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

Steppers or Servos?

  • Please log in to reply
29 replies to this topic

#1 zakry3323

zakry3323

    Viking 1

  • *****
  • topic starter
  • Posts: 848
  • Joined: 11 Apr 2016
  • Loc: Pittsburgh

Posted 16 February 2019 - 10:41 PM

I'm about 3/4 of my way to being able to afford a new (maybe presently affordable used, though I have a lot of reservations about buying a used), I suppose what is considered a mid-tier (around 50lbs capacity) mount, upgrading from an AVX primarily for astrophotography use. I've narrowed my selections down considerably over the last 8 months that I've been able to save for it. 

Besides audible noise, slewing speed, and torque, are there any other considerations I should be thinking about when when thinking about motors?
 


Edited by zakry3323, 16 February 2019 - 11:47 PM.


#2 DuncanM

DuncanM

    Mercury-Atlas

  • -----
  • Posts: 2502
  • Joined: 03 Nov 2009
  • Loc: Arizona Sky Village or the rain forest

Posted 17 February 2019 - 12:01 AM

I like steppers because they're nearly bullet proof, and have far fewer moving parts than a servo. They also have  less torque and do little or no damage if they run away.  Slew speeds are a bit less but worth the trade-off IMHO.


  • zakry3323 likes this

#3 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 15458
  • Joined: 27 Oct 2014

Posted 17 February 2019 - 12:13 AM

Some high end manufacturers (better mounts than you're considering) use steppers, some servos.   This would not be a mount decision factor for me.


  • zakry3323 likes this

#4 TxStars

TxStars

    Surveyor 1

  • *****
  • Posts: 1753
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 17 February 2019 - 01:48 AM

If you are not planning to use your mount to slew for hundreds of hours every month they will all last many years.

 

The only difference in the drive systems is the longevity of the motors.

 

The mid to lower end mounts use "brushed servo motors" which have a commutator and brushes.

These are parts that wear with use and at some point the motors will need to be replaced.

 

Some of the highend mounts use "brushless servo motors" and these have a long life span.

 

Stepper motors do not have a commutator or brushes that wear, so even the less expensive ones have a long life span.


  • zakry3323 likes this

#5 zakry3323

zakry3323

    Viking 1

  • *****
  • topic starter
  • Posts: 848
  • Joined: 11 Apr 2016
  • Loc: Pittsburgh

Posted 17 February 2019 - 08:21 AM

Thanks gents, this was exactly what I wanted to know!


Edited by zakry3323, 17 February 2019 - 08:22 AM.


#6 555aaa

555aaa

    Vendor (Xerxes Scientific)

  • *****
  • Vendors
  • Posts: 1264
  • Joined: 09 Aug 2016
  • Loc: Lynnwood, WA, USA

Posted 17 February 2019 - 11:53 AM

Servos provide more precise control because by definition they use a motor mounted encoder to control shaft angle. A stepper motor used on the RA axis is used in microstepping mode and cannot provide precise control. Servos provide torque proportional to current and can be used in clutchless mounts to sense balance. Servos can provide smooth torque delivery over a higher speed range and have much lower to zero cogging torque. Steppers are by design "all cogging."That is an advantage if you need to provide holding force when no power is supplied. Steppers are cheaper and have cheaper control electronics typically.
  • zakry3323 likes this

#7 TxStars

TxStars

    Surveyor 1

  • *****
  • Posts: 1753
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 17 February 2019 - 03:44 PM

Other than added cost there is nothing keeping a mount builder from using a closed loop stepper system which does provide precise control.

http://www.jss-motor...and-driver.html

These provide a larger rpm range as the stepper can be run as a 2 phase servo motor with the encoder providing position feedback..


Edited by TxStars, 17 February 2019 - 03:48 PM.

