Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

DIY controller for Super Polaris MD-5 motors

This topic has been archived. This means that you cannot reply to this topic.
21 replies to this topic

#1 pico

pico

    Lift Off

  • *****
  • topic starter
  • Posts: 14
  • Joined: 20 Mar 2013

Posted 24 May 2013 - 10:57 PM

I am hoping someone here can read my conclusions of my research and let me know if I am on the right track.

I have a SP mount in like new condition with 2 MD-5 motors and a broken single-axis controller. My thoughts are to create a hand controller using arduino or non-arduino atmega(I have experience with both).

The SP has a 144 tooth RA Gear. The motor-to-ra gears are 1:1. So, if I am reading correctly, the worm must rotate 144 times for a SR-Day(86164 sec) meaning slightly less than 1/10 RPM(actually 59.8 seconds) or nearly one rotation of the RA knob/post per 10 minutes for SR tracking.

The MD-5 motors are 7 degree motors giving 48 steps per rotation with a 30:1 gearbox making it 1440 steps per rotation. Which sadly is 2.4 steps per second. This seems too "jerky" to me. How were these used for photography? Or, I am way off here. Is that fine?

These are unipolar 6-wire motors. I did not think it common to microstep unipolar types. But, I have also heard you can leave the common wire disconnected and treat these as bipolar capable of microsteps?

Please, let me know if I am off on these numbers or anything else that may help guide my direction in this effort.
Thanks

#2 ccs_hello

ccs_hello

    Voyager 1

  • *****
  • Posts: 10196
  • Joined: 03 Jul 2004

Posted 25 May 2013 - 08:45 AM

Pico,

Fairly old tracking only motor (Orion's, Celestron's, etc.) are using a very low end stepper using unipolar drive. The motor unit also has a high gear-down ration thus fast slewing (or even converted GOTO) is extremely slow.


Few suggestions:
- The 6 leads in the stepper are: X+ X- Y+ Y- coils, main common power, and aux common power. The first 5 leads are well known. The sixth one is for low power/low torque drive during tracking. (More on the 6th one later.)
- Do you drive the unipolar stepping motor under bipolar drive mode. The common power lead is connected, thus not great when using bipolar configuration. Also high inductance under such special configuration is not an optimal idea.
- The electronic drive on these is under quarter-stepping rectangular waveform thus (still) very jittery, not good for high-magnification observation and AP. The 6th lead in that motor is trying to reduce the jitter.

Now the mod:
- The best option is trying to change the driving method to microstepping. (But still keep the unipolar driving electronics). Good example on this is Darren's EQ6 mod http://projects.gbdt.com.au/eq6-1/ (changed the drive board's microcontroller firmware.)
- There are other microstepping driver board (e.g., pick 1/16 microstepping.)

Clear Skies!

ccs_hello

#3 orlyandico

orlyandico

    Cosmos

  • *****
  • Moderators
  • Posts: 9328
  • Joined: 10 Aug 2009

Posted 25 May 2013 - 01:17 PM

You can half-step a unipolar. Should be an improvement..

#4 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8669
  • Joined: 02 Dec 2004

Posted 25 May 2013 - 03:08 PM

Before you give up on the original MD-5 controller I would see if can be fixed. I have repaired a number of these over the years and 9 times out of 10 the problem is the 5 volt voltage regulator is bad.

- Dave

#5 pico

pico

    Lift Off

  • *****
  • topic starter
  • Posts: 14
  • Joined: 20 Mar 2013

Posted 25 May 2013 - 03:48 PM

Thanks everyone for the feedback so far.

One thing nobody addressed... did I get the timing close to correct?

Dave- I scoped the controller and the only component I got readings other than a steady 12v was the 7805 regulator which was showing 5v. Can they be blown and still show correct voltage? It looks easy enough to replace, may give it a try.

css_hello-it is difficult for me to tell if you are asking or answering. I am planning on replacing the controller, but I am not sure if I can control these particular motors as bipolar(4 wire mode).

I am thinking unless I can get these motors to do at least 8 microsteps this may not be the way I want to go. Is it possible with something like a polulu or easydriver board? These are actually fairly small motors so I'm not worried about amperage issues on those boards.

#6 DAVIDG

DAVIDG

    Cosmos

  • *****
  • Posts: 8669
  • Joined: 02 Dec 2004

Posted 25 May 2013 - 08:52 PM

I think the motors have 300:1 gear box and not 30:1. I have three SP mounts with drives and I can not see any stepping even at high power (400x or greater ) with my schiefspeigler. With a 300:1 gear box, you need to pulse the motor at 24 Hz which makes more sense
If the crystal is dead in the controller then you won't get any signals also. If you have 5 volts coming from the 7805 then you should see clock signals at the crystal so I would scope it and see if it is alive. Also note that the IC that drives the stepper is sinking current and not sourcing it. So you need to have the motor plugged in and be sure that it is getting 12 volts when you scope the IC that controls it. If I remember correctly the crystal is feeding a binary "Divide by" IC and that slower clock signal feeds the IC's that controls the steppers. The clock signal may also be feed thru one of the slow motion buttons that is usually normal closed. So when you press it, it opens and stop the clock signal from going to the motor control IC. If that switch has failed so it open all the time, then that would cause the unit to not function.

