DIY dobsonian GOTO and tracking

[Jim09]

Hello all, 

My recent attempts to build a ROR shed for AP is on hold due to several issues. Small garden , not enough open south sky are the major ones so now I'm thinking to motorize my Skywatcher 200p dobsonian to have GOTO and tracking. I'm more interested in the DIY route rather than buying something off the shelf.

So here is my plan so far.

1. Build a mount with 18mm plywood or MDF. I got a router, jigsaw , mitre saw and a table saw at home, plus I'm fairly competent with them so this part shouldn't be much of a problem.

2. Use a lazy susan bearing for AZ axis  and use flange bearings for Alt axis.

3. Two Nema 17 100:1 geared stepper motors with MXL timing belt and pulleys which gives 8:1 ratio, I already have MXL pulleys with me.

4. Arduino mega with motor drivers to drive the telescope, I have an arduino uno and a motor driver from another project so I can reuse them if uno is able to handle these computations.

That's the basic motorized mount I have in mind

Then for the short exposure AP I have a setup with hacked 1080p webcam and a adapter which I can use. I assume I will need another one for the finders scope as well.

For on board computing to connect with webcams and controlling  I can use a RPi or a small PC with stellarium.

 

My goal is to once I set this up in the garden I can remotely connect to the RPi or the PC and then control the mount.

 

So that's the whole idea I have in my mind which I have been thinking for last couple years ever since I bought the telescope.

 

Stepper motors and pulleys will give me 200 steps per revolution + 100:1 inbuilt gear ratio + 8:1 gear ratio with pulleys + micro stepping (this part I'm not sure how it works as I never dealt with stepper motors)

200x100x8 = 160000 steps per revolution without any micro stepping, I assume this will be sufficient for basic tracking to be used with planetary AP and observing?

 

Does that sound feasible at all or are there any obstacles I haven't considered?

Thank you

 

ps.

I came across this thread which has lot of useful information for me.

 

 

 

 

 

 

 

 

 

 

[Vroobel]

Hi Jim, 

After 2 years of playing with my astro-DIY I can say:

1. Definitely use plywood instead of MDF. Moreover, use higher quality plywood. 

2. Lazy Susan is OK, I can confirm that. But I can't advise in terms of Skywatcher Alt axis, I have Bresser Messier (Explore Scientific), so the idea is different. 

3. Regarding the stepper motor with gearbox included - just be careful, I found out over one year ago that they can have backlash in every single gear, so they will add themselves together giving possibly significant error. 

Basically I used GT2 timing belt. I had also only 139,000 (micro)steps per revolution. But when I started tracking I could see shaking - you can see that in my last video. I still don't know what was a reason, but please assume that the timing belt can be more elastic than you expect. Otherwise, you possibly will use small looped belt. Just don't be surprised. Worm gears are much better! That's why I reconstructed my drives.

4. I don't know the Arduino Mega, but I used firstly a clone of Arduino and it's accuracy wasn't good enough for astronomical algorithms (Jean Meeus). Other constructors say the Arduino is OK, but they work with paralactical mounts, not Alt-Az, so the calculations are simpler. My equations are long from left to right margin of HD display, sometimes even longer. That's I replaced the Arduino with a Raspberry Pi. And presently I have much more IO in GPIO. Earlier I thought about driving the motors by Arduino controlled by RPi, but changed  the mind. 

5. I have modified MS LifeCam Studio webcam which has too big resolution of sensor and too small pixels. It's good for planets I think, but not for everything. I have also DSLR camera, but without video feature, so I consider  purchase of even Canon body able to record video. 

6. I read about some Linux software allowing to take and sent pictures from any USB camera with setting up parameters of exposure. I will try to find it again and will send it to you. 

7. Presently I have two worm gears, Az is 896:1 and Alt is 897:1 and no other gears. My stepper motor work in 8 micro steps mode. So I have final accuracy of 0.9 arcsec/micro step and I can improve it even to 0.45 arcsec/micro step using 16 micro steps mode, but it's perfect enough for me. The huge worm gears gave me a stability, even with stronger blows of wind.  But the only 160,000:1in your case are not enough, you should be at least close to 1 million:1, so at least 4 micro steps, if not 8 ones. 

... 

