Doubts regarding Qwiic Flex Glove Controller

Sensors Force & Weight
1

I am working on a project using the Qwiic Flex Glove Controller (https://www.sparkfun.com/products/14666). I was able to hook them up (using https://learn.sparkfun.com/tutorials/qw … okup-guide) and can read the values but I have some questions about it.

  • On the Sparkfun website (https://learn.sparkfun.com/tutorials/fl … okup-guide), it says that the sensor’s resistance ranges between 30kOhm to 70kOhm at 90° bend. But when I measure the sensors on the Qwiic glove controller, each measured only about 7.5kOhm and 8kOhm. How could there be such a big difference?

  • The fixed resistors on the voltage divider circuit are marked as 100kOhm in the schematic. The one’s on my board measure only 47kOhm and 52kOhm. Why would this be?

  • Can I give 5V as input to the board even though it is marked as 3.3V on it? I see that the ADS1015 and LMV358 can handle up to 5V on their input. My motivation is that giving 5V would provide a larger range as output from the voltage divider.

  • I measured input and output voltages at the op-amp. And I see a drop of 300mV even though it is in voltage follower configuration. Why would this be?

    • 2

    1. I can’t say for certain

    2. Maybe a slight misprint? 47+52 is very close to 100…

    3. I believe it’s possible but it could saturate the magnetometer (note the “Pins” section of the hookup guide

    4. Not sure

    I’d sooner try calibrating @ normal 3.3v and running code that is specifically linearized to the values than increasing the supply voltage…the main thing is trying to get it to output useful signals; please let us know if something is actually malfunctioning

    3

    Posting because this could be useful to someone else having doubts about the glove controller. Or this could also be the case that I was unfortunate to receive these boards.

    I had bought two glove controllers for my project. Even though I could read and access the registers, the data did not make sense. To investigate this, I measured the resistance of the flex sensor and the fixed resistor and calculated the voltage divider output. I also measured the voltages at the input and output of LMV358 and ADS1015.

    The conclusions that I drew from the tests were:

  • As mentioned in the question, the values of the resistors were not as expected from the datasheets and schematic.

  • The input and output at LMV358 did not match even though it was in voltage follower configuration.

    • These narrowed the issues down to LMV358. After a lot of research, I found the answer as to why.

    The datasheet of LMV358 (DC Electrical Characteristics) says that the input common-mode voltage is Vcc-1 (1.7 for 2.7V and 4 for 5V). Also discussed here (https://e2e.ti.com/support/amplifiers-g … tch=LMV358#).

    Accordingly, the LMV would only work as expected below 2.3V (Vcc = 3.3V).

    But with the combination of resistors on my board, Finger 1 : Flex sensor = 8.49kOhm to ~18kOhm and R2 = 54kOhm, and Finger 2 : Flex sensor = 9.17kOhm to ~20kOhm and R2 = 47kOhm, the expected output voltage from the divider would be Finger 1 = 2.85 - 2.47V and Finger 2 = 2.76 - 2.3V.

    This means that the input is never below the common-mode input voltage. So now I have two controllers that are defective. I am curious, aren’t products tested before they hit the market?

    My thought now is to apply 5V. If I do so, the voltage input, according to my calculations and tests will come under 2.3V between 90-180°. This way, I could at least sense a flexion and have some use for the product.

    The ICs on board are ADS1015 and LMV358. Both of which can handle 5V according to the datasheets. I am not sure if there is anything else to be considered while applying 5V.