8/14/2019 Imp Match

1/3

From time to time youll come across the termimpedance matching in various areas of electronics, andespecially in fields like RF and audio engineering.However even in these fields its often misused, probablybecause many people dont really understand theconcepts behind it.

In this primer well try to clarify what impedancematching is really all about, why its important in somesituations and not important in others.

Textbooks usually explain the idea of impedancematching with a very simple example of an electricalgenerator feeding a resistive load, as shown in Fig.1.Because the generator has an internal resistance of itsown (RG) as all real generators do this tends todissipate some of the generators output power as heat,whenever we connect a load to its output terminals. Sothe full mechanical power fed into the generator cantbe drawn from it as electrical power, because some willalways be wasted in RG.

When early electrical engineers were faced with thisproblem, they naturally enough did everything theycould to reduce the internal resistance of theirgenerators. However they were inevitably still left withSOME internal resistance, because its impossible toreduce the resistance to zero unless you run thegenerator at a temperature of close to absolute zero(0Kelvins, or -273C).

Once they had minimised the internal resistance, theirnext step was to see if there was some way that theycould minimise the amount of power wasted in it, byvarying the resistance of the load. And what theydiscovered is shown in the Fig.2, which plots the amountof power transferred into the load RL as its resistance isvaried. As you can see, the amount of power reaches apeak or MAXIMUM when the load resistance is thesame as or matches the generator resistance. It falls

away for values both higher and lower than this figure,showing that matching the two is clearly desirable if wewant to maximise the power able to reach the load.

So this is where the idea of matching the resistance ofthe load to that of the generator came from. Beforelong, it was extended to cover the general situation ofany load impedance connected to a source of electricalenergy or voltage (EMF), with its own internal source

impedance. And the idea of matching the load andgenerator resistance became one of matching the sourceand load impedance impedance matching.

Now this may sound simple and straightforward, butits important to remember where the idea came from,and also realise what exactly is going on when we domatch load and generator/source resistances orimpedances. True, the POWER transferred to the loadwill be a maximum; but at the same time, the actualpower being dissipated in the generators internalresistance is EXACTLY THE SAME as that reaching theload! In other words, HALF the total power from thegenerator is now being turned into heat inside R G,because its resistance is now half the total connectedacross the generator. The other half is the loadresistance RL.

For exactly the same reason, RG and RL will now beacting as a 2:1 voltage divider across the generator

so that only HALF the generators output voltage (EG

)will be appearing across the load. In terms of voltagetransfer, then, matching the impedances isnt particularlyefficient: it actually gives a 6dB loss.

Does this mean that impedance matching really onlyapplies to generators in power stations? No, and in factit doesnt really apply there either or at least, notsimply. All it really means is that as you draw more andmore power from a generator by reducing the load

resistance, a point is reached where half thegenerators output power is being wasted inside it.Obviously with very high power generators its not agood idea to load them even this heavily let alonedropping the RL even further, where even more poweris lost inside the generator than reaches the load. (Seethe blue curve in Fig.2, showing the power lost in the

generator.) Most power station generators are loadedwith an RL somewhat higher than RG, to waste as littlepower as possible.

So when IS impedance matching a good idea? Gladyou asked. Basically its for situations rather differentfrom that in Fig.1, where were stuck with a particularload or cable impedance, and we still want to eithermaximise the power transferred into the load, orminimise the amount of power reflected back from itinto the cable, or both.

RF CABLE MATCHINGFor example in many RF situations, we tend to have

a relatively fixed LOAD impedance say a resonant

Electus Distribution Reference Data Sheet: IMPMATCH.PDF (1)

Fig.1: A generator driving a load resistance RL and its own internal resistance RG.

Fig.2: How the power fed to the load varies as the loadresistance is varied (red plot), and what happens to thepower wasted in RG (blue plot).

IMPEDANCE MATCHING: A PRIMER

8/14/2019 Imp Match

2/3

quarter-wave antenna, with an impedance of50 ohms resistive. To minimise interferencewe also have to use coaxial cable toconnect the antenna to a transmitter orreceiver.

Now as you may be aware, coaxial cablebehaves as a transmission lineat radiofrequencies, and as a result it has its owncharacteristic impedance. This simply meansthat because of the inductance-to-capacitance (or L/C) ratio of the cable, RFenergy tends to move along it with aparticular ratio between the electric andmagnetic fields (i.e., voltage to current).

In most cases when the energy reachesthe end of the cable, we want as much aspossible to transfer into our load theantenna, in the case of a transmitter, or theinput RF stage in the case of a receiver. For atransmitter this gives the highest power efficiency, whilefor a receiver it gives the best noise performance.

And guess what? To ensure this optimum energytransfer, we need to match the characteristic impedanceof the cable to the impedance/resistance of the load. Sofor a 75 antenna or receiver input, we need to use75 coaxial cable. For a 50 antenna we need to use50 cable, and so on. (see Fig.3)

This, then, is an area where impedance matching ISquite important. Because what happens if the cable andantenna (or receiver) impedances are NOT matched isthat some of the RF energy reaching the end of thecable wont be transferred into the load, but isREFLECTED back along the cable, towards the source.This can set up standing waves in the cable (anothercause of power loss, and possibly cable damage), and can

also cause overheating in the transmitter output stage.In a receiver, the mismatch degrades the effectivereceiver gain and noise figure.

