[time-nuts] statistical distribution of initial frequency error in tcxos

[Bob kb8tq]

Hi

Ok, a bit of nomenclature: 

TCXO = temperature compensated crystal oscillator = it has a compensation network in it
XO = crystal oscillator = it does not have a compensation network in it

Why does that matter in this case? They are different products targeted at different markets and
different applications. 

A part with a +/- 50 ppm stability spec over 0 to 70C is almost certainly an XO. It might come in a 
version that is +/- 50 ppm “all inclusive for X year(s)”. In all likelihood it does not have an adjustment
“port” on it. 

A TCXO is mighty loose if it is +/- 2 ppm over 0 to 70C. It is highly likely to have an adjustment port 
(voltage or trimmer) on it. 

Since they are very different parts, it *does* matter more than a bit which one you are looking at. 

XO’s may get manufactured with an “offset” but that’s not very common. It is more likely that the
“plate to frequency” process has an offset in it and you get a batch all to one side or the other 
as a result. A part with a _/- 50 ppm overall spec might be plated +/- 15 ppm. (could be tighter 
or looser depending on the outfit doing the part). 

A TCXO might also have an offset, but it’s going to be a lot smaller. It’s TC likely will also be a 
lot smaller.  Most low cost TCXO’s don’t get an offset since it is more hassle to track through 
a high volume manufacturing process. They all *should* hit center frequency within a fairly tight
window. It’s not uncommon to see +/- 0.1 PPM spec’d. 

In both cases, it would be highly unusual for the parts to get sorted in any way to “pick the good ones”.
It costs more to do that than it’s worth. ( = there are other ways to get “good ones”). The line
is focused and set up with a target range of spec's in mind. If you want a better approach, you get
them off a different production line. 

So out of your list of specs:

Initial set = what you would expect your TCXO to read on your board, generally not an XO spec
Frequency after initial aging = same thing as initial set with some wiggle room added
Frequency over temperature = what happens if you run it 0 to 70C (or whatever the range is)

Long term aging = what happens over 5,10 or 20 years.

Again, it is not at all uncommon to roll all that up into a single “lifetime” spec on an XO. You 
rarely see that done on a TCXO.

Simple answer would be to get a cheap oscillator with an EFC port and feed various voltages
into it. More or less - spend the $5 on the lowest phase noise tunable oscillator you can find. 

Bob



> On Mar 13, 2019, at 11:39 AM, jimlux <jimlux at earthlink.net> wrote:
> 
> If I go out and buy 100 TCXOs with a spec of, say, 50 ppm, what does the distribution of the initial frequencies (and, I suppose, the frequencies after aging) look like.
> 
> I would think these days that the manufacturer sets the tolerance based on manufacturing performance, so they don't get too many fail outs.  As opposed to "binning and marking" (i.e. make a bunch and the ones that are within 5ppm get marked that way, and the rest are marked 10ppm or 50ppm).
> 
> There's also the "initial set on" frequency, the "frequency after initial aging" and the "frequency tolerance over temperature".
> 
> 
> 
> Here's the application:
> 100 (or 1000) independent nodes (in space, as it happens) - I want to calculate the probability that two nodes are within some delta f of each other.
> 
> For instance, if I'm buying 10MHz oscillators with a spec of 5ppm, they'll all fall in a band +/- 50 Hz.  But how many are within 1 Hz? within 0.1Hz?
> 
> In this specific application, I'd actually prefer that they all be different, but close, but I can see that going to a mfr well, I'd like 1000 oscillators, spread reasonably evenly over a 1000ppm range, but with 2ppm variation over temperature. Oh, and I'd like them to be really cheap, with no NRE, just send me a reel of them.
> 
> Oh, and I don't really care about the frequency variation with temperature (since that can be calibrated) but I'd like really good phase noise.  For, say, <$5 each in qty 1000.
> 
> 
> Think of it as a sort of FDMA/RFID without having to explicitly program it.
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.