Topics

Alignment / Slewing problems


Russ W
 

I'm having loads of fun testing and tuning my new OnStep running on the Bluepill kit.  OnStep is 3.16d and the app is 2.21

I've been having issues with the go-to's -- sometimes. Seems like the Dec is off by 90 degrees, sometimes. Sometimes the counterweight has ended up higher than the tube! Pretty sure that's a rookie thing to do.

I have Home and Park set the same, pointed at Polaris more-or-less.

It seems to point in the right general directions while doing an align, but I haven't had it outside yet to be sure that's really the case.

I just "faked" an alignment on Capella and Vega, both of which were about equally distant from my meridian and somewhat north of due east and west (63 and 305 degrees, respectively). When I did a go-to to Deneb Kaitos-- south of me at 189 and below the ecliptic at 29 -- it pointed north......and straight down at my feet.

I have LV8729 drivers and 400 step motors, 3:1 belt-drive and 144-tooth worm gears on a GEM. Steps are 64 for tracking and 16 for go-to. Everything runs quiet (-ish), cool, and fast 3 degrees/sec. I recently cleaned out the 30-year-old Chinese grease and it's all balanced and turns super easy.

Did I muckle-up a config somewhere? Wiring?

Teach me, please!


Howard Dutton
 

Post the config.h file.


Howard Dutton
 


Khalid Baheyeldin
 

First, the obvious things: did you enter the date/time and location correctly?
The easiest way is to do it from the Android App. Don't forget to press Upload.

Then, did you notice any high pitch sound coming from the motors are certain speeds (as OnStep accelerates), with no movement?
If so, this is the stepper motors stalling.

Another thing, did you test it indoors to make sure that the direction of movement is correct (not reversed)?


Dave Schwartz
 

As Howard says, post the config file.

I have almost the same hardware... STM32, 400 step motors, 3:1 belt, 144-tooth worm, LV8729 drivers. The only real differences are I use 32 microstep tracking and 4 slewing. I did have to muck around with axis direction reversals to have it pointing correctly. I did it using the wiring harness between the driver and RJ port so I didn't have to keep rebuilding/reflashing repeatedly (this way I also don't have to remember to get the config setting right each time I do a fresh download).

On 2020-02-12 4:57 p.m., Russ W wrote:
I'm having loads of fun testing and tuning my new OnStep running on the Bluepill kit.  OnStep is 3.16d and the app is 2.21

I've been having issues with the go-to's -- sometimes. Seems like the Dec is off by 90 degrees, sometimes. Sometimes the counterweight has ended up higher than the tube! Pretty sure that's a rookie thing to do.

I have Home and Park set the same, pointed at Polaris more-or-less.

It seems to point in the right general directions while doing an align, but I haven't had it outside yet to be sure that's really the case.

I just "faked" an alignment on Capella and Vega, both of which were about equally distant from my meridian and somewhat north of due east and west (63 and 305 degrees, respectively). When I did a go-to to Deneb Kaitos-- south of me at 189 and below the ecliptic at 29 -- it pointed north......and straight down at my feet.

I have LV8729 drivers and 400 step motors, 3:1 belt-drive and 144-tooth worm gears on a GEM. Steps are 64 for tracking and 16 for go-to. Everything runs quiet (-ish), cool, and fast 3 degrees/sec. I recently cleaned out the 30-year-old Chinese grease and it's all balanced and turns super easy.

Did I muckle-up a config somewhere? Wiring?

Teach me, please!


Russ W
 

Figured it out.

I got confused when checking how far the mount moved with one turn of the worm. It's NOT degrees of R/A and minutes of Declination. My notes are corrected now: there's one 144-tooth, and one 180-tooth worm gear.

Also I had to swap two wires to reverse one motor.

No wiki in the world can save me from my own carelessness!  At one point I had a clutch loosened just enough that it was slipping in one direction but not the other. That kept me busy flipping wires and re-flashing for a couple of days.

Thanks


Khalid Baheyeldin
 

On Wed, Feb 12, 2020 at 08:33 PM, Russ W wrote:
there's one 144-tooth, and one 180-tooth worm gear.
That is something new to me.
What mount is it?


 
Edited

there's one 144-tooth, and one 180-tooth worm gear.
That is something new to me.
What mount is it?
On some mounts, I guess you really have to count the gearing.  The Orion AstroView I've been working on has 128 teeth on the RA gear and 65 (I counted 3 times!) on DEC.  EDIT:  No--I had to count a fourth time-using the setting circle and trial by error.  It's 66 teeth, apparently!!!


douxj2003@...
 

I also encountered the same problem when testing 2.22q, and later found that it was the reason for the configuration. The following lines of code were changed to this:
#define AXIS1_DRIVER_MODEL TMC2208
#define AXIS1_MICROSTEPS 16
#define AXIS1_MICROSTEPS_GOTO 8
#define AXIS2_DRIVER_MODEL TMC2208
#define AXIS2_MICROSTEPS 16
#define AXIS2_MICROSTEPS_GOTO 8


If you follow the configuration, the problem you said will occur:
#define AXIS1_DRIVER_MODEL_TMC2208
#define AXIS1_MICROSTEPS_16
#define AXIS1_MICROSTEPS_GOTO_8
#define AXIS2_DRIVER_MODEL_TMC2208
#define AXIS2_MICROSTEPS_16
#define AXIS2_MICROSTEPS_GOTO_8


Khalid Baheyeldin
 

On Thu, Feb 13, 2020 at 09:56 AM, Richard Shagam wrote:
there's one 144-tooth, and one 180-tooth worm gear.
That is something new to me.
What mount is it?
On some mounts, I guess you really have to count the gearing.  The Orion AstroView I've been working on has 128 teeth on the RA gear and 65 (I counted 3 times!) on DEC.
Looks like it is similar to CG4 mounts, with 130 RA and 65 DEC. But not quite.

What I have seen is that:
All EQ6 type mounts are 180 teeth on both axes, and
All EQ5 type mounts are 144 teeth on both axes.

Never seen one with 144 on one axis, and something else on another.
What mount is it with 144/180?


Russ W
 

It's a Meade LXD-500 B (bronze) that I found at a second-hand store for $15

One turn of the R/A worm = 10 minutes on the setting circle. Six turns an hour times 24 hours = 144

One turn of the Dec worm = 2 degrees on the setting circle. 360 divided by 2 = 180

I hope I got those calculations right! It sure didn't work properly when I had them both at 144.


Wagner Alves da Silva
 
Edited

Olá eu tinha uma Meade lxd500b e instalei Onstep-goto nela e usei ela com sucesso não tinha nenhum erro de alinhamento perda de passo.
O que deve estar acontecendo e que  coroas dos eixos RA e DEC ambos possui 142 dentes deve ser por isso que voce está com problemas 
Eu desmontei a cabeça equatorial e contei os dentes da coroa 

Google translate:
Hello I had a Meade lxd500b and installed Onstep-goto on it and used it successfully it had no alignment error step loss.
What must be happening is that the crowns of the RA and DEC axes both have 142 teeth must be why you are in trouble
I took the equatorial head apart and counted the crown teeth

Em qui, 13 de fev de 2020 13:44, Russ W <rlw1138@... escreveu:

It's a Meade LXD-500 B (bronze) that I found at a second-hand store for $15

One turn of the R/A worm = 10 minutes on the setting circle. Six turns an hour times 24 hours = 144

One turn of the Dec worm = 2 degrees on the setting circle. 360 divided by 2 = 180

I hope I got those calculations right! It sure didn't work properly when I had them both at 144.

 

 


Russ W
 

I was afraid someone would tell me to take it all apart and start counting!!

I better get to work......


Khalid Baheyeldin
 

We've had a discussion some time back about how to get OnStep
to calculate the steps per degree for you.

I can't remember the final solution we came with, and whether it
was a custom sketch, or just a series of LX200 commands to do
the same. There are some commands that can get the raw number
of steps.

But if you make the RA axis perfectly parallel to the horizon (using
a bubble level), then you can start with the DEC pointing to the
horizon as well, then tell it (in OnStep) to move 90 degrees. You
can measure both with a bubble level that has bubbles at right
angle to each other.

Anyone remembers?


Russ W
 

StellarJourney.com has all the commands listed, yeah?


Khalid Baheyeldin
 

On Thu, Feb 13, 2020 at 05:09 PM, Russ W wrote:
StellarJourney.com has all the commands listed, yeah?
I don't think the site is up to date with the exotic/obscure commands that are only used for debugging and such.

But I did try to use them once for backlash calculation. Didn't get that project completed.

For the RA axis
:GXF8#

For the DEC axis
:GXF9#

I do have a Python framework for OnStep if you are so inclined.


 

On Thu, Feb 13, 2020 at 09:26 AM, Russ W wrote:
I was afraid someone would tell me to take it all apart and start counting!!

I better get to work......
If you can turn the worm gear, you really don't have to take anything apart.  Just turn by hand and keep turning and counting. Helps if the mount has setting circles 


George Cushing
 

You will find that the various mounts in the "EQ-3/4" class unlike the EQ-5s have various tooth counts. Pretty uniformly Dec is about half of RA.


George Cushing
 

Meade at first sold a Hino "Mizar" as the 500 and then moved production to Linn Optics in Taiwan. Mizars were black, Linn's were gray.