Printable View
I came across the angle gauge feature on my Samsung A71 android phone and wondered how good it is. So I set up an experiment.
Attachment 38767
The Wixey angle gauge =+/-0.22° and the Samsung A71 = +/-0.12°
The plot of the phone and angle gauge versus the sine bar shows both are rather good in fact
Attachment 38768
So I am assuming that the sine bar was your "known good" value?
From your graph it appears that the error starts out very low at small angles and becomes greater as you approach 60°. But it is hard to tell just how large that error is. A graph of the error with a magnified vertical scale would be more informative.
Hi Paul, yes, the sine bar was the known angle. You are correct that the errors increase as the angle gets towards 60°. Good idea for the plot and have added it below:Quote:
Originally Posted by Paul Alciatore
Attachment 38814
Best regards, Nigel
I need to measure some other phones eg iPhone 12 would be interesting
The error plot concerns me as both devices swap from a +ve error to a negative error around the same absolute angle - I think I should set this up again and do a lot more measurements around this angle to see what is happening.
First off, remove the snap on case from the phone so the phone chassis is in direct contact with the sine bar.
This concerned me at first, but the case fits tight on the phone and doesn't rock. Therefore, once you calibrate for zero, any effect of the case is removed.Quote:
Originally Posted by Unkle Fuzzy
Actually, a case that had a square cross-section that would allow the phone to balance on it's edge would be an advantage in this application.
It's a variable.
Take it or leave it.
Yes, that plot of the errors is interesting and more informative. It shows that with the phone and the Wixey angle gauge
https://www.rockler.com/wixey-digita...13bcbf835ec14f
the first thing that jumps out to me is the fact that they both have errors on the positive side for angles less than or equal to 30° and both have negative errors for angles that are greater than that. This suggests two things to me to me. Please notice I said "suggests". First, there may be a common element in this. Two devices would definitely have different circuit boards and different outer cases but perhaps those elements are not in the mix. The other thing that it suggests is that there may be two different processes at work here: one being more predominant at lower angles and the other at higher angles.
It could be the sensor itself that is at work here and the same sensor may be used in both of them. Another possibility is that most electronic parts will have manufacturer's recommended or suggested circuits. It could be that the two devices both use the manufacture's suggested circuit and that circuit is responsible for one component of the errors.
A third interesting point is the exact matching of the first four readings from the zero point. And in this small range, the error seems to be going in a decreasing direction. But then the other readings under 30°, with only one seem to increase from the low value of the fourth of these four angles. Could this be another point of interest at about 3° or 4°?
All in all, the errors seem to be within a range of +0.25° to - 0.35° and that is good for many purposes. We just need to keep in mind what they may be and use better methods when more accuracy is needed.
Paul A.
Quote:
Originally Posted by editor@glue-it.com
I'll trust my wooden measuring stick as my angle gauge.
