Hi I am still to buy my arduino not really sure which one to purchase, The project I am going to be working towards needs to have an accuracy to within 50 ppm 0.005%
I know all processors have some inaccuracy and an external source would be best but to keep things simple if possible I would prefer to just use an Arduino which meets my requirements.
I am struggling to find any details on line of the accuracy of the clocks on the various arduino's can anyone help me purchase the right one. I know all
Change your post title first. The title should describe the nature of your questions. Your post title won't get any results.
Try changing it to "Need help with RTC /ARDUINO APP"
If the forum won't let you change the post title then use the REMOVE button in the upper right corner of the post window to delete this post and create a new one with the correct title.
You are more likely to get help if you give your post a suitable title - did you read the sticky at the top of the forums?
What do you mean by 'accurately'? All the Arduinos are dependent on the timing source for their accuracy, so there will (nearly) always be some degree of inaccuracy.
The Uno at least uses a 'ceramic resonator' as opposed to a crystal. I went to my favourite electronics supplier and looked up the spec for a similar part:
Frequency tolerance ±0.5%
Frequency stability ±0.3%
Ageing ±0.3% over 10 years
dannable:
The Uno at least uses a 'ceramic resonator' as opposed to a crystal. I went to my favourite electronics supplier and looked up the spec for a similar part:
Frequency tolerance ±0.5%
Frequency stability ±0.3%
Ageing ±0.3% over 10 years
The spec sheet for the Arduino Mega says +/-10% on page 371
dannable:
The Uno at least uses a 'ceramic resonator' as opposed to a crystal. I went to my favourite electronics supplier and looked up the spec for a similar part:
Frequency tolerance ±0.5%
Frequency stability ±0.3%
Ageing ±0.3% over 10 years
Then the Arduino is probably not what you need, at least as your measurement device. Its not something that can be improved without spinning your own PCB.
What is your project? You say you need 0.005% accuracy on timing, what time scale is this for?
Am looking at building a timing system for flyball; think dog relay racing; very fast most races are sub 25 seconds on the whole; this is from crufts this year.Crufts 2014 | Day 1 LIVE - YouTube
the 0.005% comes from the association which oversees the sport in the UK to have the system approved by them it needs to equal or be below this.
An Arduino is fine for this. Unless you need microsecond precision. In which case it isn't.
Have a way to detect the start and stop (A switch, etc), and just record the millis() value when each part gets triggered. That should give you plenty of accuracy.
mirith:
Keep in mind, EE's think of milliseconds as slow.
EE's ??
mirith:
An Arduino is fine for this. Unless you need microsecond precision. In which case it isn't.
Need it to be within 3 milliseconds.
mirith:
Have a way to detect the start and stop (A switch, etc), and just record the millis() value when each part gets triggered. That should give you plenty of accuracy.
Yes am going to use photo-reflective sensors to trip (lap) the time when the dogs cross the line
10% is 10% if its 20 seconds its 2 seconds or if its 20000millis its 2000millis or am I thinking about this all wrong?
I need to prove the accuracy the easiest way would be with documentation but thats showing 10%
Arduino comes with no calibration data/certification. You'll need to find something or someone to do that.
I would possibly look for really accurate RTCs though. They will give you a bit better accuracy than an arduino on its own.
I assume more than one of these devices will exist? So, just eliminate the issue, build your own "boarduinio" and use a 30ppm or 20 ppm crystals instead if the stinking resonator used. You develop on an Arduino and deploy on custom hardware. The manufacturer's data sheet will clearly show the freq drift against temperature.
But, that is not the whole story... Your code will have latency in the start and the finish. So, you must ensure that the code utilized for start and finish creates a (near) identical delay and that no conditional or interrupt routines interfere with this code.