Loading ...
Sorry, an error occurred while loading the content.

Re: [SeattleRobotics] Re: Engineering problem for discussion

Expand Messages
  • David Buckley
    If you buy a professional lever actuated pot such as is used on golf carts, and mobility chairs, or an industrial quality servopot (~$15) you won t have to
    Message 1 of 67 , Feb 3 12:37 PM
    • 0 Attachment
      If you buy a professional lever actuated pot such as is used on golf carts, and mobility chairs, or an industrial quality servopot (~$15)  you won't have to test it, that will already have been done for you by the manufacturer.
      If you buy a standard pot - like for an old fashioned volume control, it won't last.
      DAvid
       
      ----- Original Message -----
      Sent: Sunday, February 03, 2013 7:40 PM
      Subject: Re: [SeattleRobotics] Re: Engineering problem for discussion

       

      Michael,


      One good thing about your application is that (I presume) it is low duty cycle -- an organ only gets played a few hours a day and probably only a few days a week. The swell action is not constantly moving during playing, either. That means you can run a "life test" on your mechanism running it back and forth 24/7 and put many years equivalent service on it in a week's time. While there is much worthwhile to learn from those of us who have built systems in the past, there is no substitute for building prototypes or partial mockups and running torture tests with them. In your budget you should be prepared to destroy a few things in the learning process!

      I would also torture test the swell pedal -- maybe pushing on it with a pneumatic cylinder if you have that sort of gear around. It sounds simple to hook up a pot to a pedal but the devil is in the details!

      Chuck

      On Sun, Feb 3, 2013 at 10:50 AM, David Buckley <david@...> wrote:
       

      The problem with brushless motors is you need fancy controllers, easy if you are in the hobby/model world but not as common and more expensive in the industrial engineering world.
      Brushed motors in RCservos normally use plain metal brushes and as you found out, they don't last.
      Even if you use brushless RCservos you will find the gearbox bearings wear out.
      Nevertheless I find that nowadays, if you are not making the servos work hard, brushed Hitec servos can last for 200,000 cycles or so. Maybe the technology has improved since 1993, who knows.
       
      If you use carbon brushed motors and underrate them and run them slowly they last forever and are ridiculously cheap. Even at rated voltage the life of an industrial brushed (DC) motor is between 2000 and 5000 hours - eg http://www.anaheimautomation.com/manuals/forms/brush-dc-motor-guide.php
      DAvid
       
       
      ----- Original Message -----
      Sent: Sunday, February 03, 2013 6:29 AM
      Subject: RE: [SeattleRobotics] Re: Engineering problem for discussion

       

      We (Karl Lunt, Bill Harrison, Keith Payea, and myself and perhaps other SRS folks I can't remember) built a display for the Pacific Science Center back in 1993. We found that the servos worked great, but at about 100 hours of running, the brushes on the motors failed. We showed the people who maintain the displays how to change them out. We gathered a lot of data on the fact that a Futaba servo motor is good for 100 hours of operation until it fails. Failures where only +- about 5 hours, so at least it was consistent and predictable. The display ran for 3 years, and the PSC went through 8 servos a week.
       
      Point being: If I were to do something like this, I would highly recommend a brushless motor. They to last a LOT longer.
       

      To: seattlerobotics@yahoogroups.com
      From: kb7wnz@...
      Date: Sat, 2 Feb 2013 20:04:05 -0800
      Subject: RE: [SeattleRobotics] Re: Engineering problem for discussion



      Ok, I'm glad to hear this isn't viewed as a universal replace all option.  What resources do you have in your shop?  Must be an iteresting mix of wood and metal working.
       
      Rob




    • Michael
      Thanks David and everyone. I will wade through those asap and see what I can continue to learn. HEY, I did want to let everyone know. I had been having a bit
      Message 67 of 67 , Feb 14 11:12 PM
      • 0 Attachment
        Thanks David and everyone. I will wade through those asap and see what I can continue to learn. HEY, I did want to let everyone know. I had been having a bit of trouble with Yahoo. For a while now it has made every effort to re-associate my account with old unused ones and today I found I could not get into ANY of them. After some gyrations I determined they would not help me so rather than go through this at every password change and cache clearing I set up a new account so my new user name is "michaelorgel" formerly "hesperion22". email is still the same however. In my organ related group of course it caused me lose privileges over my photos etc. and the admin has no idea what to do about it. I thought I would just announce it here and move on, as they say. Anyway, I'm still here with a new ID which hopefully will be less troublesome. Thanks for sticking with me.

