Identifying the source of closely-spaced harmonics
Kenneth Wyatt- July 26, 2012When troubleshooting a new product, one of the first things you should do is to list all the crystal oscillators, along with a listing of their harmonics. This will help you identify the source for a particular harmonic which might lie above the radiated emissions limit. But what if there are two, or more, crystal oscillators that create the same harmonic? For example, if there are 20, 30 and 60 MHz oscillators in the product, which one(s) might be the source of the 120 MHz harmonic?
One clue that you may have multiple sources is by looking carefully at the harmonic itself. For example, you may see the following peak with what appears to be a ripple on top. This either means the harmonic is frequency-modulated (not common) or that you're really seeing two separate harmonics within the same resolution bandwidth.
If you were narrow down the resolution bandwidth, you'd actually observe two very closely-spaced harmonics.
We can also see that one of these harmonics is about 5 dB higher than the other and reducing the dominant one could potentially be the difference between passing and failing the radiated emissions limit at that frequency. So, how would you determine the source of the dominant harmonic?
There are a couple troubleshooting techniques I've used - loading the oscillator output pin with a lead pencil tip (or simply touching it with my finger), or using "freeze spray" on each oscillator package. Either method will slightly cause the associated harmonic to move slightly. From this simple test, you'll be able to determine the dominant source and deal with reducing the resulting harmonics by rolling off the edge rate a bit with a low-pass filter.