88Views0Replies

Author Options:

Hobby CNC router/engraver idea Answered

I recently had to fix a friends (chinese) mini router.
Long story short short: I realised that those things are only good for a small Dremel but not a real router.

After looking up what is available in the tinker areas it seemed the only way out is the expensive way of using servo drives.
Not really easy to handle either if you want some Arduino or Raspberry do the computing parts...
So what would actually be required to make a cheap but still usable tabletop router that can handle 10mm playwood or work with (at least) Aluminium?

In most if not all cases the tool will limit the movement speeds quite drastically.
If neither makes any sense to force the rotating tool through the material, nor to make several high speed passes to get the results.
Even when engraving stone or aluminium (dry) you can only do it so fast.
The hobby market is full of Arduino or Raspberry based router projects but most, if not all use standard Nema motors for the positioning.
As a lot came from the 3D printing area I guess it just made sense.
And if you want to run servo drives or similar you won't control them with some cheap Arduino.
Nema motors however suffer from the problem of limited torque at higher speeds.
And using normal gearboxes only goes so far as the higher the reduction, the lower the resulting max speed will be.
Means in the real world a Nema motor would struggle with any gearbox to get a reasonable speed to work with high loads.

I think the way out could be to use external encoders for the positioning.
Like on the early ink jet printers the slotted band attached to the printhead did.
Or if you are old enough you still remember the mechanical mouse that stopped working once the code wheels were full of dust.
The Idea came to me when servicing a geared motor for a 8-way valve.
Just a cheap Mabuchi RS555 style motor and a planetary gear box.
Together they produce a whopping 800nM of torque and literally no backlash at all on the output side.
High resolution encoder wheels with sensor are available quite cheap as well.
I did some very basic math on the accuracy side of things based on a table size of 1.2x1.2m.
Using a chain drive we are looking at less than 1/2 mm using standard chains and below 1/10 of a mm using tight, precision chains.
Going lead screw and heavy duty, high torque the accuracy is about the same as on a good 3D printer but the speed will be quite slow - I guesstimated around 3-4mm/s.
A heavy duty belt drive system seems to be perfect as we can adjust the drive pulley size to match the gear ratio and speed requirements.
Chains work fine as well here but usually mean a higher overall cost.

The benefits are really high torque with more than reasonable speed is possible.
The mechanics and electronics can be kept very simple.
No high speed processor required, a simple 8-bit Arduino Mega will do fine.
No expensive motor controllers, just a simple DC motor driver with PWM control.
Extreme holding torque when an axis is not moving thanks to the gearbox.
Code sensor can be driven by the belt or chain to get direct distance measurements.

Downsides of the approach....
The initial costs will be higher than standard Nema motor based solutions.
To my knowledge there is no Open Source project for this kind of approach available (yet).
With the heavy duty mechanics it won't be very portable, best to have it stationary mounted somewhere in the workshop.
Quick reversing of the direction will require suitable DC motor drivers that can not only reverse the polarity but also provide the option for a break resistor to allow a fast stop of the motor without backfeeding generated electricity.
Getting 3 dimensions would require a motorised table which can be a challenge if you need to support a loaf of weight and stay steady while a tool causes a lot of pushing in all sorts of directions.
Probably a sissor lift system in all four corners would work best here.

Happy tinkering! ;)
And if the idea is already out and in use just forget about it LOL

Discussions