hello
love your robots. amazing. really, like no others i've seen.

btw, i was thinking, you said you use force sensor on your robots to detect whether the leg is touching the ground. are you having it on the tip of the toe?

i was thinking of an alternative. what about current sensor on the power line for the servos? might just be some small smd for each of the legs (6 of them). it should be able to detect the pressure applied on the legs by the amount of current the draw. just a sudden thought. i dont know of any real advatages yet compared to the pressure sensors.

what do you think?

Hi,

how can you difference between movement and pressure?
I had the same thought and with this you have to know the power while the leg is moving to ground and have to build the difference between actual power and "only moving power". I don't know if this is excactly enough for this case.

Stephan

I was thinking of trying to use a measurment of the drawn current by a leg to give an indication of "touch", although this will obviously not be anywhere near as usefull as a leg which incoperates a proper foce sensor, it could allow for a lateral touch sensor, ie: if a leg bumps into something whilst trying to move forwards.

Thanks for the kind words :)

You can use the current in the motor for current sensing, but there are some disadvantages. Primarily the fact that for optimum battery usage, you want to design your leg to draw very little current when in a neutral position, however, most of my design are a trade off between aesthetics and practicality, and so current sensors should work.

In order for current sensors to work properly you really need to know the direction of the current flowing through the motor, and not just the current drawn by the power lead of the servo. This way you can determine the direction of the force applied. The current sense technique is particularly useful for bump sensing of obstacles during the swing phase of a leg.

Matt.