How do you ensure correct impedance matching inthis type of RF situation? Generally the cable impedanceis more or less fixed, and the antenna impedance may bethe same. But quite a few techniques have been evolvedto tweak the matching between the two: tuned stubs,quarter-wave transformers and so on. Similar things canbe done at the input of a receiver. For details of theseRF matching techniques youll have to refer to a goodRF textbook, like The ARRL Handbook.

Notice though that so far weve only considered thesituation at the LOAD end of the RF cable. How aboutthe source end isnt impedance matching importantthere too? Less so, especially for transmitters. The main

thing is to ensure that the transmitter output stage willfeed as much RF energy as possible into the cablesinput impedance. There can even be an advantage indeliberately mismatching the impedances (i.e., having thetransmitter impedance much lower than the cable), tominimise power loss in the final stage and also ensurethat if RF is reflected back from the antenna end, most

of it is bounced right back up again. So this situation is abit like the generator in a power station...

VIDEO INTERCONNECTIONSNow lets consider another area where impedance

matching again tends to be quite important: videointerconnections. Here were dealing with signals whichspan from DC up to about 6MHz or so well into theRF range. And we also tend to find ourselves usingcoaxial cables, to reduce interference. So again we need

to match the cable impedance and theload impedance, to prevent signalreflection. With video, these reflectionscan cause ringingand ghostingin thefinal picture. (RingIng is multiple edges

on outlines in the picture, whileghosting is multiple images eachshifted horizontally.)

Most video equipment is designed tobe interconnected with 75 cables, andhas inputs which are designed topresent this same input impedance. Somatching tends to occur automatically,providing you use the correct cables.

How about video outputs arethese impedance matched too? Yes,generally they are, not because itresults in maximum signal transfer but

because with video signals we DONT want any signalreflected back from the load to be reflected back all

over again this would make ringing even worse. Sooften the video outputs of cameras, VCRs, DVD playersand so on are fitted with a 75 series resistor inside, toprovide back termination for the cable (Fig.4). This isjust another name for impedance matching at the sourceend of the output cable.

Note that just as with our original generator in Fig.1,this added impedance matching resistor produces aninevitable 6dB loss of signal half the video output islost in the resistor. Thats the penalty of impedancematching at the source end, and its why the outputbuffer amplifier in video equipment is usually given again of twice what is needed, to allow for theunavoidable 6dB loss when a cable and correctlyterminated load are connected.

Electus Distribution Reference Data Sheet: IMPMATCH.PDF (2)

Fig.3: A transmitter or other source of RF feeding power to anantenna, via a coaxial cable or other transmission line. Hereswhere impedance matching IS important...

Fig.4: Video interconnections, where impedance matching is againquite important.

8/14/2019 Imp Match

3/3

AUDIO IMPEDANCE MATCHINGHow about impedance matching in audio

applications? There are a few applicationswhere its important, but perhaps not asmany as you might think.

Because audio signals are quite low infrequency, its generally only where they haveto be sent over quite long cables thattransmission-line effects make it necessary to

perform impedance matching to preventreflections. And in most cases, we can getquite efficient signal transfer simply byarranging for the output impedance of ouraudio source (such as an amplifier) to bemuch lower than that of our load (such as aloudspeaker).

In the case of most hifi amplifiers and speakers, forexample, we generally arrange for the amplifier outputimpedance to be very much LOWER than the speakerimpedance. A typical speaker impedance is 8, forexample, but most hifi amplifiers have an outputimpedance of 0.1 or less (Fig.5). This not only ensuresthat most of the audio energy is transferred to thespeaker, but also that the amplifiers low output

impedance provides good electrical DAMPING for thespeakers moving voice coil giving higher fidelity.

Older valve amplifiers needed a different form ofimpedance matching, because output valves generallyhad a fairly fixed and relatively high output impedance,so they couldnt deliver audio energy efficiently into thelow load impedance of a typical speaker. So an outputtransformer had to be used, to produce a closerimpedance match. The transformer stepped up theimpedance of the speaker, so that it gave the outputvalve an effective load of a few thousand ohms; this wasat least comparable with the valves own outputimpedance, so only a small amount of energy waswasted as heat in the valve.

The only other area in audio where impedance

matching (of a different kind) tends to be important iswith transducers like microphones, gramophone pickups,tape heads and so on. Here the transducer often needs

to be provided with a particular load impedance, butnot in order to maximise power or signal transfer.Generally its to ensure that the transducer performanceis better controlled by the electrical damping effect ofthe load.

For example when correctly loaded these transducersmight have cleaner output, with fewer unwantedresonances and hence a much flatter response.

SUMMARYHopefully this has given you a better understanding of

the idea of impedance matching, where it came from andwhat its really designed to achieve.

As you can see, true impedance matching is generallyonly needed for RF and video interconnections andmainly at the LOAD end of coaxial cables or othertransmission lines.

Remember too that when impedance matching isperformed at the SOURCE end of a cable, theres alwaysa penalty: loss of power (-3dB) or signal level (-6dB),because when the generator impedance and theimpedance of its load are equal, half the power is

inevitably lost in the generator resistance.(Copyright 2001, Electus Distribution)

Electus Distribution Reference Data Sheet: IMPMATCH.PDF (3)

Fig.5: When a speaker is connected to a hifi amp, the impedancesare NOT matched.