# Using a crystal to regulate a PIC error

#### camerart

Joined Feb 25, 2013
3,283
Hi,
I use crystals to regulate PICs.
If the crystal has 'say' a possible error of 20PPM, how many seconds/milliseconds could it error in 1Hr.
Camerart.

#### Irving

Joined Jan 30, 2016
3,190
20ppm = 20 parts per million or 1 second in 50000 = 1sec in 13h 53m 20sec or 0.072sec in 1hr

#### MrChips

Joined Oct 2, 2009
27,679
There are 3600 seconds in 1 hour, or 3,600,000 milliseconds in 1 hour.

Thus 1 ppm is equivalent to 3.6 milliseconds in 1 hour
20 ppm is 72 milliseconds in 1 hour.

btw, crystals can be trimmed by adjusting the load capacitors on the crystal.
Watch crystals are designed to have a resonant frequency slightly higher than the nominal 32768Hz frequency. Adding trim capacitors will allow one to bring the frequency down closer match the desired frequency.

#### camerart

Joined Feb 25, 2013
3,283
There are 3600 seconds in 1 hour, or 3,600,000 milliseconds in 1 hour.

Thus 1 ppm is equivalent to 3.6 milliseconds in 1 hour
20 ppm is 72 milliseconds in 1 hour.
Hi Mc and I,
This means that, what I am trying to figure out, which is a GPS synchronisation, is feasable, if needed.
I've been trying to work it out for hours
Thanks,
C.

Last edited:

#### jaack

Joined Sep 13, 2013
3
Many times the ppm specs for quartz crystals for oscillator control vary with their temperature and should be part of the spec. Maybe you have an incomplete spec. i.e, 20 ppm/degree Cent.
When the oscillator circuitry heats up, the crystal's frequency varies. To get a precision oscillator, the crystal will put into a tiny oven or a precision-isolated heat source. The heat keeps a constant temperature of the crystal's structure ensuring stable frequency.
See this link for more details https://www.everythingrf.com/community/tcxo-vs-ocxo

#### camerart

Joined Feb 25, 2013
3,283
Many times the ppm specs for quartz crystals for oscillator control vary with their temperature and should be part of the spec. Maybe you have an incomplete spec. i.e, 20 ppm/degree Cent.
When the oscillator circuitry heats up, the crystal's frequency varies. To get a precision oscillator, the crystal will put into a tiny oven or a precision-isolated heat source. The heat keeps a constant temperature of the crystal's structure ensuring stable frequency.
See this link for more details https://www.everythingrf.com/community/tcxo-vs-ocxo
Hi J,
I have 2x options, for the project I'm working on, and needed a ball park figure, to help me compare both.
Thanks for the temp info, but it pobably won't be needed, as I've chosen the 'other' option to try first, but could use 10PPM, crystals or a temp controller.
Thanks, C.