I have a few very basic doubts regarding impedance matching. One or more of them might sound funny to most of you. But if it is so, please clear the doubts.
1. If impedance matching is all about matching the impedance of transmission line with the load, ie an antenna, why is this done only at this junction? Cant there be reflection of signal at other junctions?? like from source to transmission line, antenna to the medium (atmosphere). Y dont the waves reflect at these junctions? Y dont we do impedance matching for them?

2. In a practical case, If the signal comes out of an IC, how am I supposed to know the impeadance behind the pin of IC (if not given in datasheet)

 

Impedance matching is done at both ends of the transmission line in many, many cases.

 

Provided the load end of a line is matched then there should be no standing waves on the line. From the perspective of an ideal source at the driving end of a matched line, it should see the characteristic impedance [Zo]. One doesn't need to match the driving end in such an ideal case. If this were a 50 ohm line one wouldn't insert a 50Ω in series with the ideal source. One would lose 50% of the source voltage for no obvious purpose.
Mind you perfect load matching is rare as is the likelyhood of an ideal RF source.

 

Any time there are changes in impedance, there will be reflected energy. This will be true at couplings and connectors, which is why connectors are intended for use with particular characteristic impedances. For instance, you have 50Ω BNC connectors and 75Ω BNC connectors -- don't mix them up. Even bends in a coax line will cause local changes in the impedance and result in reflections (and which is why minimum bend radii are spec'ed).

RF sources are impedance matched to the transmission line. Look at signal generators. Many have a fixed 50Ω output and some have switches that you can use to get other common outputs. If you are using a source that has a different impedance than your transmission line, then you should ideally use an impedance transformer to couple it into the transmission line. Under a lot of situations, you can be pretty sloppy and still get something that works good enough, though it probably isn't working nearly as good as it could.

One thing to keep in mind is that impedance matching isn't as important when you are dealing with distances that are very short compared to the wavelength. So, bring a signal out of an IC and then taking it into an amplifier might not need a lot of attention paid to impedance matching in that part of the path unless you are talking really high (GHz+) frequencies. Another thing to keep in mind is that, in many cases, quite a bit of loss can be tolerated when working with the active low-powered portions of the signal chain. It's the largely passive portions and/or the portions dealing with very high power that need the most attention.

 

I have a few very basic doubts regarding impedance matching. One or more of them might sound funny to most of you. But if it is so, please clear the doubts.
1. If impedance matching is all about matching the impedance of transmission line with the load, ie an antenna, why is this done only at this junction? Cant there be reflection of signal at other junctions?? like from source to transmission line, antenna to the medium (atmosphere). Y dont the waves reflect at these junctions? Y dont we do impedance matching for them?

2. In a practical case, If the signal comes out of an IC, how am I supposed to know the impeadance behind the pin of IC (if not given in datasheet)

Hi Singh:

I highly recommend getting a copy of "REFLECTIONS" by Walt Maxwell, W2DU. It is the classic reference for transmission line theory.

In order to know the impedance that your driver IC "sees" you have to know the load impedance, the transmission line impedance, AND the length of the transmission line. The SMITH CHART makes this quite simple.

I have an entire chapter dedicated to this topic in my upcoming book, Radio Science for the Radio Amateur. It is not a simple topic, but it's also not rocket science!

In the special case that the load impedance is purely resistive and is equal to the transmission line impedance, the length of the transmission line is irrelevant. In this special case, your IC will always see the same impedance as a load. This is known as a "FLAT" transmission line, and is the ideal condition.

Hope this helps!

Eric

 

If a datasheet does not mention the output impedance, then they manufacturer probably doesn't think that its important for application for that device.

 

If a datasheet does not mention the output impedance, then they manufacturer probably doesn't think that its important for application for that device.

At relatively low speeds, the output impedance of a digital gate is generally assumed to be Zero with a low and infinite with a high. Usually a "build out" resistance is incorporated to reduce double reflections in case the LOAD impedance isn't perfectly matched.

In this situation, the main interest isn't maximum power transfer, but rather minimal distortion.

For radio frequency circuits, maximum power transfer is generally the main goal, so source and load impedances become more critical. This is a book in itself.

Eric

 

At relatively low speeds, the output impedance of a digital gate is generally assumed to be Zero with a low and infinite with a high.

I'm not following this. If the output impedance is infinite when the output is high, then it couldn't provide any current at all without deviating from the nominal output level drastically.

 

He probably meant zero impedance when a gate is an output and infinite when a gate is an input.