Arduino Rotary Table

[froudeg490]

I have finished the electrical portion of this tool and I'm now ready to buy a 4" rotary table.

Such as:

https://www.amazon.com/Inches-Regular-Rotary-Milling-Machines/dp/B099KMBKTH/ref=sr_1_10?crid=3M7PEN1GQT71X&keywords=rotary+table&qid=1682803738&sprefix=rotary%2520table%2Caps%2C150&sr=8-10

Anyone purchased the same one (or very similar)?

I'm trying to decide if I should get this one. Perhaps there is a version that is better?

 

[Scott_M]

I am not sure which code you are running on the arduino, but some of the various code that is out there has rounding errors. This table that you linked to has an odd 36:1 ratio. Most rotary tables have 90:1 or 40:1. Before you buy I would check your code and see how it will handle that ratio.
It has been years since I messed with rotary table code but I do remember the above.
Hopefully someone will be along and ask for your code and check it. If you are unable to make that call.
If you already have your stepper, you could install a pointer and zero mark and check it with various divisions and see if it will repeat and return to zero.
I would make sure of this before I bought a table.

Just me 2 cents worth ( not inflation corrected )

Scott

 

[froudeg490]

Thanks. That's good to know.

I do have the system working with an old NEMA 17 stepper I had lying around. It's 200 steps per 360 degrees.

I'll put a pointer on the stepper shaft and use my old degree wheel to check it out.

I was planning on making the degree wheel the face of a clock for my workshop, so the timing is perfect (no pun intended).....

 

[ignator]

I have finished the electrical portion of this tool and I'm now ready to buy a 4" rotary table.

Such as:

https://www.amazon.com/Inches-Regular-Rotary-Milling-Machines/dp/B099KMBKTH/ref=sr_1_10?crid=3M7PEN1GQT71X&keywords=rotary+table&qid=1682803738&sprefix=rotary%2520table%2Caps%2C150&sr=8-10

Anyone purchased the same one (or very similar)?

I'm trying to decide if I should get this one. Perhaps there is a version that is better?

I purchased this same rotary table about 15 years ago from Harbor Freight. It was made in India. It is very poor quality. I recall paying $90 for mine. The crank drive to the worm gear is very stiff, and there was too much backlash.
Maybe the quality has improved. But mine sits on the shelf unused.
The link from Amazon, no one has left a review for the 4star or 1star ratings that two people did give.

 

[weir-smith]

It doesn't matter what the ratio is, it can be adjusted on line 34 of the sketch and I provided a list variables for each ratio.
I have provided the sketch to well over one hundred people so far and I haven't had any adverse feedback. I modified the original sketch to remove the rounding errors and other problems/issues that I found when testing.
I have used my setup to machine a lot of gears for model IC engines, so if you experence any problems, just send me an email.

Bruce
Western Australia

 

[Toymaker]

Since you do not already own a manual rotary table, you don't need to buy one just so you can modify it by bolting a stepper motor onto it. All you need is a ridged spindle with a round table attached to one end and a method of coupling the spindle to a stepper motor. The stepper and software are doing all the calculations and moving the table the correct distances.

Some years back I built the rotary "device" you see in this video: Rotary Device. It consists of a stepper motor turning a spindle coupled together using a timing belt and two pulleys. After the splines were finished, I used my rotary device to hold the disk shown in the video while the blades shown in the below pic were machined.

Timing belts and steppers are used in virtually all printers today,...the system is extremely accurate,...you need only to ensure your design is also extremely ridged.

 

[froudeg490]

It doesn't matter what the ratio is, it can be adjusted on line 34 of the sketch and I provided a list variables for each ratio.
I have provided the sketch to well over one hundred people so far and I haven't had any adverse feedback. I modified the original sketch to remove the rounding errors and other problems/issues that I found when testing.
I have used my setup to machine a lot of gears for model IC engines, so if you experence any problems, just send me an email.

Bruce
Western Australia

Bruce,

The electronics worked perfectly, first try. Using default settings for gear ratio. Thank you.

I wanted to retain the the handwheel so I 3D printed a bracket and used a 2GT belt drive with a 3:1 ratio.

When I update the code to include the added 3:1 ratio. The code stopped working.....

No worries. The number of steps for 360 degrees (43,200) now exceeds the range of the Int. data type (32,767).

I just changed the driver to single step so now everything works as it should. 1 Deg. = 60 Steps.

Phil

Phoenix Arizona (it's only 48 degrees today).

 

[Toymaker]

Bruce,

The electronics worked perfectly, first try. Using default settings for gear ratio. Thank you.

I wanted to retain the the handwheel so I 3D printed a bracket and used a 2GT belt drive with a 3:1 ratio.

When I update the code to include the added 3:1 ratio. The code stopped working.....

No worries. The number of steps for 360 degrees (43,200) now exceeds the range of the Int. data type (32,767).

I just changed the driver to single step so now everything works as it should. 1 Deg. = 60 Steps.

Phil

Phoenix Arizona (it's only 48 degrees today).

Why not change the data type to long which has a range of 2,147,483,647,... the syntax is identical to Int and would look like this:

long var = val; // where var is the step name and val is the integer value you assign to var.

 

[froudeg490]

Why not change the data type to long which has a range of 2,147,483,647,... the syntax is identical to Int and would look like this:

long var = val; // where var is the step name and val is the integer value you assign to var.

I considered changing the data type to long, but because of the 3:1 belt drive reduction, the table moved very slowly. It was an easy way to speed things up. I always install a pin header and use shorting jumpers for enable, Mode0, 1 and 2. So I just pulled the jumper off M0.

Thanks for the suggestion. I will probably add the STL files and list of hardware to Thingiverse, in case someone wants to copy/improve this method of driving the table with a belt....

 

[froudeg490]

I added the STL Files to:

Rotary Table Stepper Motor Adapter by froudeg490

This is a method to add a NEMA 17 Stepper motor to a Rotary Table: https://www.amazon.com/gp/product/B01MY9H1EO/ref=ppx_od_dt_b_asin_title_s00?ie=UTF8&psc=1 It utilizes a 2GT-200 belt and two timing belt pulleys, 60T and 20T. The 60T pulley was printed using PLA and is a push fit onto the...

www.thingiverse.com

 

[weir-smith]

Phil

If I remember correctly your ratio was 4:1 and the divisor was 4.4444. Now you have added a further 3:1 eg 7:1 total and the new divisor is 7.7778.
All you needed to do was edit line 34 with the new value. The stepper driver should be left with a division of 2. The output from the Arduino is 400 steps per revolution however, the stepper requires 200 steps per revolution. At 7:1, it requires 1400 steps at the stepper motor or 2800 steps from the Arduino output. You should not have needed to use long Integers as you are well below the limit.

In some instances, people have used the system to make multiple full revolutions eg continuous rotation. In this case, you will eventually suffer overflow and using "long int" for some of the variables would be required.

Bruce W-S

 

[froudeg490]

Bruce,

The table I purchased has a ratio of 36:1.

I calculated that with the driver set at 1/2 step, the number of steps required for 360 degrees would be 36*3*400=43,200

As a quick test I changed the driver to full step so that 360 degrees would be 36*3*200=21,600 steps.

I didn't consider multiple rotations. In that case "long int" would definitely be a required.

Phil S

 

[mcostello]

You Guys blew past My pay grade at supersonic speed.

 

[gjrepesh]

I used a Sherline 4" rotary table. They are 72:1. Worked out quite well.