        --- In SeattleRobotics@yahoogroups.com, "David Buckley" wrote:
        >
        > Hi Michael
        >
        > That size shouldn't be a problem for a wiper motor.
        > Kevin is right you could try your junkyard and the window winder motors are good for smaller projects.
        > I once use two Ford XR3i window motors for an emergency repair, they were great with nice long output shafts, most car motor/gearboxes have shafts which are difficult to connect to.
        > However you can buy wiper motors new for probably the same price, and for a commercial item you would want to use the new parts anyway.
        > I did a quick Google search -
        > http://www.scary-terry.com/wipmtr/wipmtr.htm
        > http://www.scary-terry.com/wipmtr/wipmtr2.htm
        > http://monsterguts.com/index.php?act=viewCat&catId=2
        > http://monsterguts.com/index.php?act=viewProd&productId=306
        > http://monsterguts.com/index.php?act=viewProd&productId=4
        > http://monsterguts.com/index.php?act=viewProd&productId=105
        > http://monsterguts.com/index.php?act=viewProd&productId=11
        > With the monsterguts parts you could rig-up a prototype controlled by switches.
        > You will need to source some ball-ends for the linkages, Try Amazon Small Parts.
        > Arrange the linkage (just like in a car) so as the motor rotates continuously the vanes go from shut to open to shut to open and never hit any end stops, that way nothing breaks. You might want to add an obvious way that an untrained person could disconnect the linkage so the vanes could be adjusted to a fixed position by hand just in case something goes wrong - would it matter? Could the volume be too loud or too quiet?
        > When that works you could add a servo pot, use the type with a 1/4" shaft and end stops. (at least servopots haven't yet gone metric - here in the UK volume control pots have 6mm shafts and all the knobs still have 1/4" holes so the knobs rotate off axis!!!!).
        > Connect the pot by a linkage arm so that it rotates no more than 90degrees as the vanes go from open to shut. Arrange the pot so when the vanes are in the mid position the pot is in its mid position (just move it by hand from one stop to another and guess the middle position just like you would do with a volume control. - If you connect the pot by a tooth belt drive or gears you could go to 180 degree rotation from vanes open to vanes closed. Don't worry about a bit of backlash in gears etc, it won't matter, you are not building a CNC mill. The important thing is you don't want the pot to rotate to anywhere near its end stops and you are using one with end stops so you can easily find its mid position.
        > Add a servopot to the swell pedal just the same. It is easier if both pots move 90degrees or 180 degrees etc.
        > Next get a servo amplifier which will handle the voltage (12V) and current of the motor and which has easy to understand instructions to set up a position feedback system. You might want to come back and ask Chuck which are good companies to deal with and suitable parts to buy.
        >
        > DAvid
        >
        >
        > ----- Original Message -----
        > From: hesperion22
        > To: SeattleRobotics@yahoogroups.com
        > Sent: Thursday, February 14, 2013 4:20 AM
        > Subject: [SeattleRobotics] Re: Engineering problem for discussion
        >
        >
        >
        > Hi David,
        > The opening in this proto is about 4'(wide) x 3' (high) with transverse shades each about 5" wide, total of 5 shades. Do you really feel it will be that simple? I am liking this if so. LOL. What is an example of a windshield wiper motor for reference? Do you mean an actually automotive part?
        >
        > --- In SeattleRobotics@yahoogroups.com, "David Buckley" wrote:
        > >
        > > Michael
        > > Should have asked also - how big is the shutter area - in general?
        > > The answer is looking more and more like a windscreen-wiper-motor sized geared-motor plus a small servo-amplifier from Chuck's links, a power supply, two 5K servo pots and 10 pieces of wire plus some brackets and levers.
        > > DAvid
        > >
        > > ----- Original Message -----
        > > From: hesperion22
        > > To: SeattleRobotics@yahoogroups.com
        > > Sent: Wednesday, February 13, 2013 6:17 AM
        > > Subject: [SeattleRobotics] Re: Engineering problem for discussion
        > >
        > >
        > >
        > > Answers: It is always preferred to have a balanced swell action. The player would expect that and I want the experience to be what they would expect. The time, open to closed, of the shades would rarely be any quicker than a second or so. I can not think it would be less. The sense of it is: The foot can move to open or close the box and one could sort of imitate that motion to visualize the time it takes. Rarely is it ever slammed rapidly. Not good form.
        > >
        > > --- In SeattleRobotics@yahoogroups.com, "David Buckley" wrote:
        > > >
        > > > Michael
        > > > One thing not mentioned so far is what sort of Swell Pedal you wish to implement. Presumably it would be a Balanced Swell Pedal rather than Ratchet or Infinite Speed types.
        > > > I take the movement of the Shutters is normally about 90 degrees. It would be useful to know how fast most Swell Shutters can move, that is how quickly can they move from fully closed to fully open? For example 1/10 second, 1/4 second, 1/2 second, 1 second, longer?
        > > > DAvid
        > > >
        > > >
        > > > ----- Original Message -----
        > > > From: hesperion22
        > > > To: SeattleRobotics@yahoogroups.com
        > > > Sent: Saturday, February 09, 2013 10:27 PM
        > > > Subject: [SeattleRobotics] Re: Engineering problem for discussion
        > > >
        > > >
        > > >
        > > > Still trying to comprehend all the details but to attempt to answer: The entire motion (almost without exception) is within 90 degrees and it would not have to be extremely fine increments. I am gathering a short-list of components and adjunct parts based on the 2 or 3 schemes that have been suggested. The arduino idea is still strong in my mind but, as some have pointed out, it is probably overkill in light of simpler options. Run the components with lots of margin and at lower than rated power and the optical rotary "pot" is very attractive for its longevity. I hope to demonstrate my understanding by listing the parts for your approval soon. THANKS TO ALL!!
        > > >
        > > > --- In SeattleRobotics@yahoogroups.com, Rob Purdy wrote:
        > > > >
        > > > >
        > > > > How accurate do you need those steps to be? @ 256 steps you will see some vaiance due to inaccuaracies in the motor build. There's also some torque loss. That's part of the reason Gecko doesn't do 256 micro steps. Rob
        > > > > To: SeattleRobotics@yahoogroups.com
        > > > > From: james@
        > > > > Date: Tue, 5 Feb 2013 16:37:38 -0800
        > > > > Subject: Re: [SeattleRobotics] Re: Engineering problem for discussion
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > > I'm familiar with Gecko, but all of their drivers are 1/10 uStep at the finest setting. I was specifically curious about 1/256 uStep because it gets you to a finer resolution without introducing a gear reduction in the system somewhere.
        > > > >
        > > > > I have a space-constrained application which needs motors to turn a shaft at ~8192 steps per revolution at a very slow speed (e.g. 24 hours per revolution). I'll probably end up going DC with an absolute encoder, but the stepper option is attractive since I've already got boilerplate arduino code to drive 2 axis steppers.
        > > > >
        > > > >
        > > > > On Tue, Feb 5, 2013 at 4:18 PM, Rob Purdy wrote:
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > > Start with Geckodrive.com The white pages Mariss wrote should be required reaqding for any one gettinginto CNC.
        > > > >
        > > > > That noise can be mitigated many ways but mostly it will depend on what Freq. The chopper circuit is running.
        > > > >
        > > > >
        > > > > Rob
        > > > >
        > > > > To: SeattleRobotics@yahoogroups.com
        > > > > From: james@
        > > > >
        > > > > Date: Tue, 5 Feb 2013 15:30:04 -0800
        > > > > Subject: Re: [SeattleRobotics] Re: Engineering problem for discussion
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > > On Mon, Feb 4, 2013 at 8:12 PM, David Wyland wrote:
        > > > >
        > > > >
        > > > > That said, a stepper motor driven by a micro-stepping controller can
        > > > >
        > > > > achieve very fine resolution, e.g. 200 x 256 =>51,200 steps/ revolution.
        > > > >
        > > > > And stepper motors can be fast. I know of a 200 step/rev stepper motor
        > > > >
        > > > > application for camera pointing that runs at up to 20,000 steps per
        > > > >
        > > > > second. This works out to 6,000 RPM! So, I recommend looking at using a
        > > > >
        > > > > stepper motor as your servo. They are available in a wide variety of
        > > > >
        > > > > sizes, both commercially and surplus.
        > > > >
        > > > > I was not aware there were stepper drivers that could microstep down to 1/256. Most of the stuff out there is 1/10. Can anyone recommend a good brand?
        > > > >
        > > > >
        > > > > Another thing to keep in mind, since this is a musical application, is noise. Stepper motors put out a distinct tone relative to the speed they're being driven - even the little ones. DC motors and servos are almost always quieter.
        > > > >
        > > > >
        > > > > I'm new to the list, but happy to find some like-minded folks. My name is James, and I work in the Center for Digital Arts and Experimental Media at the UW.
        > > > > Best,
        > > > >
        > > > > James
        > > > >
        > > >
        > >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.