- Dave

#7 ccs_hello

ccs_hello

    Voyager 1

  • *****
  • Posts: 10196
  • Joined: 03 Jul 2004

Posted 25 May 2013 - 11:06 PM

Classic EQ6's gearbox has the 132.8125 or 131.8756346 ratio. I don't know your Celestron's, but let's assume it's 130:1 and no more reduction in the external gear train.
Assuming you'll use 8-microstep/step drive and use unipolar drive arrangement.

X (microsteps/sec) = 144 (worm) x 130 (gearbox) x 48 (48 step/rotation) x 8 (microstep/step) / 86164 (sec per sidereal day)

X = 83.42788 Hz

<edit> when gearbox is 30:1, X = 19.2526Hz </edit>

Clear Skies!

ccs_hello

#8 Arjan

Arjan

    Viking 1

  • -----
  • Posts: 906
  • Joined: 21 Jan 2009

Posted 26 May 2013 - 04:44 AM

I have a SP mount in like new condition with 2 MD-5 motors and a broken single-axis controller. My thoughts are to create a hand controller using arduino or non-arduino atmega(I have experience with both).

This is easy, I did one with a PIC in a day or so. See here, if you need sources let me know.

But, I have also heard you can leave the common wire disconnected and treat these as bipolar capable of microsteps?

Yes this should be possible.
You can half step or even microstep unipolars.

#9 orlyandico

orlyandico

    Cosmos

  • *****
  • Moderators
  • Posts: 9328
  • Joined: 10 Aug 2009

Posted 26 May 2013 - 11:11 AM

I still think microstepping is overkill.

The DD-1 controller (MT-1 motors) used a 120:1 gearbox and 48ppr unipolar steppers. That gives you 5760 steps per revolution. since the SP has a 144-tooth worm (9000"/revolution) then with a DD-1 you'd get 9000/5760 = 1.56"/pulse.

If the MD-5 has a greater reduction than 120:1 then even better. I am almost 100% sure they don't use a 30:1 gearbox. For one thing it wouldn't have enough torque. To turn a worm with a 30:1 gearbox you need hybrid steppers. The dinky tin can 48ppr steppers won't cut it.

According to this link - http://stargazerslou...nd-controllers/

the MD5 has a 300:1 gearhead.

#10 Arjan

Arjan

    Viking 1

  • -----
  • Posts: 906
  • Joined: 21 Jan 2009

Posted 26 May 2013 - 03:23 PM

Agree; I've seen 300:1 in other places as well. Gives you 28800 half steps per rev.

On my Q&D drive I have a 200:1 reduction gear and half step a 1.8deg stepper, giving me 80000 half steps per rev, well below an arcsec per step.

#11 pico

pico

    Lift Off

  • *****
  • topic starter
  • Posts: 14
  • Joined: 20 Mar 2013

Posted 26 May 2013 - 11:16 PM

Except that this one is 1:30
Posted Image

Here are the RA and Dec motors. Not sure if it matters, but they are the darker color with 5-pin DIN connector. Early SPC-8 types.

Posted Image

#12 Arjan

Arjan

    Viking 1

  • -----
  • Posts: 906
  • Joined: 21 Jan 2009

Posted 27 May 2013 - 02:29 AM

I see...
Googling a bit further, using PF42-48 and vixen as terms, you will find more on this. I see DM-5 with 1/300 ratio, MT-1 with 1/120 ratio, and now yours with 1/30.
It seems a bit rough as you say, at least for tracking RA. If you are going for a new controller anyway, why not also look for a replacement stepper?

#13 orlyandico

orlyandico

    Cosmos

  • *****
  • Moderators
  • Posts: 9328
  • Joined: 10 Aug 2009

Posted 27 May 2013 - 03:07 AM

Agree, the Nippon Pulse PF42-48I3G has a lot of compatible motors with different gear ratios.

#14 Nils Olof Carlin

Nils Olof Carlin

    Vanguard

  • *****
  • In Memoriam
  • Posts: 2227
  • Joined: 26 Jul 2004

Posted 27 May 2013 - 08:30 AM

I just looked at my very old SP mount. The DEC motor is type # PF42-4813G but mine says a gear ratio 1/120. That would give 9.6 steps/second (at 48 steps/turn).
Looks more reasonable, if there is a bit of inertia in the OTA.

