Go Down

Topic: resonator vs crystal (Read 5146 times) previous topic - next topic

db2db


Is there any kind of gauge how much better timing is with a crystal compared to using a ceramic resonator?

oric_dan

Yes, look at the datasheets. Ceramic resonators have 0.5% tolerance, while
crystals are 20 - 50 PPM.

FYI ---->  *always* look at the datasheets.

db2db

#2
Feb 21, 2013, 08:14 am Last Edit: Feb 21, 2013, 08:22 am by db2db Reason: 1
So at 50ppm it's about 0.005% for a crystal, and .5% for a resonator.

The real question though - does this manifest as jitter, or just a constant variation from the spec?  
That really affects the answer for some apps.

(Can anyone list what crystal and caps to use? A Digikey/Mouser part number would be even nicer!)

dc42


So at 50ppm it's about 0.005% for a crystal, and .5% for a resonator.

The real question though - does this manifest as jitter, or just a constant variation from the spec?  


I don't think you'll see much (if any) jitter with a ceramic resonator, but you will probably see a greater variation of frequency with temperature than with a crystal. The datasheets should tell you.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

majenko



So at 50ppm it's about 0.005% for a crystal, and .5% for a resonator.

The real question though - does this manifest as jitter, or just a constant variation from the spec?  


I don't think you'll see much (if any) jitter with a ceramic resonator, but you will probably see a greater variation of frequency with temperature than with a crystal. The datasheets should tell you.

I don't know about ceramic resonators, but with ceramic capacitors you can get quite big variations in capacitance just from sound.  A friend created a simple Colpitts oscillator using cheap ceramics once, and by shouting at the capacitors the frequency would change quite noticeably.  Quite an eye opener.

I *always* use crystals when decent timing stability is needed.

afremont

I had an interesting experience with my Uno R3.  I keep hearing that it uses a resonator, and there truly is a tiny tiny metal can on the board.  There is also a 16Mhz xtal for the USB chip.

Anyway, I ported a PIC program to the Uno that I did some years ago.  It uses Timer1 Capture facility to time incoming pulse spacing.  I use it to analyze and adjust old clocks and watches.

I got the program working and was using it to time the 1Hz square wave output of a Dallas 1307 RTC.  Depending upon room temps, I consistently measure from 999,965 uS to 999,988uS.  I assumed that the Uno had a crystal on board.  Imagine my surprise to find that a resonator was performing within specs for a crystal.  If I put my finger on the resonator to heat it, my numbers start to climb faster than a xtal would deviate.

I'm wondering if this is just a fluke or if resonators can be expected to perform like this.  I bought my Uno for $15 at microcenter, it's made in Italy.  I plan to pick up another just to run this same test on it.

To summarize, the resonator on my Uno is well within xtal specs as far as room temp accuracy.  How can a device speced for 5000ppm accuracy be achieving 50ppm accuracy?

Since I measured the output of the RTC, I'm not measuring my own resonator against itself, but against a 32.768kHz clock crystal.  I'm impressed if this is what can be expected from resonators now.
Experience, it's what you get when you were expecting something else.

retrolefty

#6
Feb 21, 2013, 02:16 pm Last Edit: Feb 21, 2013, 02:18 pm by retrolefty Reason: 1
Quote
To summarize, the resonator on my Uno is well within xtal specs as far as room temp accuracy.  How can a device speced for 5000ppm accuracy be achieving 50ppm accuracy?


Because such specifications describe a range of possible frequency variation for the device and still be within the manufactures ratings for the device, not the specific frequency of a specific example at a specific temperature that you end up with.

PS: Real Arduino Uno boards don't retail for $15, you own a clone disguised as a real Uno board.

Lefty

Grumpy_Mike

Quote
I bought my Uno for $15 at microcenter, it's made in Italy. 

