Motorized Router Lift

[Chems]

A bit of a different one. I've been a bit unhappy with my last 2 router tables, either the table or the router. So this time around I'm doing it properly. First I wanted a cheap powerful router so got this for £60something:

Which seems to be a very nice router, coupled with a extended collet. Soft start, very powerful nice functions. The first job was to remove the springs, couldn't find a way to take them out of the top, so yanked them out through the bottom which was great fun. There aren't many pictures of this bit.

Onto the table, I used a sign making material for my top, its a aluminium composite, .3mm ally with plastic sandwiched between. Glued to an old kitchen worktop using mitre fast, made sure it was nicely flat this time as last time I used some fireback it had some bobbles in it. Chamfered the edges to ensure not sticking as things get caught on the edge of the table, a problem with previous ones:

This has since been put into a nice oak veneered router cabinet, odd for me, no photos! Next I removed the fine height adjuster from the router, and used that nice channel left to put a M12 threaded rod through. On this I drilled through a bolt to captivate that at the top within the bottom of the base. And then use a bolt at the top of the adjuster tunnel. So when the rod turns the bolt lifts the router up and down.

I could just hook it up to like an old drill or multi tool which I thought I might originally, but I decided to go a bit further. What I'm after is having 1 button to do a +1mm another to do a -1mm and another 2 to do fully up and fully down. Perhaps a 10mm up 10mm, the possibilities are endless! The problem with a standard DC motor is they just spin theres no accuracy, so I decided to use a stepper motor which can be controlled to the degree. This particular one has 42 steps at 1.7 degrees.

After a bit of soldering and learning Arduino (google it if your a tech lover) I finished doing a quick program to get it working.
Red Buttons = +1mm
Green = -1mm
White = All the way up
Blue = All the way down

http://www.youtube.com/watch?v=n2djhhEnGqU
The bit to note here is when pressing the white and blue buttons at the end the masking taped shaft goes exactly back to where it was which is the idea of the extra work involved in using the stepper motor vs normal.

And heres the basic code, I'll next implement a counter so it knows where it is so it can do the fully up fully down without limit switches.

#include <AFMotor.h>
AF_Stepper motor(48, 2);

void setup() {
  Serial.begin(9600);           
  pinMode(2, INPUT);
  pinMode(3, INPUT);
  pinMode(11, INPUT);
  pinMode(10, INPUT);
  motor.setSpeed(500);   
}

void fturn(int amount){
      motor.step(amount, FORWARD, MICROSTEP);
}

void bturn(int amount){
      motor.step(amount, BACKWARD, MICROSTEP);   
}

void loop() {
  int btn_one = digitalRead(2);
  int btn_two = digitalRead(3);
  int btn_three = digitalRead(11);
  int btn_four = digitalRead(10);


 if(btn_one == HIGH){
    fturn(1);
    Serial.print("Button 1 Pressed: ");
    Serial.println(btn_one, DEC);
 }
  
 if(btn_two == HIGH){ 
    bturn(1);
    Serial.print("Button 2 Pressed: ");   
    Serial.println(btn_two, DEC);
 }
 
  if(btn_three == HIGH){ 
    fturn(20);
    Serial.print("Button 3 Pressed: ");   
    Serial.println(btn_three, DEC);
 }

  if(btn_four == HIGH){ 
    bturn(20);
    Serial.print("Button 4 Pressed: ");   
    Serial.println(btn_four, DEC);
 }
 
}

So next up is to remove the tiny motor, and couple it with a powerful motor and controller. The problem there was working out how much torque it was taking to turn and lift the 9kg router using just a standard threaded rod. So I bought a proper measuring torque wrench after the only torque wrench I had its lowest setting was 5 feet pounds. The new one didn't even register so I'm just going to go with a 4nm stepper and hope for the best which looks like this:

So sorry for the long post with little saw dust, but hopefully interesting to others and a future alternative to:
http://www.mlcswoodworking.com/ordersta ... rlift.html

 

[monkeybiter]

Sounds pretty good, maybe you could add a graduated knob for fine feed [twentyfourths of a millimeter]

To do it right you ought to incorporate rotational feedback from the lifting screw, just counting can't account for missed steps due to uneven torque/lifting force.

That's a good price for a 1/2" router, does it seem ok ?

 

[Chems]

Yeah I think your right re the rotational, even just working with the small motor its becoming obvious. I can hook it up to a potentiometer easily enough to do that, I even have one here!

I think at the moment, the 1.7 degrees got by pressing the +1/-1 buttons probably would equate on the acme screw to even less than a twenty fourth of a millimetre, but I've no real way of know until I hook it up to the actual table.

