Star Tracker For Misers - Because the money is better off in your pocket

[EwanV]

The SkyWatcher Star Adventurer 2 Pro is £335 at the time of writing.  I want to make a star tracker for a DSLR for under £35.  "Why?" I hear you ask.  "Because if it's made for astro, it's cheaper to make it yourself".  I want my design to be something that you can buy the parts for, but I want as much of it as possible to be 3D printable too.

 

There are already loads of designs for 3D printed star trackers - some designed to hold small scopes, tabletop ones, lightweight, etc.  I decided to follow a completely new approach - belts and gears!  Now for 3D printing, I decided to go for the GT2 tooth setting, but the important thing is the number of teeth on each gear.  For simplicity in design (and so that people can take advantage of buying multipacks of gears if they're not 3D printing them) I decided to go for a ratio of 4:1.  So the small gear has x teeth and the large gear has 4x teeth.  I've gone for 20 and 80, because that's big enough that you can see any errors in the final object and small enough to not take ages to print.  I made a pillar of the gears so that you have an 80t gear, then a hub with two holes for M4 grubs and the 20t gear on top.  Each one takes 2 hours to print on my printer.  People buying gears can go for whatever teeth they want, as long as they maintain the ratio.

 

So how many gears?

 

There are 86,164.0905 seconds in a sidereal day, so no matter what we need our last gear completing one revolution in that time (and only one revolution).  This final gear has 80 teeth and is being driven by the 20t gear on one of our towers.  That tower rotates 4x for every 1 of the final gear, so (86,164.0905/4) every 21,541.0226 seconds.  Adding in more gear towers gives us:

 

Final Gear - 86,164.0905 seconds

Tower 1 - 21,541.0226 seconds

Tower 2 - 5,385,25566 seconds

Tower 3 - 1,346.31391 seconds

Tower 4 - 336.578479 seconds.

 

Attaching a 20t gear to the motor and using it to drive the 80t gear on tower 4 gives us (336.578479/4) 84.1446196 seconds for the motor to complete one revolution or 0.713058 revolutions per minute.  You can specify the RPM of a stepper motor, so that's what we'll do. 

 

Please follow this thread and I'll be back with pictures, updates, file locations for printing everything and so on.

 

 

 

[EwanV]

Gears are ready. I’m using a metal 20t gear as I happened to have some spare. I just need to order some bearings, print the belts, design the belt tensioner design the case…..

[EwanV]

Design for the gear chasis!!  This is just a first run through so that I've got enough room around everything to get to the belts.  I'm going to replace the 5mm threaded bar (in the pictures above) with a smooth bar that I'll grind down to give the grub screws in the gears something to hold on to and then each bar will be held in place top and bottom by a 5x11x5 bearing (size chosen because a pack of 10 is less than a fiver).  The four gear towers will run one to the other and the final gear will, for the test, be the lone 80 tooth gear.  In the end, the spot where the final gear is will be removed and the motor mounted instead.  The final gear will run on a separate plate so that if something goes wrong with one of the belts/motors you don't have to dismantle everything to fix it - just take the chasis out of the device.

 

This chasis for the 4 tower gears is 90mm on the short side.

 

[MarkAR]

Good little project 👍

[EwanV]

22 hours ago, MarkAR said:

Good little project 👍

Thanks Mark.  It's coming along nicely.  I've decided that for the "actual" build, I'm going to reduce the teeth on the gears to 60/15 instead of 80/20.  It makes no difference to the ratios, but will make the device smaller.  I thing 60/15 is probably the smallest I want to go, as I want a good tooth/belt contact on the smaller gear to help prevent slippage.

[EwanV]

We have a new question to deal with: 5V or 12V for the stepper motor? 

 

Both have their advantages.  Firstly, the 5V

 

• It and the Arduino can be powered via a USB powerbank
• It's cheap to buy and replace
• It has an internal gearbox meaning it does over 4000 steps to a single revolution

 

It's disadvantages are:

 

• It's not very powerful
• It's not as accurate in its steps as a bigger motor

 

Now the 12V

 

• Much more powerful than the 5V
• Better build quality
• Better reliability
• Accurate steps

 

Disadvantages:

 

• Will need 12V power rather than 5, so either an Arduino Uno with its built in higher voltage port or a Nano and separate supply
• Either way, it's going to mean carrying a power brick rather than USB jobbie
• Vibrations can be an issue

 

I've decided to test with the 5V first, as I happen to have that motor spare (why buy the bigger motor if I don't have to?) but my test code has been written for both.  I've also re-designed the chasis to take both motors, whilst also switching to the 60:15 tooth gears from the 80:20 too (less plastic to print each gear).  I've also picked up a 5 guage rod (about 5.3mm diameter) that's great for testing with.

 

Finally, I've changed my mind about printing the drive belts as I can buy 10 for £5 and printing 6 or 7 was going to cost me £3.99 in materials.

[EwanV]

The 5v motor works, but I’m not happy with the level of torque the motor generates. Just in the test with one belt attached, the motor stalls occasionally despite the lack of a load on the system. The belt isn’t tight so that’s not a problem. 
 

looks like we’re switching to 12v whether we like it or not!!

[MarkAR]

I was going to say 12V would be better. 

[EwanV]

11 hours ago, MarkAR said:

I was going to say 12V would be better. 

To be honest, Mark, I didn't really think the 5V would have anywhere near the power needed and I was only using it because I had one to play with!!  I'm pleased I designed the chasis and wrote the code to work with both motors!