Jump to content


Photo

Servo vs Stepper

  • Please log in to reply
86 replies to this topic

#1 Stew57

Stew57

    Mercury-Atlas

  • *****
  • Posts: 2537
  • Joined: 03 May 2009
  • Loc: Silsbee Texas

Posted 19 August 2012 - 02:27 PM

Interesting article on cheaper servos used in the CGEM
http://www.photodady...t-motor-drives/

#2 rmollise

rmollise

    Hubble

  • *****
  • Posts: 15562
  • Joined: 06 Jul 2007

Posted 19 August 2012 - 03:09 PM

Interesting, yeah, I guess. He says "You get what you pay for," but it doesn't sound like he knew what he was buying. Anybody expecting outstanding UNGUIDED performance in this mount or any of the others in its class has misplaced expectations. ;)

You guide and you take good pictures and you are aware you don't own an AP or Bisque.

#3 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5457
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 19 August 2012 - 03:14 PM

Meh. This is why I chose stepper drive for my GoTo conversion - because my CGEM experience has really soured me on servos.

Steppers suffer from a lack of dynamic range (if you want precision, you can't get fast slews). My GoTo conversion is limited to 160X sidereal, which is pretty slow but surprisingly can be lived with.

Steppers are generally easier to work with, the circuitry can be simpler, etc - but they are open-loop. However one can buy steppers which ALSO have an encoder. I believe Taks do this.

#4 Stew57

Stew57

    Mercury-Atlas

  • *****
  • Posts: 2537
  • Joined: 03 May 2009
  • Loc: Silsbee Texas

Posted 19 August 2012 - 03:21 PM

Agree I can't fully support his conclusion especially if you factor price/performance. He does reveal the weak link in the CGEM as the cheap servos correctly. They are the culprit behind the cogging effect, 8/3 harmonics, and generally what ails the CGEM!

For the price I am quite content with my current CGEM but just wanted to set the record strait on the knock against the Atlas.

You don't always get what you pay for in this day and age, but you never get more. That is where most of us make our mistake.

#5 photodady

photodady

    Lift Off

  • -----
  • Posts: 7
  • Joined: 03 Sep 2012

Posted 04 September 2012 - 10:18 AM

Thanks for using the link to my blog, I'm responding here because of the number of hits I received from this forum. I agree with most of the posts here. First and foremost I am a photographer FIRST, not an astrophotographer. I am relatively new to this area but I must say that I am a perfectionist and my background is in engineering and an aircraft technician.

I do use a guide scope system but I come from a perspective that any system used to improve upon, using any control system compensations, should ONLY be applied AFTER mechanical system optimizations. That being said I have found the CGEM to be more than adequate for short exposure astrophotography and for visual use. Even when using the guide scope it is still necessary to use software post-processing for star alignments.

Would you find it acceptable for your car to immediately veer to the ditch when you let go of the wheel? Of course not, you would you expect to be able to let go of the wheel for a few seconds without peeing your pants?

In my opinion the CGEM falls into the category of it being necessary to wear diapers (guide system) in order to make it acceptable for astrophotography for anything other than multi-short exposure or lunar photography.

Simply take a look at Celestron's pro mount and you will see that the RA gear is twice the diameter of the CGEM mount, the same is true for ANY other RESPECTABLE astrophotography mount. (period) The Bisque Paramount actually uses an eleven inch gear thus providing at least THREE times as much precision possible from the CGEM mount. Please show me ANY published high end fine art photos that were taken using a CGEM, the vast majority of them have been taken wit an AP mount.

The tradeoff here is, of course, CASH. So for the money the CGEM (or Orion Atlas) is as good as it gets - for the money. The Losmandy G-11 with an upgraded worm gear (and twice the RA gear diameter) will cost upwards of four thousand US dollars, the Celestron pro more than that.

Again, thanks for using the link to my site and I wish upon all – clear and dark seeing.

If there is something here that I am missing please don't hesitate to correct and/or educate me. :)

#6 austin.grant

austin.grant

    Viking 1

  • *****
  • Posts: 667
  • Joined: 18 Oct 2010
  • Loc: Shreveport, LA

Posted 04 September 2012 - 10:58 AM

Unfortunately I am finding out that there are basic design flaws that will most likely prevent me from taking this mount from consumer grade into the level of precision that might be called laboratory or research grade.


.... :foreheadslap:

#7 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2545
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 04 September 2012 - 11:05 AM

Just a few things.

First, guiding is not using diapers, it is a common method for AP that is used by many people with much more expensive mounts.

I think that you would find that there are some Takahashi mounts that are considered excellent for AP use even though they do not have particularly large worm wheels in them. Large worm wheels are great (I think the old rule was something like the diameter of the RA wheel should be the same as the diameter of the OTA being used), but precision engineering and machining play a very important (if not more important) part as well. It won't matter if a worm wheel is 3 times the size, if it's quality is poor. Hauling around a mount with 11 inch worm wheels tends to be a bit of a pain. Thus, they are most often permanently mounted.

The CGEM/Atlas/EQ6 ARE respectable AP mounts within the confines of appropriate expectations. I'm not sure what you mean by "published" but you can look here www.jwalk.smugmug.com to see some pretty impressive images taken on a CGEM and there are many others out there doing both planetary and deep sky imaging with the CGEM and other mid-level mounts that is quite impressive. Some of them even using C11 OTAs. Most people who own equipment in the CGEM range don't feel the need to pay for or justify their equipment expenditures by "publishing" "fine art" images. I know people who have thrown $50K+ at the astroimaging problem and have taken some astounding images and gotten them published in some of the magazines. I know others who have spent a fraction of that amount to take images that are just as impressive but they don't worry about getting published. There is a significant element of operator/imager talent in imaging and post-processing that comes into play with AP. If all it took was a great setup to snap incredible pics, then all of us would be doing it.

For the record, the G11 does not have an RA worm wheel that is twice the diameter of the CGEM ring gear. Maybe about 50% larger, but still quite small in reality and it uses a worm that is at most 50% of the diameter of the CGEM worm.

There are pluses and minuses to everything. If you have the money to buy high end equipment and are really serious about the hobby, then I certainly recommend pumping that money into the economy. But the average person does not have to spend that much (comparitively) to do some great work and the CGEM/Atlas mounts can do it. Expectations are very important. It would appear that you went into the CGEM with unrealistic expectations and as a result, your results are not surprising.

#8 brianb11213

brianb11213

    Cosmos

  • *****
  • Posts: 9047
  • Joined: 25 Feb 2009
  • Loc: 55.215N 6.554W

Posted 04 September 2012 - 12:36 PM

Simply take a look at Celestron's pro mount and you will see that the RA gear is twice the diameter of the CGEM mount

Look, there's no way that I'm claiming that the CGEM is the greatest mount in the world, but the worm gear diameter isn't the big issue. It's the tolerance of the parts. There's a lot of backlash in there, and that makes work for the autoguider.

The worm gear diameter may have an issue on how much torque the drive can supply without wearing the worm/gear at an excessive rate but this has little to do with precision.

#9 jrcrilly

jrcrilly

    Refractor wienie no more

  • *****
  • Administrators
  • Posts: 33792
  • Joined: 30 Apr 2003
  • Loc: NE Ohio

Posted 04 September 2012 - 12:45 PM

Simply take a look at Celestron's pro mount and you will see that the RA gear is twice the diameter of the CGEM mount, the same is true for ANY other RESPECTABLE astrophotography mount. (period)


Don't read too much onto the worm wheel diameters. There have been plenty of poorly-performing mounts with great big gears in them. Bear in mind that whenever Synta makes a knockoff of a high-quality mount they retain the gear diameters - but not the performance. If you want a high-quality, precision version of the Atlas/CGEM it's available, and was for years before the knockoffs happened. Just go back to the original - the EM-200. Same gear diameters and ratios, but +-3.5 arcsecond tracking out of the box. It would cost far more than the price difference to upgrade the Chinese mounts to compete with the originals in performance.

#10 Stew57

Stew57

    Mercury-Atlas

  • *****
  • Posts: 2537
  • Joined: 03 May 2009
  • Loc: Silsbee Texas

Posted 04 September 2012 - 12:46 PM

The rate of change in the PE is the determing factor in guiding. The diameter og the worm and worm gear may affect this but the precision of the machining determines this. There are work arounds for backlash. Balance and guiding in one direction only for the DEC can help. I took string and affixed it to the RA with a weight on it. Keeps a constant amount of torque on the RA axis no matter the position of the RA axis. Keeping a slight east heavy balance does the same thing.

While backlash can be dealt with, the servos in the cgem are stilled handled as steppers at low voltage/speeds. It makes the servos not more accurate than the steppers used in the Atlac hence my reference to the article. many erroneously believe the Cgem is more accurate and tracks better (no stepper jumps) than the Atlas. This article should dispell that.

#11 Lorence

Lorence

    Viking 1

  • -----
  • Posts: 843
  • Joined: 15 Sep 2008

Posted 04 September 2012 - 12:52 PM

Interesting article on cheaper servos used in the CGEM


You might find this interesting.

http://www.siderealtechnology.com/

No idea if the controller would be practical in a CGEM but it looks like it will control an LX200. Whether the end result would be worth the effort is unclear but I'm thinking about trying one.

#12 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2545
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 04 September 2012 - 01:15 PM

Here's an excellent example of the CGEM with a C11HD being used for planetary work: http://www.flickr.co...n/photostream/.

#13 EFT

EFT

    Vendor - Deep Space Products

  • *****
  • Vendors
  • Posts: 2545
  • Joined: 07 May 2007
  • Loc: Phoenix, AZ

Posted 04 September 2012 - 01:18 PM

Many erroneously believe the Cgem is more accurate and tracks better (no stepper jumps) than the Atlas.



I agree. Little or no difference in regards to accuracy and tracking. The positives and negatives of both designs even them out.

#14 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5457
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 04 September 2012 - 01:28 PM

A stepper that is micro-stepped is probably just as smooth as a servo. My current Go-To conversion microsteps at 1/64 - the biggest lumps in the fundamental are due to the gearbox.

And we all know the famous CGEM gearbox with its 8" amplitude 8/3. For comparison, the (surplus, 33-buck-but-220 new) steppers in my Go-To conversion have about 4" p-p gear noise.

At least, I cannot get the native PE of 15" p-p below about 5" using PEC - so I'm assuming that's due to the gearbox.

#15 psandelle

psandelle

    Viking 1

  • -----
  • Posts: 660
  • Joined: 18 Jun 2008
  • Loc: Los Angeles

Posted 04 September 2012 - 01:35 PM

People with far more experience and bigger brains than I are already answering for the CGEM's ability to do AP (I've seen wonderful work with them), but I will make comment on the analogy thrown out by photodady, which I feel is a misleading one:

"Would you find it acceptable for your car to immediately veer to the ditch when you let go of the wheel? Of course not, you would you expect to be able to let go of the wheel for a few seconds without peeing your pants?"

For what they do, CGEMs do NOT "veer off the road" for visual, or for short focal length AP, or even when worked by a "master" for longer focal length AP. I think a more accurate analogy in the automotive world would be comparing a Toyota Tercel (I used to have one) and a Ferrari (used to have them, too - let's say a 328GTS, which is what I had when I had my Tercel).

Neither veered off the road (in general) when the wheel was let go (at least not if I had them balanced correctly). They both went from point A to point B very well (let's analogize that to GOTO's). And they both were fun to drive in their own way (let's call that short focal length AP). BUT, take them to the track, and the Toyota could not compare to the Ferrari. One would have to add a lot of "extras" (just like one would have to do for a CGEM) to get the Toyota even close to keeping up with the Ferrari, and, in the end, you get what you pay for.

So, there, I think, is a more accurate analogy. Right now, I have a Volvo station wagon because neither the Toyota nor the Ferrari could fit all the astro gear I have when I head to a darksite.

Paul

#16 jrcrilly

jrcrilly

    Refractor wienie no more

  • *****
  • Administrators
  • Posts: 33792
  • Joined: 30 Apr 2003
  • Loc: NE Ohio

Posted 04 September 2012 - 01:41 PM

many erroneously believe the Cgem is more accurate and tracks better (no stepper jumps) than the Atlas. This article should dispell that.


Stepper jumps were a concern decades ago. The original Vixen Sky Sensor (NOT the Sky Sensor 2000) used steps that were too coarse for the drive ratio and caused "jumps". Ever since, manufacturers have known enough to avoid the issue by selecting step sizes appropriate to the gearing. It's just not a concern today.

That said, PWM driven servos are now used instead on most mounts, from the bottom of the range to the top. The article implies that this is a flawed concept, then proceeds to recommend two superior mounts using the same concept. It's not the scheme that limits peformance, it's the implementation. Being in the same price class, and from the same manufacturer, I'd expect the Atlas and CGEM to offer similar precision.

#17 ccs_hello

ccs_hello

    Fly Me to the Moon

  • *****
  • Posts: 6450
  • Joined: 03 Jul 2004

Posted 04 September 2012 - 08:29 PM

John C is right. Recent high-end GOTO stepper based is microstepping based. High-torque (thus requires high current to drive) hybrid perm-mag stepping motor is used. Since microstepping drive method with sinusoidal waveform is used, the motor is tracking just like a regular perm-mag DC motor but with a great precision because the movement is timed.

High torque also eliminates the need for a high gear-down mechanical gearbox thus less coffee grinder noise and slops caused by multiple gears.

However, the drive electronics is not so easy. It has to handle ramp-up and ramp-down on the speed to avoid missing a step. The firmware has to handle both the position and precision timing for tracking including necessary fine-tune adjustment.


P.S. changed the title to properly reflect the recent art, not the old stepper technology.


Clear Skies!

ccs_hello

#18 orlyandico

orlyandico

    Fly Me to the Moon

  • *****
  • Posts: 5457
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 04 September 2012 - 09:38 PM

I don't really see why there is this story floating around that precision microstepping is "hard" - you can do this with an Arduino and an A4988 driver, which is a single $10 chip.

#19 gnowellsct

gnowellsct

    Fly Me to the Moon

  • *****
  • Posts: 6756
  • Joined: 24 Jun 2009

Posted 04 September 2012 - 10:01 PM

Meh. This is why I chose stepper drive for my GoTo conversion - because my CGEM experience has really soured me on servos.

Steppers suffer from a lack of dynamic range (if you want precision, you can't get fast slews). My GoTo conversion is limited to 160X sidereal, which is pretty slow but surprisingly can be lived with.

Steppers are generally easier to work with, the circuitry can be simpler, etc - but they are open-loop. However one can buy steppers which ALSO have an encoder. I believe Taks do this.


Steppers are very tolerant, but they are not main stream any more. I am 100% stepper on my current mounts. But I do push-to not go-to. Greg N

#20 gnowellsct

gnowellsct

    Fly Me to the Moon

  • *****
  • Posts: 6756
  • Joined: 24 Jun 2009

Posted 04 September 2012 - 10:08 PM

in the automotive world would be comparing a Toyota Tercel (I used to have one) and a Ferrari (used to have them, too - let's say a 328GTS, which is what I had when I had my Tercel).


Paul


I understand. You would drive the Ferrari for 90 minutes and then it would have to be taken to the shop. While it was in the shop getting fixed for three or four months, you drove the Tercel. Then you drove the Ferrari again for 90 minutes till it broke.

I understand because my brother used to own a Lotus and a white Ford pick up. He would use the white pick up to tow the Lotus to the shop and then drive it while it was being fixed. From your choice of second vehicle, I infer you had a AAA membership in good standing.

Greg N

#21 gnowellsct

gnowellsct

    Fly Me to the Moon

  • *****
  • Posts: 6756
  • Joined: 24 Jun 2009

Posted 04 September 2012 - 10:21 PM

The tradeoff here is, of course, CASH. So for the money the CGEM (or Orion Atlas) is as good as it gets - for the money. The Losmandy G-11 with an upgraded worm gear (and twice the RA gear diameter) will cost upwards of four thousand US dollars, the Celestron pro more than that.

Again, thanks for using the link to my site and I wish upon all – clear and dark seeing.

If there is something here that I am missing please don't hesitate to correct and/or educate me. :)


Well, $3800 unless you're counting shipping. ($3300 plus $500 for an Ovision).

A lot of top high end work is done on planets with lower end mounts.

There is something to your argument; in fact, I think it's right. Nonetheless, the selection bias in saying this is what the people who are in the magazines use is apparent. There is a certain class of people who have the wherewithal to go high end on everything which makes their choices easy. The best CCD systems, Peltier coolers, best optics, best software, one can go north of $100k without much effort, but in any case, $20 to $50k.

The people working on the less expensive mounts have also made other decisions which may make them less competitive in high end astrophotography.

Many, however, are doing work which equals or exceeds the pics Palomar used to send to the textbooks. But if you compare everything pixel by pixel the high end equipment is sure sweet when it is in the hands of people who know how to use it.

Greg N

#22 psandelle

psandelle

    Viking 1

  • -----
  • Posts: 660
  • Joined: 18 Jun 2008
  • Loc: Los Angeles

Posted 04 September 2012 - 10:48 PM

Greg N - actually, I had a great Ferrari mechanic and bought the 328GTS new, so except for tune-ups/oil changes, I never had a minute's trouble for almost 8 years (never should have gotten rid of it). But I couldn't put groceries or anything in it, hence the Toyota. Now a Lotus!!! That thar's trouble! :)

#23 photodady

photodady

    Lift Off

  • -----
  • Posts: 7
  • Joined: 03 Sep 2012

Posted 04 September 2012 - 11:04 PM

The CGEM mount is what it is, no more, no less.

Interesting that the link you provided shows that most of the photos, that had the info provided, were captured using a AP 1200 mount. I do understand that the machining tolerances are important but being machined with equal tolerances the larger diameter gear WILL ALWAYS be more precise.

A single ten minute exposure will exceed, by far, ten one-minute stacked exposures. With AP mounts this IS easily doable without guiding.

I can easily produce a Horse-head nebula photo that rivals one shot from the Hubble derived solely from a dark frame exposure given enough time and effort post-processing and an example to work from. That is the difference between multi-short exposure composites and true long exposure photos that still retain intact data that might be scientifically useful. Most amateur astrophotography, that you see on the web, is merely an artistic rendering subjectively highly manipulated, aligned, and warped on several levels in order to achieve the final result.

The SOLE reason I posted my reviews in my blog is that there is a lack of TRUTHFUL REALISTIC reviews on the web that would possibly have prevented me from wasting my money on a CGEM. By the way if you are interested mine is for sale while I put my name on the AP waiting list.

In the mean time I will be activating my SkyX TPoint trial and will be posting extensive data on numerous tolerances (or lack thereof) in many aspects of the CGEM mount, watch my blog for updates probably sometime later this month or possibly early next month.

http://www.photodady.com/blog

The latest development from AP with precision encoders on the AP 3600:
"The measured tracking error was well below 1/2 arc second for a 20 minute time period" (only in the wildest dreams of a a CGEM user)
http://www.astro-phy...3600gto/3600gto

Good luck and clear seeing.

#24 Falcon-

Falcon-

    Aurora

  • *****
  • Posts: 4869
  • Joined: 11 Sep 2009
  • Loc: Gambier Island, BC, Canada

Posted 04 September 2012 - 11:42 PM

Most amateur astrophotography, that you see on the web, is merely an artistic rendering subjectively highly manipulated, aligned, and warped on several levels in order to achieve the final result.


Yes, true.

And so are most of the pretty pictures released for PR use from NASA, ESO, etc, and by those using mounts such as the AP3200 that may grace a magazine cover.

Astrophotography *IS* about the end result image, not scientific data gathering after all...

The SOLE reason I posted my reviews in my blog is that there is a lack of TRUTHFUL REALISTIC reviews on the web that would possibly have prevented me from wasting my money on a CGEM.


I get the strong impression that what is considered realistic in this case is very much a mater of perspective. Nearly the entirety of my imaging has been done on a CG-5 mount! For me a CGEM's failings would be a wonderful blessing! From my perspective I *expect* auto-guiding to be required to achieve 10-minute exposures. I would consider the CGEM well worth praise if it does allow guided 10 minute exposures at relatively long focal lengths.

I can understand being disappointed if you come from the perspective that no guiding should be necessary, but to my mind that is not a realistic expectation at this price point.


(BTW - While I do not consider myself the equal of the top-end astrophotographers, I still do have a lot to learn with processing especially, but I think I am also justified in using myself as an example that good AP can be done with "poor" mounts.)

Anyway, just to be clear I am not trying to be confrontational, just trying to point out the disparity of perspectives. :)

#25 jrcrilly

jrcrilly

    Refractor wienie no more

  • *****
  • Administrators
  • Posts: 33792
  • Joined: 30 Apr 2003
  • Loc: NE Ohio

Posted 04 September 2012 - 11:57 PM

I do understand that the machining tolerances are important but being machined with equal tolerances the larger diameter gear WILL ALWAYS be more precise.


A larger worm wheel diameter will provide greater torque but doesn't affect precision. For a given worm accuracy, it's the tooth count on the worm wheel that determines drive precision. The more teeth per 360 degree rotation, the less erroneous axis motion results from a given worm error.

A single ten minute exposure will exceed, by far, ten one-minute stacked exposures.


It'll certainly exceed the stack in terms of random noise - a real image killer. Otherwise the result is pretty similar so the noise is by far the most significant difference. Stacking is an essential part of CCD imaging because of its ability to cancel random noise.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics