Shop OBEX P1 Docs P2 Docs Learn Events
measuring rpm off a tesla turbine — Parallax Forums

measuring rpm off a tesla turbine

science_geekscience_geek Posts: 247
edited 2009-03-02 06:58 in General Discussion
i need to measure the rpm of a tesla turbine, one problem, i cant touch the shaft. one idea i had was to shine a laser at the exhaust holes and then when there was no hole it would act like a mirror and bounce the laser back at a sensor, so that the angle of reflection would be different if there was a hole compared to no hole. my question is would a light to frequency converter work better than a photoresistor, i plan on using a propellor chip that way i have the speed to read the sensor, any comments or suggestion will be very appreciated

Comments

  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-03-01 18:46
    You could always use a hall sensor, and a tiny magnet on the side of the shaft or part that is spinning.
  • Alex41Alex41 Posts: 112
    edited 2009-03-01 18:51
    I'd like to hear more about your Tesla turbine. Any pictures, what size is it and what are using to power it?

    I've thought about building one for grins, did you get the disks CNC cut?


    Thanks for any info,

    Alex

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    If at first you don't succeed, then skydiving is not for you.
  • science_geekscience_geek Posts: 247
    edited 2009-03-01 19:14
    cant do magnets they fly off and cause imbalaces,

    will post pictures when done working on scholarships, currently i am using compressed air to power it, i am working on a nozzle to direct air into all disk spaces, its about 2" in diameter there·are 4·holes· 1/4" in diameter i have built about 5 turbines, the first time i ever built one i made it out of cd cases and had a styrofoam housing, no bearings and paper spacers, i got it spinning so fast it melted the cd case faceplate and when the blades hit the houseing the paper spacers exploded and it was in shambles, the next one was the same but with bushings, i exploaded without touching the housing wall, the next few were made from sheet metal cut with a hole saw, they worked good,i have use fender washers, but·you need to get a bunch of them because there is no precision to the thickness,·3" disks seem to run awesome with 100-120 psi, 2" disks are good at lower air speeds with a smaller nozzle, the one im working with right now i machined from 2.5" aluminum and used a cut off tool to cut the disks, the spacing is a little to big, but when i get parts fro my lathe i plan on making an even better one
  • kwinnkwinn Posts: 8,697
    edited 2009-03-01 20:10
    Try a reflective sensor. It consists of an LED/phototransistor in a housing. It is made specifically for this type of application. Typically all you need to add is 2 resistors, one to limit LED current and one on the collector of the transistor. Something like the OPTEK OPB608A should work if there is a large difference in reflectivity between the holes and shaft. If the difference is low you may need to add an op amp or comparator to get a usable signal.

    PS - Don't forget if you have 4 exhaust holes in the shaft you will get 4 pulses per revolution.
  • Carl HayesCarl Hayes Posts: 841
    edited 2009-03-02 06:58
    Jay, you can use a reflection sensor or a Hall sensor or any of several other approaches, most of which give you a series of pulses. For quick rotation it may be sufficient to measure the number of pulses in some period of time. For greater precision, especially if revolution is slow, you can measure the time between pulses instead.

    In general, if the number of pulses per second, say, is greater than the square root of the·number of counts per second you can count, then you count pulses; otherwise you count the time between pulses.

    Example: Say the shaft is turning 100 rps, and you can count in microseconds. Then if you count shaft rotations for a second, you'll count 100, getting a resolution of 0.01 second (which is 1% of the measurement). But if you count microseconds from pulse to pulse, your measurement will be 10,000 microseconds, with a resolution of 0.0001 second, or 0.01% of the measurement. You would probably choose to count the time between pulses.

    But if the shaft is turning 10,000 rps, which is pretty fast, then you'd only count 100 microseconds per pulse, getting 1% accuracy. Counting the pulses for one second would give you 10,000 pulses, with 0.01% resolution.

    So you should figure it both ways: How accurately can you measure time, and how accurately can you (instead) count pulses?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · -- Carl, nn5i@arrl.net

    Post Edited (Carl Hayes) : 3/3/2009 1:52:06 AM GMT
Sign In or Register to comment.