US military designation for one variant.
You’ll find the F connector in cable TV systems, satellite TV installations, and as the antenna connector on American TV sets. The F is probably one of the easiest connectors to fit, using the centre conductor of solid-core 75 ohm coax cable as its centre pin. It has surprisingly good performance for such a simple design, with a bandwidth of well over 1 GHz.
The N connector was one of the earliest constant impedance connectors, and dates from the 1940s. Available in 50 ohm and 75 ohm variants, capable of handling high RF power, and with a bandwidth over 10GHz, this connector appears on a wide variety of UHF and microwave transmitters and other equipment.
The BNC dates from the early 1950s, and has become ubiquitous everywhere from video systems through test equipment, handheld transmitters, Thinwire Ethernet, and many more applications. It is a constant impedance connector available in both 50 ohm and 75 ohm versions, and has a bandwidth somewhere around 5GHz.
SMA connectors were developed in the 1960s, have a 50 ohm constant impedance and a bandwidth of up to 18GHz. Being intended for use with semi-rigid coax on RF paths inside equipment they are not designed for more than 500 mating cycles. You will encounter them on some hand-held radios and most WiFi equipment, and they seem also to have become common on software defined radios. There is a variant of the SMA that you will also encounter, the Reverse SMA has its pin on the threaded side of the connector. This was an attempt by the manufacturers of WiFi equipment to stop end-users fitting Yagis and similar high-gain antennas, thus breaking radio emissions laws.
And finally, the MCX connector is a sub-miniature connector developed in the 1980s. They have a 6GHz bandwidth and are available in 50 ohm and 75 ohm variants. The MCX is included here because it is commonly used on RTL-SDR USB digital TV sticks, though if you own an RTL-SDR the chances are you use it with an adaptor to another connector series.
That’s the list of common coax connectors. Now which should you be using? Of course, as with any component choice, it depends on your application. But since few of us are working at the bleeding edge, it’s likely that none of our applications are going to be so demanding that any of a selection of the above couldn’t do the job. Thus there is another factor: personal preference. My personal preference will be different from yours: you may love the simplicity of an F while hating the soldered centre pin of a BNC, for example.
So here follows my preferences — what I reach for when I want a coax connector.
If rugged dependability is your requirement and you’re not looking into the microwaves or making a tiny device, the UHF is by far the best choice. They’re not too expensive, they’re extremely easy to wire, almost indestructible, and they mate very solidly. This is the connector I will carry to the top of a mountain with me, the one I know I can fix with a gas pen soldering iron by the light of a mobile phone screen and still achieve a usable SWR. My first ever transmitter I built in 1987 had one, I know they aren’t the best connector but I also know I can rely on them completely.
My go-to connector is the BNC. Cheaper than an N connector and with way more bandwidth than I am likely to ever need, I’ll concede they’re a bit fiddly to wire up but they’re not too big and I know they’ll readily plug into almost anything I’ll bring onto my bench. They plug directly into my oscilloscope, for instance. One day I’ll get round to ordering the right tool to do up those nuts.
The connector I prefer to avoid is the SMA. I’ve used them where I have to, but through convention rather than preference. So for instance I made an upconverter for RTL-SDRs using them, because SDR owners are more likely to have SMA than BNC in their armoury. They are nice and small, but not at all cheap (they are usually gold-plated, after all!), and I find the plugs to be very fiddly to wire. Plus with that 500-mating-cycle life you know that some time in the future you’ll have to replace the connector. So in the picture at the top of the page you’ll see a selection of SMA adapters that are first out of the box when an SDR lands in front of me so I can connect a BNC to the system.
It’s important to note though that there is no wrong answer in this field. Open up an RSGB or ARRL VHF or UHF handbook from the 1950s and you’ll see a plethora of components and techniques that you’d be forgiven for looking at in amazement from 2016 that they ever worked at all. If there’s a lesson to be taken away from that it’s this: your radio is only as good as what you manage to do with it.