  • brave_ulysses likes this

#8 DuncanM

DuncanM

    Mercury-Atlas

  • -----
  • Posts: 2502
  • Joined: 03 Nov 2009
  • Loc: Arizona Sky Village or the rain forest

Posted 17 February 2019 - 04:43 PM

Lets examine a couple of cases.

 

CGE PRO versus HDX110/EQ8

 

The CGE PRO has a servo motor = [encoder-motor-9.9-1 gearbox]-> ~4-1 transfer gears->.worm shaft (RA & Dec)

 

HDX110 has a stepper motor = motor -> 1-1 transfer pulley -> worm shaft (RA)

 

HDX110 has a stepper motor = motor  -> worm shaft (Dec)

 

The simplicity of the HDX110 is evident. The CEM120 has a very similar drive arrangement to the HDX110 RA drive. To achieve higher accuracy position and speed control for the stepper motor(s) the CEM120 has optional high res encoders and the total cost of the CEM120EC mount is roughly equal to the pre-sale cost of the CGE PRO.

 

Hypothetically 

 

We fit an AP 1100 with an HDX110 stepper drive system and we add the AP absolute encoders. 

 

Would the above mount be less accurate than with the stock AP servo drive with absolute encoders? 

 

Could AP deliver the AP1100 with stepper drives and absolute encoders at the same cost as the current AP1100 servo drive system?


Edited by DuncanM, 17 February 2019 - 04:45 PM.


#9 Pauls72

Pauls72

    Vanguard

  • *****
  • Posts: 2341
  • Joined: 28 Oct 2007
  • Loc: LaPorte, IN USA

Posted 17 February 2019 - 09:25 PM

At our level of products, either will work just fine in a properly designed mount.


  • 555aaa likes this

#10 rgsalinger

rgsalinger

    Skylab

  • *****
  • Moderators
  • Posts: 4453
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 17 February 2019 - 09:35 PM

I own two mounts. One uses steppers, one uses brushless servos. There's no practical difference between them in terms of performance but neither mount is more than 4 years old at this point. 

Rgrds-Ross


  • OldManSky likes this

#11 gnowellsct

gnowellsct

    Hubble

  • *****
  • Posts: 15174
  • Joined: 24 Jun 2009

Posted 19 February 2019 - 07:58 AM

Servos stall and burn out. Yes there is software protection but it doesn't always work. Steppers are very uncomplaining. If they encounter a severe imbalance or hit against something they just stop.

About 98% of the market is servos. If you like steppers you're weird like someone who drives a manual clutch.

I have a manual Accord and prefer steppers.
  • elmiko likes this

#12 gnowellsct

gnowellsct

    Hubble

  • *****
  • Posts: 15174
  • Joined: 24 Jun 2009

Posted 19 February 2019 - 08:00 AM


Could AP deliver the AP1100 with stepper drives and absolute encoders at the same cost as the current AP1100 servo drive system?


Dunno but I bought a 1993 used astro physics 900 qmd because of my allergy to servos.

#13 zakry3323

zakry3323

    Viking 1

  • *****
  • topic starter
  • Posts: 848
  • Joined: 11 Apr 2016
  • Loc: Pittsburgh

Posted 19 February 2019 - 08:26 AM

Servos provide more precise control because by definition they use a motor mounted encoder to control shaft angle. A stepper motor used on the RA axis is used in microstepping mode and cannot provide precise control. Servos provide torque proportional to current and can be used in clutchless mounts to sense balance. Servos can provide smooth torque delivery over a higher speed range and have much lower to zero cogging torque. Steppers are by design "all cogging."That is an advantage if you need to provide holding force when no power is supplied. Steppers are cheaper and have cheaper control electronics typically.

Is this why mounts within my range with servos are able to move to a "home position" and park after a session, whereas mounts with steppers cannot, unless absolute encoders are also used (bumping up the pricetag considerably)?  