That sounds like a fake Arduino to me, I suspect it wasn't.

afremont

I bought it a microcenter, they have a slew of them in stock in houston.  They are on sale for 14.99, check it out.  They are R3 and work just like they are supposed to.  I really doubt that a fake would perform better than the real thing.

http://www.microcenter.com/product/392614/Arduino_Uno_Rev_3

Experience, it's what you get when you were expecting something else.

retrolefty


I bought it a microcenter, they have a slew of them in stock in houston.  They are on sale for 14.99, check it out.  They are R3 and work just like they are supposed to.  I really doubt that a fake would perform better than the real thing.

http://www.microcenter.com/product/392614/Arduino_Uno_Rev_3




I stand by my statement. Real Uno's don't even wholesale for $15, so microcenter can't make it up in volume sales. Not an issue if a clone is better or not then the real deal (most work just fine), it's a matter of clones using arduino trademark in violation. The Asians have been cloning arduino boards from the beginning. Follow the money is usually the best indicator.

Lefty

afremont

Did you even look at the link?  It's obviously a loss leader.  They have scores if not hundreds of them and they also have tons of parallax, maker shed and other arduino things.  If it's fake, then it's the best one ever.  Complete with white print on the back of the board.
Experience, it's what you get when you were expecting something else.

Grumpy_Mike

Quote
I really doubt that a fake would perform better than the real thing.

Why not all we are talking about is branding, the parts would be the same.

Quote
It's obviously a loss leader.

Funny that only normally works when you sell other stuff as well.

Quote
If it's fake, then it's the best one ever

Might be, is the on LED green or have they sorted that out as well?

afremont

#12
Feb 22, 2013, 03:10 pm Last Edit: Feb 22, 2013, 03:14 pm by afremont Reason: 1
They do sell tons of other stuff grumpymike, I don't know what you are talking about.  I spent some time yesterday afternoon looking at fakes and how to identify them.  My board came in a blue/white box with the seal.  It came with the six stickers and the thank you pamphlet.  It has a green power led and yellow rx, tx, and Pin 13 LEDs.  It says "made in italy" and has the trademark symbol on the back side.  It has a red reset button and everything else that I could check.  It's simply not a fake, but if you all would rather pay twice as much, then forget I said anything.

EDIT:  Microcenter is a huge outfit that has been in the computer parts retail business for over twenty years.  They now sell tons of maker, parallax and arduino doo-dads.  I doubt they would throw it all away to sell some fake Unos.  I invite anyone to actually prove that these are fakes instead of just making comments.  Show me one thing that is wrong with them.
Experience, it's what you get when you were expecting something else.

Papa G

The ad I saw online lists them as Gheo Electronics Arduino Unos.

retrolefty

#14
Feb 22, 2013, 03:28 pm Last Edit: Feb 22, 2013, 03:35 pm by retrolefty Reason: 1
Quote
EDIT:  Microcenter is a huge outfit that has been in the computer parts retail business for over twenty years.  They now sell tons of maker, parallax and arduino doo-dads.  I doubt they would throw it all away to sell some fake Unos.  I invite anyone to actually prove that these are fakes instead of just making comments.  Show me one thing that is wrong with them.


I would suspect that the only 'proof' could come from the Arduino folks, if they sell wholesale to Microcenter or not? I don't see them listed as official distributors of arduino products. Again it's just my opinion, but any new $15 Uno R3 board is a clone. We have seen many clone arduino boards selling on Amazon.com, are they any smaller or hold less risk aversion then Microcenter?

Quote

USA: SparkFun, Maker Store, Adafruit Industries, Little Bird Electronics, Modern Device, FunGizmos, NKC Electronics, Gravitech, RobotShop, Liquidware, Hacktronics, MakerBot Industries, Microcontroller Pros, Curious Inventor, AeroQuad, CuteDigi, EIO, Teach me to make, UltiMachine, Electrojoystick.com, Electronics is Fun, AME - After Midnight Engineering, Trossen Robotics, Jameco, Zagros Robotics, Advanced Micro Circuits Corp, iHeartEngineering, 3D Robotics, Jaycon Systems, Elexp, Abra-electronics, Reuseum, Pololu, 411 Technology Systems, Handmadecircuits, ManyLabs, Epictinker, Woodman Studios, LLC, oddWires, Siliphi LLC, Karlsson Robotic, RadioShack, Banana Robotics, SuperDroid Robots


There are some real small retailers in that list, why would fail to list a "huge outit" like Microcenter?

http://arduino.cc/en/Main/Buy

Lefty

Go Up