Android test results

[GoogleCodeExporter]

I have my hijack board working on iOS with the example apps and it works well. 
I've run some initial tests on Android (just sampling mic input) and the shape 
of the manchester-encoded signal varies widely across the different phone 
models I've tried. Any ideas why and what could be done to improve this? 

At the moment, I don't see how data can easily be decoded on some Android 
devices. It seems a long way from the clean square wave on iOS.

Attached is a ZIP with some examples. In each image, the sampled mic input is 
shown in the top channel.

Original issue reported on code.google.com by siwat...@gmail.com on 19 Jul 2013 at 11:41

Attachments:

• [Android Tests.zip](https://storage.googleapis.com/google-code-attachments/hijack-main/issue-13/comment-0/Android Tests.zip)

[GoogleCodeExporter]

HI guys,
Now I am working on a prototype project that use  this hijack base android 
platform.I have searched many docs and can't find some demo project.So would 
you please send the "Android Tests.zip" file to my email "joyopeng@gmail.com".

Many thanks.

Original comment by joyopeng@gmail.com on 12 Aug 2013 at 3:02

[GoogleCodeExporter]

> I've run some initial tests on Android (just sampling mic input) and the 
shape of the manchester-encoded signal varies widely across the different phone 
models I've tried

Perhaps this is because the amount of power provided through the jack is 
different depending on the model?  

I'm also interested in using this for Android, in particular I want to build a 
temperature sensor board to plug into a HiJack for building affordable open 
source temperature alarms for greenhouses.

- R.J. Steinert http://github.com/rjsteinert

Original comment by r...@rjsteinert.com on 22 May 2014 at 7:38

[GoogleCodeExporter]

Here's an Android example from Andrew Robinson 
https://github.com/ab500/hijack-infinity

He notes that not all Android devices provide enough power.

> Power draw still remains high, this device will only be compatible with a 
select number of Android devices. For all other devices you will have to supply 
power externally. Devices that have Beats by Dr. Dre have been found to work 
well with HiJack, the extra amplification provides enough energy to drive the 
board. Supply HiJack with 2V of regulated power and disable the portion of 
Android code that generates the 22kHz tone for power and this code should work 
with many more phones.

This makes me wonder if a 2v battery mount would be appropriate for this design 
to overcome this power variability challenge.

Original comment by r...@rjsteinert.com on 22 May 2014 at 10:32

[GoogleCodeExporter]

The Thermodo project ran into a problem with different results from each device 
that they fixed by looking at the relative difference between the two channels. 
 Not sure exactly what they are referring to and their hardware design is 
probably very different but it might be a lead.

See 2:22 in this video -> http://vimeo.com/75071515

Original comment by r...@rjsteinert.com on 23 May 2014 at 12:30