Measuring battery life of Bluetooth headphone AMP

[LeoTech]

Hi,

I want to measure the realistic battery life of a Bluetooth headphone AMP - it is an FiiO BTR1K, if anybody is interested - with a specific headset connected. The problem with measuring battery life is obviously the fact, that it will highly depend on the load, i.e. the headset.

My first idea was to simply charge up the battery, connect a headset and play some music or just a tone, but I don't really like the idea of me having to listen to it for several hours straight to know when to stop the time. Instead I would connect a high impedance probe - like one from an oscilloscope - maybe amplify the signal and then connect it to an arduino to record the signal and measure the time, when the signal is gone, the battery would be dead and the timer would be stopped.

This should work, I am however concerned, that the probe will have an impact on the load, so my question is, is this a reasonable idea? Will the extra load presented by the probe have a major impact on the battery lfie?

Thanks in advance, and if any of you have another idea of how to accomplish this, please let me know.

Leo

[amyk]

An oscilloscope has 10M of input impedance when used with a probe in 10x mode. Considering that headphones have impedances in the low hundreds of ohms at most, an extra 10M in parallel will have no discernable effect.

[magic]

For a one-off job?

Record with a mic and load into Audacity to see where the signal ends

[LeoTech]

An oscilloscope has 10M of input impedance when used with a probe in 10x mode. Considering that headphones have impedances in the low hundreds of ohms at most, an extra 10M in parallel will have no discernable effect.

Yeah, I should have realized that myself, thanks for pointing it out!¨

For a one-off job?

Record with a mic and load into Audacity to see where the signal ends

Good idea, I just checked out Audacity and if I were to turn down the sample rate it wouldn't even require a lot of data. The problem, however, arises in the fact that the audio output of the connected headset will be really small and therefore not really detectable and discernible from other sounds by a normal laptop mic - I don't really have anything else. In other words, this might not work reliable and I would like a proper solution for more than a one-off job.

Thanks for the suggestion however, as I could simply connect the oscilloscope probe to a sound card and still do the signal analysis in audacity, that would get rid of the arduino and the problems with using a mic.

Any thougts?

Leo

If I were to do this for several jobs and repeatable, what would you recommend?

[magic]

See if Audacity has an option to automatically identify and remove silence at the end of file

Otherwise, I can offer a humble remark that anything that can be done on Arduino can also be done on a PC. Probably a simple algorithm would do which scans the file looking for signal peaks and proclaims the battery dead if signal amplitude stays very low for more than a few seconds or whatever, the details would depend on what kind of music you test with.

PC has enough storage capacity to simply record the whole signal, which may be helpful if your analysis software turns out to be buggy
Lossy compression could further reduce file size to almost nothing.