Welcome Guest,Register Now
Log In

ANT Forum

Welcome guest, please Login or Register

You are here: Forum Home → ANT+ Forums → ANT+ Bike Power → Thread

   

Recommended method for sensitivity calibration?

Rank

Total Posts: 2

Joined 2013-03-12

PM

What is the recommended method for a user to enter/update the sensitivity-calibration of the bike power sensor?

As with other linear systems, two constants define everything: Offset and Sensitivity. The Offset value is dealt with using one of the "zero-calibration" techniques described in the spec.

But what about the Sensitivity constant? For a crank based system this is proportional to the length of the crank arm and some constant of the sensor system. If this calibration is done at the factory, no problem. But what if the user should be able to calibrate this? Or update a previous factory calibration?

What is the recommended method for the user to enter some calibration value to be sent from the display to the sensor device at a given time?      
Avatar
RankRankRankRank

Total Posts: 235

Joined 2012-08-31

PM

Hi Rolf,

A method is defined in the profile for the display to provide the crank length to the power meter (get/set parameters using data page 0x02). The same page can be used by the sensor to tell the display the crank length by sensors that have previous values stored or can self-detect. In addition to this a mechanism is provided for custom calibration parameters as different manufacturers needed different calibration information.

The general method for communicating this to the user is via the product manual - i.e. calibrate offset every ride / day, and return the sensor to the manufacturer every two years for a sensitivity calibration. There is a also a flag that the sensor can set in the status page to indicate whether custom calibration is required.

Kind regards

Kat      
Rank

Total Posts: 2

Joined 2013-03-12

PM

Kat,

Thanks.

Do I read the spec correctly that a Display can choose not to implement the full range of custom calibration features provided via page 0x02?

So you could have a fine sensor that requires the user to enter a calibration value on the display and a nice display that supports the power profile - yet not be able to make it work (because the display have chosen not to implement the full calibration, which the sensor requires)?

Thanks in advance for clarifying this.      
Avatar
RankRankRankRank

Total Posts: 235

Joined 2012-08-31

PM

Rolfo,

Yes, it is correct that data page 2 is optional. However using a power meter that requires custom calibration with a display that doesn't support requesting custom calibration should not result in the power meter not being able to be used - it should still transmit power, only it will be less accurate.

If the power meter manufacturer expects that the accuracy will be too bad without a custom calibration, and/or that a custom calibration must be performed regularly by the user then there are several steps they could take to ensure this is acheivable:

1) provide an input method for the user to begin the custom calibration that does not rely on the display - e.g. pedal backwards 5 times

2) provide a display software that the user can install on a laptop to handle the extra calibration steps and guide them through the calibration process

3) recommend that the user choose a display that does support custom calibration and suggest some examples

...

The power meter manufacturers that we work with have not raised this as an issue, and I expect the main reason for this is that to perform a sensitivity calibration is not a simple task to require a user to perform. Multiple known power inputs would need to be applied to the pedals and compared with the current reading, which is likely to involve specialist equipment and multi-step procedures. This is a bit much for the average consumer to take on - hence the preference for calibrating sensitivity in the factory, and designing the sensor such that the drift in this value is minimal.

Kind regards,

Kat