Hi All,

For stability it is enough to place the dongle in the
middle of a big bucket filled with sand.

Look at figure 3 here:
http://www.sm5bsz.com/linuxdsp/hware/mirisdr/thermal.htm
With 10 litres of sand you can expect the frequency to
stay well within 3 Hz at 144 MHz at room temperature.

There is a sweet point where the temperature coefficient
goes through zero. One can put a thermometer in the bucket 
and a heat source on top of it to find the sweet point
and keep the temperature near for even better stability.

The tests were done on the Mirics chip, but I guess the
crystal oscillator is similar in all the dongles except for the
FUN Cube Pro+ which has temperature control.

Note that the supply voltage affects the frequency quite a lot.
That is true even for the FUN Cube Pro+. It may be a good idea to
supply the dongles with 5 V from an extra 5V supply that
is very well regulated.

Since the dongle is small we can temperature stabilize the 
whole thing, not just the crystal:-) That way we get both stability
and reference accuracy. Note that changing the state of the
dongle (current consumption) can cause a small frequency shift
with this arrangement.

Regards

Leif / SM5BSZ



> To achieve stability and reference accuracy, the time base crystal must be
> removed and placed in a temperature controlled enclosure, or provided by a
> commercially obtained TCXO.
> 
> http://www.qsl.net/m0ayf/Crystal-Ovens.html
> 
> http://www.crystek.com/documents/appnotes/Understanding_TCXOs.pdf
> 
> 
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Dan
> Sent: Monday, March 31, 2014 9:58 PM
> To: [email protected]
> Subject: TCXO other than 28.8Mhz
> 
> Hi,
> 
> I need rtl-sdr TCXO with 1ppm. I need replace 28.8MHz at someone with 1ppm.
> My sample rate is 2.048MHz. 28.8MHz with 1ppm is hard to find at digikey or
> mouser. I would like to know is anyone has success replace TCXO other than
> 28.8MHz
> 
> Thanks
> Dan 
> 
> Sent from my iPad
> -----
> No virus found in this message.
> Checked by AVG - www.avg.com
> Version: 2013.0.3465 / Virus Database: 3722/7286 - Release Date: 04/02/14
> 
> 

Reply via email to