#14 OldManSky

OldManSky

    Viking 1

  • -----
  • Posts: 831
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 19 February 2019 - 09:32 AM

Is this why mounts within my range with servos are able to move to a "home position" and park after a session, whereas mounts with steppers cannot, unless absolute encoders are also used (bumping up the pricetag considerably)?  

My lower-price-tag CEM25P, with steppers, is able to move to a "home position" and park after a session.

Doing so on the rare occasions (lately) that I could leave it setup for more than one night means on night #2, I just turn it on and go.  No star alignment needed.  No absolute encoders needed.


  • elmiko and zakry3323 like this

#15 zakry3323

zakry3323

    Viking 1

  • *****
  • topic starter
  • Posts: 848
  • Joined: 11 Apr 2016
  • Loc: Pittsburgh

Posted 19 February 2019 - 09:50 AM

My lower-price-tag CEM25P, with steppers, is able to move to a "home position" and park after a session.

Doing so on the rare occasions (lately) that I could leave it setup for more than one night means on night #2, I just turn it on and go.  No star alignment needed.  No absolute encoders needed.

I wonder how it's able to do this? Did you have to input a home position during your initial setup? 



#16 gotak

gotak

    Surveyor 1

  • -----
  • Posts: 1804
  • Joined: 18 Sep 2016
  • Loc: Toronto, CA

Posted 19 February 2019 - 10:01 AM

The advantages of servos are overstated around here. 

 

Steppers are more or less the better solution for open loop control. Yes servos have an encoder but how accurate are those anyhow? 

 

The other big advantage of steppers are that you don't have to care about pier contact for the most part, all that would happen is the mount would stall. Nothing would burn out etc.

 

All said as mentioned it doesn't really matter as long as it works.


  • Skymind likes this

#17 DuncanM

DuncanM

    Mercury-Atlas

  • -----
  • Posts: 2502
  • Joined: 03 Nov 2009
  • Loc: Arizona Sky Village or the rain forest

Posted 19 February 2019 - 10:26 AM

Is this why mounts within my range with servos are able to move to a "home position" and park after a session, whereas mounts with steppers cannot, unless absolute encoders are also used (bumping up the pricetag considerably)?  

All GOTO mounts with open loop steppers such as the EQ-5 are able to move to predefined positions as long as the stepper control system hasn't lost track (counting steps) of the motor - this isn't much different than the servo control system in a low cost servo mount, such as an AVX. I would suspect that the AVX would have slighter more accurate positioning due to the feedback from the servo encoders. However, some stepper motor mounts (and some servo motor mounts including the CGE PRO) such as the CEM60 and 120 have home position (Zero position) sensors, and thus the mount can reorient itself by finding the sensor to reestablish a point of reference.


  • zakry3323 likes this

#18 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 15458
  • Joined: 27 Oct 2014

Posted 19 February 2019 - 11:10 AM

I wonder how it's able to do this? Did you have to input a home position during your initial setup? 

I've had a couple of iOptron mounts.

 

It automatically saves where the zero position was last time.  It saves the sky model that was generated by aligning to one or more stars last time.  It also has a pretty good default sky model that you can access by the operation "clear alignment data".

 

It does it with my CEM60 (steppers, no absolute encoder).   Since I rely on platesolving for precise GOTO, I never star align my CEM60 any more.  I do manually reset the zero position when I've disengaged the worms for balancing.  There is a "find zero position" button, I've never gotten it to work well and fast enough that just eyeballing zero position isn't better.

 

Servos/steppers is a Ford v Chevy deal.  People get passionate and negative about one or the other, but generally incorrectly so.  Neither is "bad".  The issues with modest mounts are the gears and software, not the motors.

 

Don't buy a CEM60 used, unless you're confident it's not one of the early models.  I waited a couple of years to buy mine, until the initial reports of issues went away.


Edited by bobzeq25, 19 February 2019 - 11:23 AM.

  • zakry3323 likes this

#19 rgsalinger

rgsalinger

    Skylab

  • *****
  • Moderators
  • Posts: 4453
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 19 February 2019 - 12:26 PM

As to the potential accuracy of servos, I don't think that every Paramount would come with them if they were somehow inferior to steppers. Now those are expensive brushless servos. AP says .05 arc seconds http://www.astro-phy...ch1gto/mach1gto here. I can't speak for a dollars to dollar comparison. What I can say is that my Paramount with servers and my CEM120EC2 work just as well as you would hope they would, each using a different type of motor. Still, when you get to less expensive motors I can't comment.

Rgrds-Ross



#20 OldManSky

OldManSky

    Viking 1

  • -----
  • Posts: 831
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 19 February 2019 - 12:47 PM

I wonder how it's able to do this? Did you have to input a home position during your initial setup? 

You do have to set a home position.  Just once.  You move the mount to what you want the home position to be, then use the hand controller to "Set Home Position."

 

Once that's done, you can park it at the home position, turn off the power, and come back anytime later and just turn on and go.  As long as you don't move the mount. :)


  • zakry3323 likes this

#21 rgsalinger

rgsalinger

    Skylab

  • *****
  • Moderators
  • Posts: 4453
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 19 February 2019 - 01:02 PM

It's also necessary that the time be correct when you start up any mount. If it's not then the mount's ability to translate between alt/az (home position) and DEC/RA is compromised. A second won't matter but a few minutes will. I routinely check the time on my Paramount system and actually push the internet time into both of the other systems that I use. Timebase differences are also one reason why that first goto every night can be a bit off. With most systems this is fixed by a plate solve and a sync. 


  • zakry3323 likes this

#22 zakry3323

zakry3323

    Viking 1

  • *****
  • topic starter
  • Posts: 848
  • Joined: 11 Apr 2016
  • Loc: Pittsburgh

Posted 19 February 2019 - 01:51 PM

 

Don't buy a CEM60 used, unless you're confident it's not one of the early models.  I waited a couple of years to buy mine, until the initial reports of issues went away.

I very much appreciate this advice! A CEM60 is one of the two mounts that I've narrowed down my search to!



#23 DuncanM

DuncanM

    Mercury-Atlas

  • -----
  • Posts: 2502
  • Joined: 03 Nov 2009
  • Loc: Arizona Sky Village or the rain forest

Posted 19 February 2019 - 02:41 PM

I've had a couple of iOptron mounts.

 

It automatically saves where the zero position was last time.  It saves the sky model that was generated by aligning to one or more stars last time.  It also has a pretty good default sky model that you can access by the operation "clear alignment data".

 

It does it with my CEM60 (steppers, no absolute encoder).   Since I rely on platesolving for precise GOTO, I never star align my CEM60 any more.  I do manually reset the zero position when I've disengaged the worms for balancing.  There is a "find zero position" button, I've never gotten it to work well and fast enough that just eyeballing zero position isn't better.

 

Servos/steppers is a Ford v Chevy deal.  People get passionate and negative about one or the other, but generally incorrectly so.  Neither is "bad".  The issues with modest mounts are the gears and software, not the motors.

 

Don't buy a CEM60 used, unless you're confident it's not one of the early models.  I waited a couple of years to buy mine, until the initial reports of issues went away.

I now control my mounts from the comfort of my family room via Chrome Remote Desktop.

 

Whilst experimenting and learning to use the plate solving engine in AstroArt 7 with my CEM60 I inadvertently generated a false solve on the current image (using a "blind solve") and synced the mount to that incorrect position... Doh!  The next goto ended up somewhere very far from where I wanted to be and since I had synced the mount to the wrong position any saved reference positions in the mount would also be incorrect...confused1.gif  The mount was also now aimed below the effective horizon...

 

No problem. I issued the "find Zero position sensors" command via the iOptron Commander interface and after a minute or so the mount was back in the Zero (home) position and the mount controller now knew where it was pointing. I did another goto to a target and then did another plate solve and I was back in business.


