phyphox Forums

...and this should also be fixed in the beta 1.1.7-beta3 that I just published.

Bonjour Sebastian,

all works perfectly well with 1.1.7-beta3. No peak anymore on the gausssmooth curve (with sigma=20), see attached figure. But a new problem has appeared: phyphox does not want to register new programs from e-mail...

Mikhail

(11-27-2020, 12:41 PM)solid Wrote: [ -> ]Bonjour Sebastian,

all works perfectly well with 1.1.7-beta3. No peak anymore on the gausssmooth curve (with sigma=20), see attached figure. But a new problem has appeared: phyphox does not want to register new programs from e-mail...

Mikhail

Sorry to hear that, but that's what Beta versions are for and I am very happy that you test and report every problem. Unfortunately, I could not reproduce it with gmail (tried phyphox-files, zips and phyphox-links). Could you give a few more details? Is it a phyphox-file or a zip-file? Which email app did you use and what happens if you try to open the file? I assume it is an attachment?

It is OK, I am sorry for warning but maybe it is file specific. Look the discussion in the requested options... With a new rectified file (attached here) there is no problem.

In previous examples of phyphox-files the version was not indicated and the experiment could be used only by Android smartphones, sorry.. Here is a corrected version of the pressure measurements using SensorTag of Texas Instruments with data smoothing by Gausssmooth and LOESS. Also the average of last 20 points of the raw pressure is given (the latter two additions were proposed by Sebastian). It works well for our students laboratory experiment of Clément and Desormes (determination of Cp/Cv of air).
[attachment=363]

Here is a compact version without smoothing and without temperature reading. Only average of last 20 point has remained for a more stable reading of pressure.

Good evening.

Another problem has appeared. It is not related to smoothing of the measurements of pressure using SensorTag, but to the value read by iphones. Everything works perfectly well for two last experiments given above with Android phones. An iphone accept the experiment and lunch it but the values are too high as seen on the attached screen capture ! The pressure value should be around 1000 Pa ...

What it could be and how to solve the problem? Unfortunately I do not have an iphone myself ...

Mikhail

Damn, someone noticed before we could fix it. I am preparing an update right now that fixes this problem, which should be available in a few days.

The problem was that the binary conversion for 24 bit integers was a bit mangled up after the last migration to Swift 4. So, all values from the sensors with 24bit data are messed up in the iPhone, which to my knowledge only affects the pressure sensor on the SensorTag.

Thank you, Sebastian!
Poor students with iphones feel not so happy now ...

On my iDevices it has been exactly a factor of 256 by the way…