If you are familiar with tools and like the DIY, and didn't begin anything yet, try to consider that:

http://astromaniak.pl/viewtopic.php?f=5&t=48957

It's fantastic Polish project, so use the Google Translate, but pictures says almost everything. 

I'm happy to read that my texts, pictures and videos are useful. That's the reason why I do it. 

Let me know your final solutions. 

Good luck! 

[Jim09]

Thanks very much for detailed post Vroobel 

Regarding the  large worm gear , I would struggle to do your way  because I don't have proper equipment, but I guess I have to rethink about gear combinations to get the gear ratio I want . As you said if the gearbox has backlash then whole thing fails.

The link to the other project looks really cool, thanks for that. 

[Jim09]

I have a question regarding stepper motors,

I'm sketching a gear system to achieve 2880:1 gear ratio before the stepper motor, which will be archived with three sets of pulley systems which has gear ratios of 45*8*8

There are multiple stepper motors available on the market which I'm not sure which one will be best for me, 

eg

Nema 17 0.9deg 400 steps 1.68A 0.44Nm

Nema 17 0.9deg 400 steps  2A 0.46Nm

Nema 17 1.8deg 200 steps  2A 59Ncm

Nema 17 1.8deg 200 steps  Bipolar 1A 13Ncm

 

I'm not familiar with stepper motors and how to choose the right one for this project ,

my worry is with my gear ratio will it take forever for the scope to goto the location before it starts tracking?

or should I forget about steppers and go with a DC motor and a separate encorder?

Is there any reason why I should consider 200 steps one over 400 steps one when they both have similar torque ratings? do these have different max speeds?

Thanks

 

[Vroobel]

11 hours ago, Jim09 said:

Nema 17 0.9deg 400 steps 1.68A 0.44Nm

Nema 17 0.9deg 400 steps  2A 0.46Nm

Nema 17 1.8deg 200 steps  2A 59Ncm

Nema 17 1.8deg 200 steps  Bipolar 1A 13Ncm. 

... 

my worry is with my gear ratio will it take forever for the scope to goto the location before it starts tracking?

or should I forget about steppers and go with a DC motor and a separate encorder?

Is there any reason why I should consider 200 steps one over 400 steps one when they both have similar torque ratings? do these have different max speeds? 

I'm a little bit surprised, because I have Nema 17 1.8deg 65Ncm, which is none of listed above. 

You are right in your worrying. If I use 8 micro steps mode the slew of telescope needs around 2 minutes to reach opposite Az position. I have a feature allowing to change the mode to full step, but then the set generates terrible noise, especially in Alt axis. Please remember the tube is like a bell, it is empty inside! 

I'd like to reach a speed around 60 seconds for full resolution, but it is a choice between the speed and the noise. The micro steps are always silent. 

I don't know if there is any benefit of usage the 400 steps motor. I thought about it earlier, before I made the worm gears, but I have a fantastic accuracy now, so I stopped the consideration. 

Don't change the idea, stay with the stepper motors. 

[Jim09]

Thanks Vroobel,

I just picked few examples from ebay for stepper motors, that's why you don't see the model you have in that list.

I'm worried about the noise, I've seen youtube  videos a terrible clanking noise coming from Nema 17 motors when they are horizontal, is that what you mean? or just the general motor whining noise? if it's the first one I will have to rethink as I don;t want to disturb my neighbours middle of the night.

[Vroobel]

I mean the noise generated by the set of stepper motor + worm gear + aluminium skids + tube which is empty inside, that is Alt axis. The Az axis is not so noisy.

But don't worry, I just work on 8 micro steps mode and equation of ramp for first and last 1500 steps of the slew. I use the Excel and its chart to see a ahape of the chart, then move the equation to Python script. The slew do the first 1500 steps with acceleration, then do for example 25,000 steps with last calculated speed (delay in ms), then do the final 1500 with deceleration.

I'm not ready yet, so won't write too much now, but I based my thinking and attepmts on this idea:

https://guy.carpenter.id.au/gaugette/2012/01/26/simplified-acceleration-model/

After many days of working on the table, the Excel, charts and Python I stopped working on tables and moved towards the Excel, equations and charts only. Finally the noise is acceptable, but 1/4 full slew in Az axis takes 50 seconds. Just a compromise.