Nils Olof

#15 Nils Olof Carlin

Nils Olof Carlin

    Vanguard

  • *****
  • In Memoriam
  • Posts: 2227
  • Joined: 26 Jul 2004

Posted 27 May 2013 - 08:44 AM

I made an Arduino microstepping driver for an EQ platform:
http://web.telia.com...per/Stepper.htm
The stepper is a 200 steps/turn (ex floppy), running at about 5 to 6 full steps/sec, at 12 microsteps/step (this was found nice enough for me so I left it there - but just about any number/step is useful - see comments to code). There is also compensation for the arcsine error, not neded for you (hardly for me either, but I did it for fun). It approximates a linear current ramp (triangle waveform) as it suits my motor well, and seems better suited than the sin-cos waveform often used.

There might be some useful ideas there.

Nils Olof

#16 pico

pico

    Lift Off

  • *****
  • topic starter
  • Posts: 14
  • Joined: 20 Mar 2013

Posted 27 May 2013 - 09:02 AM

Thanks Nils.

I am at the point of perhaps selling all my SP parts and mount and just going with a new setup. I enjoy a good project, but also want to get to photography, etc.

Thanks for all the help everyone

#17 AV in CMH

AV in CMH

    Viking 1

  • *****
  • Posts: 661
  • Joined: 07 Nov 2005

Posted 27 May 2013 - 10:10 AM

Hi,

A few years ago I ran into a similar issue with an SP mount for a C8. I had the old Celestron diver unit but got tired of setting it up. So, I purchased an easy driver unit, about 10-$12 per motor. I made a small box that mated with the RA motor to house the EZ driver and a clock circuit I built. Very compact. Here is link to the original source for the PCB: http://www.schmalzha...erExamples.html

After having the Easy driver I created a small Pic 12F509 based circuit to send a clock pulse to the driver. The circuit I built uses two buttons to speed up and slow down the step rate sent by the EZ driver to the SP motor. I also put some LEDs in the circuit to track if the step rate was at normal, tweeked to track planets, or a little faster or slower. Pressing both switches reset the rate to normal. The Leds flash to confirm a switch was pressed. The easy driver has a direction control feature. With one switch you can reverse the direction.

You can make a clock for the EZ drive with a 555 too.

Good luck, have fun,

Anthony

PS,

The EZ drives produced much smoother tracking than the old Celestron unit.

#18 orlyandico

orlyandico

    Cosmos

  • *****
  • Moderators
  • Posts: 9328
  • Joined: 10 Aug 2009

Posted 27 May 2013 - 08:18 PM

pico, do note that an SP with non-working drives will fetch you less than $200. not really enough to get a new setup unless you fork out more cash.

#19 ccs_hello

ccs_hello

    Voyager 1

  • *****
  • Posts: 10196
  • Joined: 03 Jul 2004

Posted 27 May 2013 - 09:21 PM

These are unipolar 6-wire motors.

Thinking this over again, this most likely means the center tap in X and Y axis are not tied together (as in 5-wire), i.e., X1 and X2 can truly separate from Y1 and Y2. In this case, leave the center taps alone, then you have a bipolar (course step: 48/rotation) stepper.

... Which sadly is 2.4 steps per second. This seems too "jerky" to me. How were these used for photography? Or, I am way off here. Is that fine?...


The magic is more than just the step pulse frequency but the driving mode.
For the former, increase frequency will simply "shake" more often. With higher rate, one may no longer detect "step"-like jitter, but the jitter is still there (just hard to count.) But the observed image (though won't be jumpy but) tends to smear.

For the latter,
- full step, half-step, and quarter step modes are all under rectangular waveform and step motor moves in steps (jumpy and creates jitters).
- microstepping is driving the motor in sinusoidal form as if its a regular AC syncho motor. I.e., the stepping motor no longer moves in "steps" but run smoothly.
(Higher microstepping steps/step is better. E.g., 16 ustep/step. Production quantity do fine-tune the coil drive current on each microstepping step to build a smooth near circular movement.)

Hope these details help.

Clear Skies!

ccs_hello

#20 John Carruthers

John Carruthers

    Skiprat

  • *****
  • Posts: 3574
  • Joined: 02 Feb 2007

Posted 29 May 2013 - 03:20 AM

http://proto-pic.co....or-Controllers/

any use ?

some usb and serial stepper controllers.

#21 ccs_hello

ccs_hello

    Voyager 1

  • *****
  • Posts: 10196
  • Joined: 03 Jul 2004

Posted 29 May 2013 - 08:08 PM

Jc,

Some of the boards are microstepping driver boards.

The USB/serial boards are simple PWM DC motor driver boards and not usable.

Clear Skies!

ccs_hello

#22 pico

pico

    Lift Off

  • *****
  • topic starter
  • Posts: 14
  • Joined: 20 Mar 2013

Posted 29 May 2013 - 10:30 PM

the Allegro A4988 used in the Easydriver board uses a sine wave microstepping scheme according to the pdf docs for the chip. Up to 1/16 microsteps, so it is pretty good depending on the gearing.


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics