Return to the Cloudy Nights Telescope Reviews home pageAstronomics discounts for Cloudy Nights members
· Get a Cloudy Nights T-Shirt · Submit a Review / Article

Click here if you are having trouble logging into the forums

Privacy Policy | Please read our Terms of Service | Signup and Troubleshooting FAQ | Problems? PM a Red or a Green Gu… uh, User

Equipment Discussions >> Mounts

Pages: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | (show all)
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: Starhawk]
      #5779818 - 04/05/13 02:57 PM

I think the problem is in the interpolation somewhere.

I bought some more IEEE articles but they are beyond my understanding. That's what comes of not being practicing EE for 15 years.

I thought of a way to "sanity check" the interpolation. By wiring comparators to the A and B analog inputs, I can now get A and B digital signals. With quadrature decoding this is a ~ 65" accuracy every time one of these inputs change.

So whenever there is a rising or falling edge on the A and B, I know that the interpolated angle must be 0, 90, 180, or 270 degrees (for sure! to within the limit of the encoder) and I can use that information to check the interpolation algorithm.

Later this weekend.. I have to finish my SAO exam..


Post Extras: Print Post   Remind Me!   Notify Moderator  
Starhawk
Space Ranger
*****

Reged: 09/16/08

Loc: Tucson, Arizona
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5781774 - 04/06/13 12:33 PM

Perfect data doesn't help if the controller doesn't work. I expect it is struggling with backlash, reversals, and slope rates. Your encoder agrees pretty well with a completely independent measurement, so I am thinking your encoder works well enough already.

-Rich


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: Starhawk]
      #5782065 - 04/06/13 03:36 PM

I wired up the comparators and triggered a couple interrupts whenever there is a zero crossing. And I found that the interpolated values were in agreement to within 0.8" of the zero crossing.

So the interpolation algorithm is only capable of about between 162X and 324X. This should still be enough - if we assume 162X as the worst case, this means we get 810K ticks, or 1.6" per tick.

That does mean that I should not correct any error < 1.6" because it might be interpolation error. Which in turn would limit the performance of the device. But I'm not getting anywhere near 3.2" p-p so I need to look at the control algorithm...


Post Extras: Print Post   Remind Me!   Notify Moderator  
Starhawk
Space Ranger
*****

Reged: 09/16/08

Loc: Tucson, Arizona
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5784258 - 04/07/13 04:33 PM

Here's where the fun starts!

-Rich


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mert
Post Laureate
*****

Reged: 08/31/05

Loc: Spain, Pamplona
Re: Encoder-based PE Correction on the cheap new [Re: Starhawk]
      #5786403 - 04/08/13 04:57 PM

Hi Orlando!

I've read that you are using the MCP3304 ADC is that true?
In the specs I see that the reference voltage should be
controlled very closely in order to get 13 bit resolution.
The encoder output sin/cos-wave is -1 to +1 V?
Then also from the datasheet, "Conversion speed (FSAMPLE) is
100 ksps with FCLK = 21*FSAMPLE"!
Are you using the clock signal from the Arduino? ( can't see
any external clock driver on the breadboard )

Just a couple of questions running through my head trying to
uderstand what's going (wr)on(g).

Keep up the great work


Post Extras: Print Post   Remind Me!   Notify Moderator  
vdb
sage


Reged: 12/08/09

Re: Encoder-based PE Correction on the cheap new [Re: Raginar]
      #5794517 - 04/12/13 01:52 PM

Are you sure the physical mounting of the encoder is not introducing some error as well?
As with many of these kind of projects, after initial progress a blockage of some kind is encountered, time will solve this, I'm sure ...


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: vdb]
      #5795734 - 04/13/13 12:51 AM

Hi Mert, my LM385 reference is supposed to be 1.24V but my multimeter says 1.2V or so.

Doesn't matter, when I try to get the full 13-bit resolution the last 4 bits or so are jittering. There's no point trying to get the full span given that there are no input analog buffers. Also the encoder output is not 1Vpp, it's more like 0.85Vpp (according to my ADC) but that is still within spec for the encoder.

More importantly, the encoder's Lissajous pattern is almost perfectly circular, and has zero offset. I've added some algorithm to throw out any encoder readings whose radius is not within 3 sigmas of the average encoder radius reading (as 99.7% of encoder readings are within 3 sigmas).

I also noted that, from back-calculating the angle when the sinusoid input reverses direction (used a comparator to detect the zero crossing) that there is about 0.8" of jitter in the interpolated angle, even after all the pre-processing.

What this means is with the current interpolation electronics and code, 0.8" is the best resolution possible (324X interpolation) which means about 1.6" is the best tracking accuracy that can be obtained. I am not surprised by this, as the much more expensive and complex Heidenhain IBV660B does 400X. If I can get even 162X (1.6") I would be extremely happy.

Have been traveling a lot so no updates here. Will try to get another measurement run in over the next few days.


Post Extras: Print Post   Remind Me!   Notify Moderator  
fetoma
Carpal Tunnel
*****

Reged: 09/26/06

Loc: Southern New Jersey
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5796768 - 04/13/13 03:03 PM

Would this help:

http://www.ebay.com/itm/Heidenhain-IBV-660B-Interpolator-Converter-Box-/12108...


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: fetoma]
      #5796982 - 04/13/13 04:19 PM

I have one of those, Frank.

I am consciously not using them because they cost 1000 Euro ($1300) new. So any solution using that box would be useless to the rest of the world.

I want to get my cheap interpolator box working so that other people can duplicate the results cheaply. I don't want a solution that depends on getting expensive equipment by scrounging. This isn't Storage Wars.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Raginar
Postmaster
*****

Reged: 10/19/10

Loc: Rapid CIty, SD
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5797413 - 04/13/13 07:59 PM

Hey Orly,

I noticed that the Gemini-2 can support an external encoder. You ever seen anyone figure that out?

Chris


Post Extras: Print Post   Remind Me!   Notify Moderator  
jrcrillyAdministrator
Refractor wienie no more
*****

Reged: 04/30/03

Loc: NE Ohio
Re: Encoder-based PE Correction on the cheap new [Re: Raginar]
      #5797613 - 04/13/13 10:43 PM

Quote:

I noticed that the Gemini-2 can support an external encoder. You ever seen anyone figure that out?

Chris




It is probably the same as the original Gemini in that respect. It can read typical push-to style encoders to recover goto alignment after a slipped clutch. They would have insufficient resolution for any tracking feedback.


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: jrcrilly]
      #5797801 - 04/14/13 02:19 AM

John is correct, the Gemini supports external encoders but only for pointing. These are 2500 CPR (10000 tick) encoders, which cost $50 a pop.

The Littlefoot Elegance Photo is the only GoTo controller I know of that can use an external encoder for both pointing and PE correction. But it requires a very high tick digital encoder.


Post Extras: Print Post   Remind Me!   Notify Moderator  
jrcrillyAdministrator
Refractor wienie no more
*****

Reged: 04/30/03

Loc: NE Ohio
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5797862 - 04/14/13 05:06 AM

Quote:

The Littlefoot Elegance Photo is the only GoTo controller I know of that can use an external encoder for both pointing and PE correction.




The SiTech can, too, using separate encoders for the two functions.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Raginar
Postmaster
*****

Reged: 10/19/10

Loc: Rapid CIty, SD
Re: Encoder-based PE Correction on the cheap new [Re: jrcrilly]
      #5798054 - 04/14/13 10:21 AM

I saw the thread on an individual with a sitech/mi250 who put a nice encode and was doing 20 min unguided exposures.

Pretty cool.


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: Raginar]
      #5798106 - 04/14/13 11:12 AM

Chris, they are using a 320K tick (digital) Gurley encoder which costs about $600 - $700. I could certainly use the same encoder and forget about interpolation. But it's not that interesting an approach..

Post Extras: Print Post   Remind Me!   Notify Moderator  
Starhawk
Space Ranger
*****

Reged: 09/16/08

Loc: Tucson, Arizona
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5798158 - 04/14/13 11:45 AM

Any chance to try new control algorithms? Is there any way you could build the software to have several available to try out in an evening?

-Rich


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: Starhawk]
      #5798298 - 04/14/13 01:19 PM

I've hooked up the encoder to my AP600, which has a much more tractable periodic error. Still no resolution in sight.

If I could hook the encoder to the Mach1 and run the Mach1 with PEM enabled, I could characterize the encoder error very quickly. But alas the Mach1 polar scope bore does not rotate.

I tried measuring the PE with the CGEM while it was guiding, but the CGEM's PE is so bad that even when guiding I could only get 2" RMS

If it clears up during the week, I'll try guiding the AP600 (which from experience I know guides much, much better) and measure the residual PE. That should give me an idea of what sort of error the encoder and algorithm has.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mert
Post Laureate
*****

Reged: 08/31/05

Loc: Spain, Pamplona
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5798591 - 04/14/13 03:23 PM

Hi Orlando,

Just for curiousity on how good your interpolator works,
would it be an idea to hookup the IBV660 just for a
comparison and see how the data looks?
I know your idea behind all this is an economical
solution for everybody


Post Extras: Print Post   Remind Me!   Notify Moderator  
orlyandico
Postmaster
*****

Reged: 08/10/09

Loc: Singapore
Re: Encoder-based PE Correction on the cheap new [Re: Mert]
      #5799653 - 04/14/13 11:35 PM

I have given thought to that Mert... maybe I'll do that when I've completely run out of ideas.

On the other hand... I'm beginning to suspect that the periodic error in the encoder is larger than 1 slot (i.e. 259.2 arc-seconds). In other words, the spacing between each of the 5000 slots is not perfectly consistent, because I'm seeing a very slow periodic error with a cycle of about 900 seconds. The IBV660B would not help with that since it solves the problem of interpolation between slots.

This actually makes sense because in its current state, the box smooths out small fast errors really well (much better than guiding) but does not correct long-term drift.

Things should be more clear when I measure the PE while guiding the AP600.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mert
Post Laureate
*****

Reged: 08/31/05

Loc: Spain, Pamplona
Re: Encoder-based PE Correction on the cheap new [Re: orlyandico]
      #5800394 - 04/15/13 12:14 PM

I see Orlando, of course the precision of the encoder
is important, but what I was worried about is your post
on page 8 of this thread ( #5777368 ) where we see the
raw PE of the GEM leaking through!!!
I can't understand how can this happen, since these are
so to say "high" frequency errors, far out of the scope
of the "PE" of the encoder IMHO

Just my thoughts, hope to read more


Post Extras: Print Post   Remind Me!   Notify Moderator  
Pages: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | (show all)


Extra information
17 registered and 34 anonymous users are browsing this forum.

Moderator:  Dave M, richard7, bilgebay 

Print Thread

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled


Thread views: 14783

Jump to

CN Forums Home


Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics