Seeing and Understanding SWR

[mbrun]

As SWR is a constant topic in radio, and since it may be difficult initially to grasp the concept of, I thought it would be helpful to provide a link to an incredibly effective video that allows you see it in action. The video is 60 years old and was produced by AT&T labs. It is worth watching from start to finish. You will get to see standing waves, reflected waves, impedance matches, mismatches, shorts and opens. It does not get visually any clearer than this.

 

 

Enjoy.

 

 

Michael

WRHS965

KE8PLM

[jgillaspy]

That was cool! Thanks for the link.

 

JG

[kidphc]

Saw that video a while ago. He was an incredible lecturer, some of other videos are just as informative. 

For me at least that video help conceptualize SWR.

• Another awesome video on antenna concepts in reality.

[kidphc]

Another great video on theories and why we do what we do.

 

[OldRadioGuy]

This is a cool demo but there is a simpler way to look at SWR.

Bad SWR is almost exactly like having your car (or bike) in the wrong gear.  

It means your engine (or legs) will not be as effective and you won't get the full horsepower from it.

If your gear is just a little high or low it's just not optimum.  It's not a big deal.

If your gear is way off it could either stall out or over-rev your engine or break your bike chain or whatever.

So you could break something but it has to be an extreme mismatch. 

With more power it becomes a bigger deal but even 50W is not a lot of power.

 

Vince

 

 

[AdmiralCochrane]

Wrong gear is more of a power mismatch - running higher power than the antenna is rated for or turning the tires more rpm/mph than they are rated for.  

SWR is more like clear glass letting all light pass vs semi-mirror reflecting some of the light back.  Full mirror would send zero signal out, all reflected. 

[OldRadioGuy]

Impedance is the ratio of current versus voltage that a source can generate when producing maximum power.  You want a load that uses current and voltage

A low Z source can make lots of current but not so much voltage.  Car audio systems often use 4 ohm and even 2 ohm speakers because the stereo is limited to 12V of the car battery.  So you need to "take" more current to get more power out of the 12V source.   Switching power converters have changed this more recently.

A hi Z source can make lots of voltage but not so much current.  As soon as you draw some current the voltage will drop.  You want a load that can make use of higher voltage but not need a lot of current.

It is almost an exact analogy to torque and RPMs in the mechanical world.  Transmissions are impedance matching devices.  They behave exactly like a transformer in the electrical world - trading torque for RPMs the way transformers trade voltage for current.  Torque being voltage and RPMs being current.

 

Vince

 

[BoxCar]

Impedance is the opposition of current flow in an AC circuit and replaces resistance in Ohm's law. E=IZ or E/I=Z and E/Z=I. There is no particular reason antennas have an impedance of 50, 75 or 300 Ohms, they were just values picked at random which could be easily replicated.

[OldRadioGuy]

The reason antenna (and therefor radio outputs) were designed to be 50 ohms is largely the coax.

50 Ohm coax is practical to make in manageable diameters and will carry decent amounts of power.  They determined that 50 ohm coax for a given size could carry more power than other impedances.  It was deemed optimum by some military research study.  Also, with low voltage solid state radios it works out very well. 

Once you choose 50 ohm coax for your radio system it makes sense to have the radio and antenna made to match it.

 

Of course there is 300 and 400 ohm twin lead transmission line (ladder line) and some radios used to be designed for it.  It made far more sense in the tube days to design higher Z finals on radios.  But that's mostly history - at least for our portable 12v radios.  

Most radio outputs are predominantly resistive impedance with little reactance (inductance or capacitance). 

 

Vince

 

[OldRadioGuy]

Here's something to ponder.

If a 2:1 SWR antenna were connected directly to the radio with no coax at all there could be no standing wave.  Standing waves need transmission line in order to exist.  A mismatch is still a mismatch even without a transmission line to allow standing waves.

You will still have the same loss of radiated power due to the same mismatch.  The standing wave itself is only a symptom of the mismatch.  The coax is just the middle man. 

When you have a "matching box" at the radio end of the coax things get really complicated.  There is still a mismatch between the antenna and the coax but the mismatch at the match box makes a "reciprocal" reflection that cancels out most or all of the bad reflection at the antenna.  The length of the coax will affect the match in this case.  This is why is it best to have the matching unit right at the antenna.   The icom AH4 does this for the HF bands and works very well with terribly mismatched wire antennas.  You do not see these things on the VHF/UHF bands where antenna size is not much of a problem.  Better to just make the antenna the right length.

Just some

Vince

[rdunajewski]

1 hour ago, OldRadioGuy said:

If a 2:1 SWR antenna were connected directly to the radio with no coax at all there could be no standing wave.  Standing waves need transmission line in order to exist.  A mismatch is still a mismatch even without a transmission line to allow standing waves.

You always have transmission line, whether it's just the short traces on the PCB going to the antenna connector, or 100' of transmission line. The standing wave can exist anywhere in that system, by the very nature of the fact the energy is not being efficiently radiated out due to the mismatch.

[OldRadioGuy]

Not true.  When the transmission line is shorter than about .1 Lambda (wavelength) it no longer behaves as a transmission line. 

I used to design RF probes and test gear for Maxim who made chips for cell phones etc.   I spent a lot of time on network analyzers and sim tools.  You don't treat every chip lead, pad, and bond wire as a transmission line... and you can't.  Chip designers do not treat anything like a transmission line.

Certainly you can call it what you like but as a practical matter it's not a transmission line anymore when it's a small fraction of a wavelength.

Vince

[OldRadioGuy]

After thinking about it..... a better way to put it is that there is no point in treating very short lines as T-lines.  There is no benefit in treating it that way.  In PCB and chassis  design it's mainly the delay of the T-line that has any meaning

When you design on a PC board, even at 1GHz you measure the S-paramterters on the PCB or at the PCB interface.  So all the parasitics related to the packaging and PCB interface are absorbed into the S-Parameters.  Matching network components are generally treated as discrete impedances - and this is pretty accurate.  Everything is surface mount today and the traces are typically as short as the pads.  Once we got past the matching network we would try to keep our microstrip in the ballpark of 50 ohms but in many cases you are only going an inch or two.  So if your microstrip is off 15% it's probably still better than your matching network. 

Even at a GHz PC boards have become too small to economically use stubs and T-lines for on board for matching.  You can buy an 0402 1pf cap or 1nH inductor for next to nothing.  PC board space is too valuable for stubs or T-lines of meaningful length. 

The only way to really see this is to plot it out on a Smith chart.   When the rotation is so small it just doesn't do anything. 

Vince

[AdmiralCochrane]

There is more tech knowledge on this forum than many of the "good" ham boards.  I was going to post that somewhere else and I still may