My friends and I are making an open-source haptic device, called the haplet, that allows you to feel virtual objects on your tablet or computer screen using force feedback.
We've been using the Stanford hapkit, which is an uno with a motor driver and h-bridge attached. We're currently building our own equivalent board with the Arduino Due.
We're looking for feedback on the idea and we're trying to gauge if people would be interested in making their own haplets using our blueprints, programming apps for it, or using pre-made apps. Let me know what you think!
Please message me for more information, or just to chat about haptics!
Are you using a similar motor control system (dc motor, dual-ended shaft, magnet and magnetoresistive sensor) for the feedback? Are the motors geared, or bare?
If you could figure out a way to do this with servos in some manner, it would open things up greatly. Most people who've played with the Arduino for any length of time have a couple of servos lying around, and if not, they are readily available. However, they don't really like to be back-driven.
If the swing of the arms isn't too great (doesn't look like it) - then maybe a version could be done using potentiometers instead of a magnetic sensor; while it might not lower the cost much, it would likely be more accessible to others - versus needing to purchase a special board and such.
Just some thoughts I'm having - it looks like a fun and interesting project!
Sorry for the delay I went on a trip and the forum slipped my mind :S
Cr0sh:
Are you using a similar motor control system (dc motor, dual-ended shaft, magnet and magnetoresistive sensor) for the feedback? Are the motors geared, or bare?
The motors are geared using a ratio of 27:1 and have built in encoders. It allows back-drivability while generating an adequate amount of torque to have meaningful feedback. The motor control system uses a torque control proportional to the current. We can get the required torques by mapping them back from the forces generated at the handle!
The design requirements are that it has to be highly backdrivable while insuring I don't get a lot of quantization from the encoders. I might tinker around with some servos to see if I could get the backdrivability. I'm a little apprehensive about the plastic gears being back driven hard though. Will maybe give it a shot for a lower fidelity option : )
Do you know if using a pot instead of the rotory encoder would provide enough resolution for the angles? I'm woried about the limitations of the ADC on the arduino.
P.s... mart256, the board is just to show how small the device is. It's using the hapkit which is an open source board built by stanford that's just an arduino uno with an H-bridge attached to drive a couple motors. Check out the "haptic paddle" if you're interested. That was the seed idea of this project but we wanted to make it a little more interesting for users a single degree of freedom : )
I want to try and make this project really accessible but also make sure that the quality is good enough that maybe it could be used as a force feedback mouse or something. Maybe some cool game mods (i'm thinking DOTA where the spells and attacks could be felt!)
