This question was posted on e2e.ti.com:
My company have designed and are now manufacturing circuits using the CC2430 in a similar circuit to the reference design. Quantities are still relatively small (runs of 500 – 1000) but we are having great difficulty sourcing the 32MHz NX3225SA crystal at the 10/15/30/16 spec. We can get 15/15/30/8 very easily and am not too worried about the initially accuracy being 15 instead of 10. The switch to the 8pF load crystal is more of a problem. I can change C15/16 from 33/27pf to 15/15pf and so the crystal should still work but are there any other possible problems? How will this affect the RF accuracy? Will the circuit still meet the IEEE specs?
Pinter Electronics Consultants’ answered:
Chris Pinter answers, “The tolerance of the crystal is only as important as the application. What I would do is go back to the business case and determine the tolerance required of the crystal. Usually the tolerance specification is related to the operating temperature. The stability of the crystal will change over temperature. What you want to do is test your design at the extremes of your specification. Once you know that the current design is working within the specification then you can reduce the tolerance of the parts until the performance drops out of specification at the two extremes.
You will want to adjust the values of the capacitors as you change the crystal to keep the performance in specification. However, you should be able to achieve some flexibility in cost and availability of parts.
The best way to look at this is to take a step back and understand what you are actually trying to do. Is the goal to know the theoretical C Load or do you want your circuit to work?
I assume the goal is to get the circuit to work. What happens when you change the crystal on the evaluation board with the new crystal you have?
Does the frequency change? By how much?
Does the waveform change?
How about the current? Is it higher?
In many cases, trying to look at the problem in more practical terms will get you to the result faster. Once you know the circuit works, then if you really feel you must solve this calculation try to change the capacitance down to your theoretical. You will most likely find that the circuit trends down to non-functional the closer you get 0pF.”
If you want to read the full discussion click here.