swampdonkeykami 13 Posted July 7, 2013 Share Posted July 7, 2013 What have you guys used for stepper motor position feedback? I'm aware of shaft encoders, hall effect sensors, magnet and reed switches... but haven't used any to actually get feedback/control. Code/Links appreciated (especially Energia sketches)! It's for use with a little open source 1000PSI hydraulic pump that I'm iterating. The stepper feedback is for accurate hydraulic flow measurement, which will help us zero in on hydraulic positioning (not an easy thing!) curious/interested (blogpost -no BS, no commercials): http://www.arduinoforgoodnotevil.blogspot.ca/2013/07/arduino-1000-pounds-per-square-inch.html THANKS! Quote Link to post Share on other sites
hvontres 22 Posted July 9, 2013 Share Posted July 9, 2013 What have you guys used for stepper motor position feedback? I'm aware of shaft encoders, hall effect sensors, magnet and reed switches... but haven't used any to actually get feedback/control. Code/Links appreciated (especially Energia sketches)! It's for use with a little open source 1000PSI hydraulic pump that I'm iterating. The stepper feedback is for accurate hydraulic flow measurement, which will help us zero in on hydraulic positioning (not an easy thing!) arduino hydraulic pump.jpg curious/interested (blogpost -no BS, no commercials): http://www.arduinoforgoodnotevil.blogspot.ca/2013/07/arduino-1000-pounds-per-square-inch.html THANKS! I would suggest a shaft encoder, especially once the new Tiva Launch pads get out. That one will have TWO quadarature counters on board, so interfacing to the encoder should be very simple. As far as controlling the position of a Hydraulic cylinder, you might be better off getting your feedback from the rod itself. One way that I played with years ago while working on the animation system for a Rose Parade Float (http://www.csupomona.edu/~library/specialcollections/rosefloat/floatimages/1993photo.html) is to use a 10 turn pot with a torstion spring, a large wheel and some steel cable. As the rod extends, the pot turns and you get absolute position feedback from the cylinder. I think if you try to just measure the flow, you will integrate up large errors. swampdonkeykami 1 Quote Link to post Share on other sites
Fred 453 Posted July 9, 2013 Share Posted July 9, 2013 Is using feedback with a stepper a great idea? I always felt you generally had a two options for positioning - either using a motor with feedback (i.e. a servo, whether off the shelf or rolling your own) or using a stepper. If a stepper is correctly chosen and powerful enough, driven properly, etc. then it should predictably go to the position you set. I can imagine if it's underpowered and a step out then sending another move pulse might even encourage it to skip backwards further. Feel free to enlighten me if my knowledge of steppers and servos is a little basic. orso2020 1 Quote Link to post Share on other sites
swampdonkeykami 13 Posted July 9, 2013 Author Share Posted July 9, 2013 I would suggest a shaft encoder, especially once the new Tiva Launch pads get out. That one will have TWO quadarature counters on board, so interfacing to the encoder should be very simple. As far as controlling the position of a Hydraulic cylinder, you might be better off getting your feedback from the rod itself. One way that I played with years ago while working on the animation system for a Rose Parade Float (http://www.csupomona.edu/~library/specialcollections/rosefloat/floatimages/1993photo.html) is to use a 10 turn pot with a torstion spring, a large wheel and some steel cable. As the rod extends, the pot turns and you get absolute position feedback from the cylinder. I think if you try to just measure the flow, you will integrate up large errors. I've seen the fishing line and spool before, but with a resolver. It had also had a plunger at both ends so the system would self calibrate every stroke to avoid compounding small errors. But the main criteria for this application is cheap and easy; discovery/experimentation type stuff... Quote Link to post Share on other sites
swampdonkeykami 13 Posted July 9, 2013 Author Share Posted July 9, 2013 Is using feedback with a stepper a great idea? I always felt you generally had a two options for positioning - either using a motor with feedback (i.e. a servo, whether off the shelf or rolling your own) or using a stepper. If a stepper is correctly chosen and powerful enough, driven properly, etc. then it should predictably go to the position you set. I can imagine if it's underpowered and a step out then sending another move pulse might even encourage it to skip backwards further. Feel free to enlighten me if my knowledge of steppers and servos is a little basic. Thanks! That's what I'm looking for, your thoughts and opinons. I too am wondering if feedback is even required; but then if you miss a step, glitch, whatever and the error compounds... Also, what if somebody wants to use a servo, then you must have feedback, no? Quote Link to post Share on other sites
Lgbeno 189 Posted July 10, 2013 Share Posted July 10, 2013 In my mind a servo implies feedback. I think it's best to use some sensor to ensure a hydraulic valve position to guard against Mis steps and glitches like you mentioned. Even in a stepper based system. The control will need to be able to tolerate the resolution of the step size though otherwise it will oscillate if it is trying to reach a position between two steps. One other sensor type for linear position would be a lvdt transducer which is used in aerospace hydraulic actuators. swampdonkeykami 1 Quote Link to post Share on other sites
hvontres 22 Posted July 10, 2013 Share Posted July 10, 2013 Thanks! That's what I'm looking for, your thoughts and opinons. I too am wondering if feedback is even required; but then if you miss a step, glitch, whatever and the error compounds... Also, what if somebody wants to use a servo, then you must have feedback, no? A DC motor (either brushed or brushless) will need some kind of feed back to control position. With a stepper, the encoder can be used for servo like control, or it can be used to detect missed steps after the motion is done. This way, the motor is run in open loop and if there were any missing steps during the move, you could command a couple of extra steps at the end. With a stepper, you are more likely to loose steps at higher speeds, since motor torque falls off pretty rapidly. Having worked around servo systems for the last 18 years, I would not reccomend them for the uninitiated. Proper tuning and hookups can be a real pain without the propper tools. Also, the effects of feedback failures can be very spectacular and dangerous (imagine an axis flying towards you with full torque applied after the encoder fails and then hitting the end stops with a loud bang... that'll wake you up) But for the really adventurous, they do offer the potential for more performance. One of these days I should set up some articles on how a servo system works and can be tuned.... but that would be a bit too much like work swampdonkeykami 1 Quote Link to post Share on other sites
abecedarian 330 Posted July 10, 2013 Share Posted July 10, 2013 Okay, I know I'm not going to be explaining it properly (10 shots of whisky isn't helping), but maybe this wall of text might make sense. Hydraulic pumps are great, positive displacement, double-lip seals and thick packing helps leakage- measure the volume of fluid pumped and if you know the volume of the cylinder you can calculate position. I think you're trying to avoid pressure relief valves- that simplifies the plumbing and also prevents blowing out the cylinder packing. You will encounter problems with trying to determine position from flow: no matter how well the pump and metering device are sealed, there will be leakage and possible back-flow when the pump is not running and the cylinder is under stress from compression (being pushed back) or extension (being pulled out). This will make your calculations drift over time unless you recalibrate periodically: run the hydraulic cylinder all the way in, zero the count, run it out and measure the flow volume to reach full extension. You could then, maybe, scale things accordingly and have reasonable accuracy calculating the cylinder's position, given you know how far the cylinder will travel for a particular volume of fluid pumped in or out. To 'sort of' make my point: Cranes do not rely on how many times a pump has turned to determine how far the crane boom has extended. They use something like what hvontres suggested- a known, fixed-length spool of cable that unwinds as the boom extends and retracts with the boom and uses a rotary encoder to determine position based on the spool's rotation and knowing that the cable is under no significant tension and won't stretch, which if it did stretch that would be analogous to your pump leaking or such). Granted, that is an extreme case where the crane boom extension can be anywhere from a few meters to 100 meters or more, but I'm guessing that is probably far more than you need. My opinion is that a linear position encoder mounted to the hydraulic cylinder would probably be your best bet. swampdonkeykami 1 Quote Link to post Share on other sites
swampdonkeykami 13 Posted July 10, 2013 Author Share Posted July 10, 2013 One other sensor type for linear position would be a lvdt transducer which is used in aerospace hydraulic actuators. One of them fancy new fangled magnetostrictive pulse gewgaws? Pricey, no? Quote Link to post Share on other sites
Lgbeno 189 Posted July 10, 2013 Share Posted July 10, 2013 I'm actually not sure what the cost is, I've only looked into the sensor circuit not the transducer itself. In theory I think it could be cheap. Austria Microsystems has cool magnetic based encoder chips too, might be worth looking into. Quote Link to post Share on other sites
orso2020 10 Posted July 18, 2013 Share Posted July 18, 2013 Thanks! That's what I'm looking for, your thoughts and opinons. I too am wondering if feedback is even required; but then if you miss a step, glitch, whatever and the error compounds... Also, what if somebody wants to use a servo, then you must have feedback, no? Also, if you turn off lose power you may lose your position of your last step. Quote Link to post Share on other sites
Hex 2 Posted July 25, 2013 Share Posted July 25, 2013 What have you guys used for stepper motor position feedback? I'm aware of shaft encoders, hall effect sensors, magnet and reed switches... but haven't used any to actually get feedback/control. Code/Links appreciated (especially Energia sketches)! It's for use with a little open source 1000PSI hydraulic pump that I'm iterating. The stepper feedback is for accurate hydraulic flow measurement, which will help us zero in on hydraulic positioning (not an easy thing!) arduino hydraulic pump.jpg curious/interested (blogpost -no BS, no commercials): http://www.arduinoforgoodnotevil.blogspot.ca/2013/07/arduino-1000-pounds-per-square-inch.html THANKS! What is used for a mouse position for a computer? They seem fairly accurate, and quite precise. I'd imagine that they should be fairly cheap, huh. Has anyone used something similar? thanks for reading this. Quote Link to post Share on other sites
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.