Edited by DuncanM, 19 February 2019 - 02:42 PM.


#24 gnowellsct

gnowellsct

    Hubble

  • *****
  • Posts: 15174
  • Joined: 24 Jun 2009

Posted 19 February 2019 - 05:17 PM

 

No problem. I issued the "find Zero position sensors" command via the iOptron Commander interface and after a minute or so the mount was back in the Zero (home) position and the mount controller now knew where it was pointing. I did another goto to a target and then did another plate solve and I was back in business.

Built in plate solving would be a bonanza for us primitives who still sight alignment stars and build models manually.  What would be nice is to do all this without needing a laptop (for field use).

 

I am puzzled at the statement (somewhere above) that servos are inaccurate.  I remember reading that servos divide the 360 degree circle into about 2 million parts--my Argo Navis divides the circle into 10,000 parts.  

 

Still thanks to the power of mount modeling I have 2 to 3 arc minute error with all sky pointing which is excellent for visual applications.  But my plate solving paramount owning buddy has all sky pointing accurate to about 2 to 3 arc seconds.  That's almost two orders of magnitude better.  

 

Anyhow my 2 to 3 arc minute error is outstanding for the level of tech I have, it used to be that dscs were accurate to about 30 arc minutes.   For people who used refractors with 3 and 4 degree (and larger) fields of view and looked mainly for bright Messiers because using a four or five inch apo, that was fine.  For a guy with a C14 and a 40 arc minute field looking for 13th and 14th magnitude off-the-beaten-path objects it was frustrating in the extreme.  With 2 to 3 arc minutes pointing error you objects will fall within the field of view of a ten arc minute eyepiece (such as an XW10 in the C14) and so you're in business.  If you can't find it, it is likely not there to be found or too dim to be seen.

 

One of the issues with servos is that starting the slew can cause some slippage between scope and mount in the clutch.  You can try to finesse that with an adjustment to the slewing software but it is a factor that servos can be so powerful that they accelerate faster than the rest of the telescope.  

 

I haven't had a slewing stall since about 2004 when I got my AP900QMD push to and I've been much happier for it.  I had a very rude introduction to go-to that did not go well for me.  

 

I decided after that if I went to go-to again it would be an astrophysics or a parmount.  Paramount doesn't really provide a smart hand paddle, meaning you have to have a laptop for control.  Astro-physics won't build mount modeling into their paddle, if you want modeling, you need a laptop.  Astro-physics wants you to rely on forced star aligns in the sector in which you are searching which is a late 1980s solution to the problem of pointing inaccuracy.  We are so far beyond that.  Everywhere except AP.  And even AP has modeling if you want to run a laptop.  They decided their visual users who want modeling in the paddle are an extremely small market segment.  Maybe they're right.

 

I don't want to run a laptop because they consume a lot of power and become one more gizmo I have to run, and one that is not really built for field use.  So Argo Navis it is.  1/2 amp or less power draw, built in modeling, unobtrusive display.  And it never ever stalls because it is push to.

 

Greg N


Edited by gnowellsct, 19 February 2019 - 05:23 PM.


#25 goldtr8

goldtr8

    Explorer 1

  • -----
  • Posts: 67
  • Joined: 29 Apr 2018

Posted 24 February 2019 - 09:19 AM

I'm about 3/4 of my way to being able to afford a new (maybe presently affordable used, though I have a lot of reservations about buying a used), I suppose what is considered a mid-tier (around 50lbs capacity) mount, upgrading from an AVX primarily for astrophotography use. I've narrowed my selections down considerably over the last 8 months that I've been able to save for it. 

Besides audible noise, slewing speed, and torque, are there any other considerations I should be thinking about when when thinking about motors?
 

What are the reasons you want to upgrade from your AVX.   I am considering purchasing one so I would like to know what you see as limitations.


  • zakry3323 likes this


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