×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Rpi DIY mount stepper control upgrade?

  • Posts: 151
  • Thank you received: 27
6 years 8 months ago #18012

Please Log in or Create an account to join the conversation.

  • Posts: 30
  • Thank you received: 3
Great stuff to listen in on guys! That 350 KG video and the implied near degree per second movement at that weight is awesome to ponder, given the stingy load limits and performance of off the shelf mounts and what they cost to get even decent AP performance.

I have used perhaps a dozen different phidget products in non-astronomical projects so the 1067 "feels" right in addition to liking the usb programmatic control and the safety features. WBIRK I bet there would be general interest for all 1067 users in a well designed heat sink for that board. I am assuming that if the stepper is driving a worm then back EMF and torque holding issues are not the issues they would otherwise be.

I can see some limited advantages in the .9 degree 400 step NEMA 23 motor in an attempt to avoid or limit microstepping, but I would think I would still need a planetary gearing and couldn't go direct at 400 steps even with 360 teeth. Although I can brute force torque, the smoothness I need is only going to come from stepping/planetary ratios (I think). Doesn't the Phidgets 1067 automatically implement stepping through microsteps anyway?

Assuming you wanted to create a mount to drive a 20-30 KG load and didn't mind a slow slew rate, is there any hope for avoiding a planetary ratio in your stepper for RA tracking? Direct drive mounts are such an illusive goal in general, but even direct stepper drive of a 360 tooth 12 inch worm would seem illusive given the desire for arc-seconds smoothness regardless of the motor size or choice. I am going to either do planetary/microstepping or both.

BTW, I have a Phidgets servo motor controller and a nice high torque CR servo with metal gears.. I have a good python program to control rotation and acceleration (use it on a lesser CR servo for focus control)..how would a high end hobby servo do for mount control I wonder.
6 years 8 months ago #18013

Please Log in or Create an account to join the conversation.

  • Posts: 151
  • Thank you received: 27
hi!
- heating of the chip never was an issue with the 1067, as opposed to other drivers. I assume they dissipate the load in a clever fashion :D.
- 1/16 microsteps are fixed with that board.

I have to say that i am quite happy with the 1067, my only concern is the size (little bit smaller than a Eurp PCB) and the price (100USD per board).
yours
wolfi
6 years 8 months ago #18022

Please Log in or Create an account to join the conversation.

  • Posts: 151
  • Thank you received: 27
hi!
one more note: being a simple physicist and not an electronics engineer, I usually take for granted what my colleagues from EE explain to me, start to think about some general physics rhubarb why it makes sense and beyond that, I simply believe them. i did the same thing with the statement "a stepper motor at rest is integrating" ... especially as it somehow was my observation that the enabled stepper at rest gets hot whereas the moving stepper does this not to the same degree.

here is the explanation - www.geckodrive.com/support/step-motor-basics.html ... "constant output power transducer" says it all :D
yours
wolfi
6 years 8 months ago #18023

Please Log in or Create an account to join the conversation.

  • Posts: 257
  • Thank you received: 22
Let's see if I can get this right... In my own words and hopefully avoiding jargon, a stepper holding a position has one or a set of coils energized and holds it so the power is constant and duty cycle 100% on that coil, where as a stepper in motion is changing coils and direction of current flow, making a 50% duty cycle and spreading out the heat load. Also, you would think the lower impedance of the coil at rest should make it roll back the output on its own in order to maintain the current setpoint. So why does it still heat up? Flux driven Hysteresis buildup in the core from the DC applied to the coil without switching. When running, the current flow changes direction every step which cancels out any residual magnetic flux/eddy current/ hysteresis, while at rest the core effectively gets a solid shot of direct current and magnetizes which creates eddy currents in the core and lots of heat...I think. :D
A tech that works with me is actually better at physics and theory, after 20+ years working together I still routinely bounce theory off him. I'll ask if I got it right when he comes over this evening. 8-D

The torque curves on that page are pretty neat!
Last edit: 6 years 8 months ago by Ray Wells.
6 years 8 months ago #18034

Please Log in or Create an account to join the conversation.

  • Posts: 151
  • Thank you received: 27
Hi! I simply think that - with constant power output, you can either put some power into motion ... force*travel/time ... or all into heat ...
Yours wolfi
6 years 8 months ago #18042

Please Log in or Create an account to join the conversation.

  • Posts: 257
  • Thank you received: 22
Yeah. That's the first thing Mark said, and it satisfies the law of conservation of energy and Occam's razor, being the simplest solution. "The energy used to create motion is subtracted from the heat produced", meaning that when it is turning the shaft some of the energy is converted to motion instead of heat, but it doesn't really solve all the variables involved or jibe with the fixed current (P=I2R), which should lower power output at standstill due to the loss of inductive reactance (XL=2piFL) ...and I Never did like that Occam guy. :P
He also mentioned that every stepper he has seen has drawn dramatically more current at standstill than while running if not current regulated. This led to us doing what we do all too often in repair, trying to guess from a distance what is wrong with something, which is sometimes helpful. In this case we both agree that your phidgets might not be limiting current correctly at zero or possibly set too high and running in voltage mode. You might try setting the limit lower while at zero (or possibly in tracking mode?), since they draw less current while moving they should still be okay at speed. Of course, if the heat isn't damaging, there's no need to change anything.

Interestingly, I've done repairs on industrial stepper drives for decades without actually knowing that much about the quirky little motor. This thread has been very educational.
I should add that I don't see them all that often in manufacturing, mostly brushless DC(really ac servo) and ac servos these days.
Last edit: 6 years 8 months ago by Ray Wells. Reason: because time speeds up as you age.
6 years 8 months ago #18061

Please Log in or Create an account to join the conversation.

  • Posts: 30
  • Thank you received: 3
Has anyone had experience with or considered a friction drive rather than toothed drive? I know it gets good reviews in high end systems but why wouldn't you be able to machine out two diameters and use stepper to work with the resultant ratio? . Some reasons might be problems:

1) The small contact area between metal rollers, while sufficient for low speed tracking performs differently at high speed slewing?
2) The risk of slipping can lead to loss of position and potential collisions
3) Exact specific ratios are easier to achieve by # of teeth rather than carefully machining two diameters to match
4) What is popular and working is tried more often by others hard to get suggestions for machining
5) Possible environmental effects on performance? I don’t know.
Seems a stepper/encoder/controller solution would work great with a friction drive at any size and be in the reach of DIY fabrication of the rollers.

found this link www.dfmengineering.com/news_telescope_gearing.html
6 years 8 months ago #18063

Please Log in or Create an account to join the conversation.

  • Posts: 151
  • Thank you received: 27
hi!
well, I learned a lot in the past 9 months, especially as this stepper controller stuff popped up in my academic work when working with a lung dosimetry phantom (i am a medical physicist), and TSC became the spin-off of that stuff. so yes, there is of course more than the first order approximation (heat vs. motion) as we do not have 100% efficacy. However, in my application it is solved by disabling the drive (this is a software command for the phidgets or a logical i/o for the drv 8815) when it is not needed. the current regulation thing sounds attractive at first sight, but my experience is that the correct current is the one where the stepper operates smooth under load without loosing steps. i never saw a reason (heating or insufficient precision at higher speeds) to tinker with current byond that. too muc current simply causes rough operation and overheating. so far for the constant output transformer - it has it's sweet spot for the load and voltage given ...

friction drives - i know a few ones here in Austria - funny enough we have quite a number of rather big fork mounted ATM instruments here, the reason might be a detailed plan that is circulating for the past 30 - 40 years called the "Austrian Precision Mount" which was designed by the head technician of Vienna University Observatory, Mr. Pressberger. Having built a fork mounted 60" RC himself, quite a number of people have made instruments in the 35-70 cm aperture range following his concepts. my houghton is one of the smaller instruments, but it is very heavy because of its front mounted corrector. my friend richard gierlinger has driven a 70 cm telescope with a friction drive - he says they are easy to control, but the roller and disk have to be made of stainless steel, and they must not be hardened. that's it, more, or less ....
yours
wolfi
6 years 8 months ago #18064

Please Log in or Create an account to join the conversation.

  • Posts: 30
  • Thank you received: 3
Wofi, I looked at some of the online resources on the Austrian Precision Mount. If your friend has any more details about the construction of a friction drive, I would be very interested in them. THe sorts of specific questions I have are:

I see references to the need for hardened steel in designs referenced online--did you mean hardened? or is the need for non-hardened stainless steel because of unique friction qualities of the Softer metal yet it retains durability? Or would regular steel or iron work except for rusting? Is a smooth contact area desired free of imperfections and perfect diameter or is a knurled or scored contact area desired? What would be a typical thickness of a wheel to ensure sufficient contact area? Any bearing recommendations? I am thinking of a GEM style mount with a 20KG payload or so stepper and TSC driven.

Sincerely grateful to all
John
6 years 8 months ago #18148

Please Log in or Create an account to join the conversation.

  • Posts: 151
  • Thank you received: 27
Hi! No personal experience, just hearsay from an experienced friend:
- unhardened, stainless steel
- bearings should be massive from my gut feeling :)
Yours wolfi
6 years 8 months ago #18161

Please Log in or Create an account to join the conversation.

  • Posts: 257
  • Thank you received: 22
Hi guys, Quick update on this project. After much discussion, deliberation and review, I finally ordered a couple nema17 steppers to do the upgrade. I also got some timing gears and belts, aiming for 5:1 reduction. Once I had the parts on hand I hit a bit of a snag with clearances. The large 60t pulley contacts the RA clutch arm so it's starting to look like the smaller cg-4 mount is going to need to use gearheads after all. I'm still fiddling with mounting and ways around it, like perhaps putting a washer stack under the main bearing to jack up the RA head a bit for clearance but that's where things stand atm.
cheers!
6 years 6 months ago #19488

Please Log in or Create an account to join the conversation.

Time to create page: 1.017 seconds