LFTVXO075802Cutt TCVCXO Oscillator 10.0MHz .28ppm 3.3V -40C+85C= GPSDO Freq Ref?

[t1d]

Re:  IQD Frequency Products LFTVXO075802Cutt TCVCXO Oscillator
https://www.mouser.com/ProductDetail/449-LFTVXO075802CUTT
Data Sheet = https://www.mouser.com/datasheet/2/741/LFTVXO075802Cutt-1115688.pdf
Caution - Noob At Work

I have been considering adding a PLL circuit to a Chinese U-Blox NEO-M8N to improve its performance as a 10MHz GPSDO Frequency Standard. The typical scenario for procuring  the oscillator seems to be to buy a used oven unit on Ebay. I haven't been able to warm up to that idea (Is there a joke here? Oven & warm up to the idea...), so I have been researching alternatives.

I came across the LFTVXO075802Cutt and, from my noob prospective, I thought it to have possibilities. I had hope to find that it was being used widely, for this purpose, and that there might be DIY circuit schematics available. However, I only found one individual offering a 35MHx-4400MHz synthesizer with its controller unit employing the LFTVXO075802Cutt. https://www.sv1afn.com/adf4351m.html

So, I am asking here what you gurus know of it and what your opinion might be for using it as I intend. Please and thank you for your help.

[mark03]

This is really above my pay grade, but since no one else has stepped in, I'll offer some thoughts anyway.  If you compare your TCXO datasheet with the data for a well regarded OCXO (Oscilloquartz 8663, http://www.sungwhatech.com/product/pdf/aging/8663.pdf), there are a couple of really significant differences (plus others which may or may not be significant depending on your application):

1) The "short-term stability" (the stability over tens of seconds) of the OCXO is one part in 10^-11.  The IQD datasheet doesn't specify the TCXO's short-term stability but it is probably much worse.  Intuitively, this is needed to smooth out the jitter in the GPS output.  Of course, the TCXO is probably still better than the GPS in this regard, but with an OCXO you can do better still, by like two orders of magnitude.  If you plotted the Allan variance of your proposed TCXO GPSDO, it would have a huge lump at short periods.

2) The tuning range of the OCXO is about 30 times smaller than that of the TCXO; when you account for the voltage ranges, the actual tuning coefficient (ppm per volt) is more like 100 times smaller.  This means the TCXO would be ~ 100 times more sensitive to any noise on the voltage control line.

What is your concern with the OCXO?  Space?  Power?  Shady ebay dealers?  w.r.t. space and power, there are emerging TCXO technologies which can rival traditional OCXOs, but they are expensive.  See the SiTime "super TCXO" series.  Phase noise is probably still inferior to a good OCXO, if you care about that.  I expect this field to continue to evolve until quartz is finally displaced by cheap, miniature atomic standards... but that is a long ways off yet.

[t1d]

Thanks, Mark. The technical details are beyond my knowledge, but I can clearly understand that there are important differences. Good job.

Personal Used Oven Issues
1) Used can equal stressed and that can be unrecognizable in simple does-it-power-up tests. The donor equipment was taken out of service for a reason; who knows why?
2) It seems to be that you grab from whatever is currently available on Ebay... There are lots of substitutions. I don't know what is acceptable and what isn't.
3) I read lots of stories about folks that ended up with the wrong oven, because of some minor difference in the model number that equaled a major difference in usability.

The answer to these issues is to buy from a dealer that has a good reputation and to pay the extra required to not buy on Ebay. So, I need to research four things... Typical makes and models that are used, reputable dealers, cost and a circuit design (I think I understand what is needed in the circuit and why. I just need to settle on one.) I'll take any tips, or suggestions, that anyone has to offer on those.

Another thing to consider is that I am a hobbyist. I really can't justify a major expense for this piece of equipment. But, it surely has my interest as a fun DIY challenge.

And, if anyone has an extra that they would like to swap/sell, I would be glad to know of it. As money is limited, maybe a couple of PIC18F4550 controller boards would be useful to someone... I have two unique designs... One is made to mate with a bread board. The other has multiple pin sockets per MCU pin, for easy plug-in access.

Thanks, Mark, for your help.

[awallin]

this picture from SRS shows what you can expect from a GPSDO with different quality/price LO:
https://www.thinksrs.com/images/instr/fs740/FS740_lockedLG.gif

- for short-term (1-1000s) stability use an OCXO
- if the interesting science/hobby/whatever you are doing with your gpsdo is on time-scales of hours and days then it doesn't matter much what LO you have.

I've used connor-winfield OH300, new from digikey, in one project with at least moderate success. Probably not as good as a used Morion MV89 from ebay (if you are lucky and get a good one), but close.
The OH300 is under 100eur IIRC.

[t1d]

Thanks, awallin...

My use is just for calibrating some vintage equipment that is not even high end stuff... An oscilloscope, frequency counter and generator... Well, I do have some nicer new stuff that will eventually need to be checked, too.

I would not be leaving it on for extended periods, unless I use it double-duty for a super-cool GPS sourced clock...

I will continue to poke around to gather information and make a decision on need vs. budget. Thanks for the help.