Crystal Aging

[Robert Reed]

Can anyone answer this for me? In regards to quartz crystal specifications, the term crystal aging is always listed. It is usually given in PPM per year or month. This is the natural drift error over time even when all other parameters are held constant.As often as I have seen this spec., it has never been clear to me if aging time is accumulated operating time or all time since point of manufacture, even if it was never powered up.As an example: xtal "A" operates for 10 years nonstop with an aging rate of 1PPM/year.
Xtal "B" (also specked at 1PPM/year) manufactured at the same time sits in its original shipping package for 10 years. Is there a difference in aging error between them?
The only thing I kmow for sure is that aging rate reduces
with each succeeding year.

[Chris Smith]

Decay is a natural process, having nothing to do with use. In radio active materials its called half life decay, while crystal structures lose their frequency [Or stability ] slightly over time for the same basic reasons of decay. I believe the crystal even shrinks its mass over time by a few atoms at a time like all materials in the universe. I don’t believe use accelerates this process by much, if any?

[jwax]

Never had to consider it before, but here's a quote from http://www.electronicproducts.com/ShowP ... a1.oct2002<p>"Crystal aging is the second major contributing factor to frequency instability. Its effects can be limited in several ways depending on the expected useful life of the target system. Since frequency change due to crystal aging is a logarithmic function, the most significant effect occurs during the first year of operation."

[Robert Reed]

Seven years ago I purchased three peices of test equipment. These are hi-end equipment (HP and SYNTONIC) all with very accurate time bases (10 Mhz).They were used commercially for thousands of hours over a ten tear period before I purchased them.These time bases have specked accuracies into the eighth order of magnitude. At the time of purchase I had access to a rubidium standard with an incredible accuracy approaching the eleventh order of magnitude (also recently factory calibrated)and these units were calibrated to that standard. These units now see about 40 hours of use each year, and I check calibration each year.The problem is I do not have access to a "standard" that exceeds the accurracies of these units. So I compare one unit against the other two in a 'round robin fashion'. With each yearly check they show less than 1 Hz difference between them (<0.1 PPM). So I am assuming one of two things is happening. First-all xtals are aging at exactly the same rate and show no difference between them (very unlikely) or second-the aging rate is almost at a standstill.This is what prompted me to post this question. It seems that highly reduced time of operation vs. real time has had no effect on actual operating frquency. Maybe I am already so far into the log curve of aging rate that changes are not as obvious?Two of these units are OCXO's (oven) and one is TCXO. Also I realize that stating accuracy without stating stability is meaning less ,therefore these test are always made under identical conditions, leaving xtal aging as the only variable.

[Chris Smith]

The only other effect on a crystal and its accuracy that I recall is stress. I have heard that over stressing a crystal can screw it up, and its accuracy. <p>But other than that I believe its purely decay, because most crystals have the same structure as radio active materials, while some crystals are slightly radio active in them self.

[Ron H]

Check outthis paper: "THE AGING OF BULK ACOUSTIC WAVE RESONATORS, FILTERS AND OSCILLATORS".

[Robert Reed]

I had remembered a former associate I woked with in the past, whose expertise was in xtal ocssillators. I E-Mailed him and he came back with a ton of info and some in depth web sites on this subject.One of these sites is the one you posted RonH (thank you). After reviewing friends info and web sites,the answer to my question boils down to this (in a nutshell) : Aging has basically 4 causes:
Releif of mechanicl stress
impurities at time of manufacture
contamination by seal leakage
loss of material such as flakeoff<p>Stress relief is caused by the actual quartz element holder relaxing its grip over time.Impuraties cause slight chemical reactions as so do leaking case seals.Loss of material is a natural physical phenomena over time.These properties are extremely minute, but so are the frequency changes we are dealing with . Aging takes place even if the crystal is idle but this process is greatly accelerated when it is active and the faster rates of change occur within the first 30 days of operation. As you might suspect, a complex subject involves a lot more detail in answer than I have typed up here,but since your replies indicated an interest and knowledge of this post, I thought I would close it with the pertinant info I found on my original posting