Go Down

Topic: UNO upgrade to crystal oscillator (Read 1 time) previous topic - next topic

jarik

Hello,
I'm making a device which counts time, so I'm willing to make it as accurate as possible. I have bought two arduino uno copies at different times and one of them has crystal oscillator (in total two) and second one has ceramic resonator. I've tried measuring how fast the time drifts away, and its more than second every hour. Arduino with crystal drifts only about 0.030-0.040 seconds every hour. Thats more or less OK for me.
The question is can I change resonator with my own crystal? Here is a photo of my device and it looks like there are places where to connect it. Am I correct that I can desolder resonator and a small smd part, and instead of them solder crystal and two 22pF capacitors?
I've read several time in forum suggestions about this steps, but couldn't find any tutorials of photos.


jarik

I've managed to remove resonator, and soldered crystal with two capacitors, and it looks like it is working fine. Only made a short test, but looks promising (about 10 times more precise).

CrossRoads

Results like that make me glad that I use crystal/caps in all my designs.
Designing & building electrical circuits for over 25 years. Check out the ATMega1284P based Bobuino and other '328P & '1284P creations & offerings at  www.crossroadsfencing.com/BobuinoRev17.
Arduino for Teens available at Amazon.com.

tim7

A little while ago I compared the clock accuracy of an Uno with a Freetronics Eleven (Uno clone with a crystal).  The crystal was 47x more accurate and 35x more stable.  That's still not really good enough for long term timekeeping, but for applications such as interval timing and event-rate measurement it's a lot better than a ceramic resonator.

http://arduino.cc/forum/index.php/topic,69316.15.html

oric_dan

For future reference, and in case they're still available for sale, my Duemilanove board
has a real crystal.

Quote
time drifts away, and its more than second every hour.

That was actually pretty good. The tolerance on ceramic resonators is 0.5%, which means
worst case error can be as much as 0.3 sec/minute, or 18 sec/hour, or 432 sec/day.
Slightly incredible.

westfw

Be aware that the newer "R3" Uno board no longer has the dual-footprint that allows easy substitution of a crystal...
:-(

afremont


Be aware that the newer "R3" Uno board no longer has the dual-footprint that allows easy substitution of a crystal...
:-(



I wondered why people thought this was so easy; the resonator used is smaller than a chip LED.  Both of my UNO R3s have reasonable accuracy.  One is well within 50ppm at room temperature, not near as stable as a crystal with temp changes though.  The other one is within about 300ppm.  My breadboard 328p at 20MHz is within 20ppm and stays there now that I used a parallel cut crystal instead of what must have been a series cut version I was using.  It was 500ppm off.
Experience, it's what you get when you were expecting something else.

oric_dan

Quote
Both of my UNO R3s have reasonable accuracy.

In general, accuracy with resonators will probably be not-too-bad, and similar to OP's
experience, since statistically they probably fall along a Gaussian curve for error,
but that doesn't mean you don't eventually end up with one of the bad boys.

Docedison

Probably the only reason to buy a clone, They are always a year or so behind... I had to pull out my R3 Genuwine... Uno and look.

Doc
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

Go Up