I have taken to attempting to align a couple ham transceivers, and the instructions often say to adjust for best waveform. I assumed that I should be trying to get things to appear the closest to a sine wave as possible, but what I see doesn't look much like a sine wave, at least not a single sine wave. As I adjust the adjustment point, I see that instead of losing the extra bumps in the waveform, there are places where the signal seems to become more consistent, meaning that I don't see lots of overlapping waves, and they mostly converge on the wave that the scope is triggering on. Is this what is meant by "best waveform", or does this usually have a set meaning?
"Best waveform" is surely ambiguous. Certainly, if the non-overlapping effect you are seeing is not a result of triggering difficulty and really exists in the signal, then reducing this effect is a good thing. This effect is
jitter, and I believe it's always bad. It can be seen using an oscilloscope. If the jitter is bad enough, it can also be seen with a spectrum analyzer (or FFT on an oscilloscope). It can be most precisely seen with a frequency counter.
So, what does "best waveform" mean in absence of an explanation, a picture, or a diagram of what the waveform should look like? If one of these instances is speaking about the shape of the wave, and the other is speaking about the frequency drift of the wave, that is somewhat confusing, and leads me to believe that there can be a nearly infinite number of ways one could consider something to be the "best waveform".
It's tough to say, but it is pretty easy to define the
ideal waveform. However, it's much easier to see how ideal a waveform is in the frequency domain (like a spectrum analyzer or FFT) rather than the time domain (like an oscilloscope display). The
ideal waveform will have only one spike in the frequency domain, at whatever the desired frequency happens to be. It would have no harmonics (integer multiples of the desired frequency) and no other content (spikes which are
not harmonics, picked up from mixing (LO feedthrough, intermodulation), power supply coupling, or other isolation problems (coupling to to other parts of the radio, etc)).
Almost all of these things are difficult to see in the time domain (oscilloscope display) unless they are very bad. Obviously, if you see a square wave, then you have lots of undesirable harmonics. It may be the case, however, that later filtering stages are going to attenuate all of these harmonics and they are not problematic yet.
In the absence of a particular problem or pictures of the waveforms, the best advice I think we could give you is:
* check for frequency stability (jitter) using a method which can make it very obvious: a frequency counter, a spectrum analyzer, etc. If you only see a few Hz change on a 10s of MHz signal, then it's pretty stable
* in general, check the signal with a spectrum analyzer or an oscilloscope with FFT function. As long as the frequency stability is fine (no jitter), then you want to reduce other spurs and harmonics as much as possible (again, assuming they are not going to be filtered out)
If you think about this, this (filtering) is why harmonics are, in general, easier to deal with than other sources of spurs (like intermodulation distortion). If you have a signal at 10 MHz, then it's pretty easy to make a filter which will get rid of the 20, 30, 40 MHz, etc, harmonics. But if you also have a crystal oscillator nearby which is at
10.230 MHz, then that is going to be much, much harder to filter out - you will need a filter with many elements and it will need to have a very sharp dropoff (to allow the 10.000 MHz signal through, but stop the very nearby 10.230 MHz signal).
Home
Help
Search
Login
Register