I've been using the router for about a month now, and it is good. The only major feature I miss from the Triton is the spindle that locks once you wind through the base. It has a nice soft start and good extraction. Can't go wrong for what it cost. Especially for a dedicated table router.

 

[monkeybiter]

If you've used an M12 threaded bar that will be a thread pitch of 1.75mm. I was going to say what that was per step, but I've just re-read your post and 42 steps at 1.7 degrees per step doesn't add up to 360 degrees. Is it perhaps [email protected]' ?

And do you know what time it is?

 

[Chems]

I don't know what I'm going on about, the final motor will be a 1.8 42 stepper! This smaller one I'm using to prototype is 7.5 at 48 steps.

My major concern is that the big motor won't be man enough to turn the rod. But I've done everything I can to measure it to no avail, its to low to register on any measuring device I have.

I know what time it is, its the Geeking HOUR! Can't do any serious coding before midnight.

 

[Paulg]

You may want to look at a quad encoder for feedback on the rotation / movement. Hook the encoder directly up to one end of the threaded bar and depending on the encoder itself, you can get extremely accurate positioning. Personally, I have access to encoders with 1024 counts per revolution, but they are expensive. I have one of these smaller ones to play with (but haven't done anything with it at the moment):
http://www.spinvent.co.uk/product info pages/Incremental_Encoder_PIP.html

You could also use one of these as input for fine movement as well as the buttons.

On the code, there are a couple of things I would suggest you look at. Firstly, you need to debounce the inputs from the buttons. If you don't, you will get more movement than you expect. Second, you may want to add in a routine to make sure that the button is released before starting your loop again. At the moment, if you press and hold one of your 1mm buttons, it will just keep going but I think what you want to do is to just move 1mm for each button press.

Hope I am not teaching my granny to suck eggs.

 

[Chems]

Paul, not at all. I was working on that very problem last night as when you press the 1mm buttons its not able to count accurately the amount of steps taken. I added in a 1 mili second delay but its still not counting exactly right if you press and hold. I think I will do as you say an write a function to catch the button up before re-sending a 1mm step.

It seems a little bit double engineering to need a rotational counter with a stepper, I thought the whole point of them was that they were deadly accurate! Do CNC machines use such things?

 

[knappers]

What you need is one of these, mate.

Si.

 

[monkeybiter]

Steppers are accurate, but only if you can guarantee that every step generated results in the required movement. If you've got surplus power [so an uneven load won't defeat the motor] and you don't generate steps too quickly [so that the loaded motor can 'keep up'] then you could get away with no feedback. The main need for feedback with a stepper is to provide 'proof' that each generated step results in the required movement, with that information you can then calculate the actual position.

 

[Paulg]

Paul, not at all. I was working on that very problem last night as when you press the 1mm buttons its not able to count accurately the amount of steps taken. I added in a 1 mili second delay but its still not counting exactly right if you press and hold. I think I will do as you say an write a function to catch the button up before re-sending a 1mm step.

Generally with debouncing inputs, you need between 20 and 50 millisecs of constant signal before deciding the signal is solid and then acting on it. Under 50mS, generally you would not be able to notice that there is a delay. Personally I would debounce both the positive (pressing) and negative (releasing) of the button.

It seems a little bit double engineering to need a rotational counter with a stepper, I thought the whole point of them was that they were deadly accurate! Do CNC machines use such things?

In general, stepper motors are open loop controlled. This means that there is no feedback. When they step, they step accurately, but there is no guarantee that they will step. This could be due to engineering reasons (i.e. the motor cannot handle the load) or it is being fed the move signal too quickly. You seem to be thinking more of a servo motor that has the integrated error correction. CNC machines use servos.

Any other info you need, give me a shout.

PG

 

[Chems]

CNC's defiantly use steppers, or at least some do, as I've seen tons of stuff with Steppers and CNC machines whilst I've been researching.

I'll code that switch method in biter and let you know how I get on, I think its defiantly the key.


@Knappers - It to much like hard work for me!

 

[Markvk]

Hi.

First I wanted a cheap powerful router so got this for £60something:

I know this thread is very old, but im just planing a router table and im wondering how it all progressed. How did the cheap router work out? and what make /model did you use? im considering the use of an old car window regulator motor for my project as i cant be bothered with code ect at this stage.

Any advice would be much appreciated.

Mark

 

[monkeybiter]

It was this one http://www.silverlinetools.com/en-GB/Products/Power Tools/Routers/124799

HTH

 

[custard]

Motorised lifts on some spindle moulders are really useful, not because of the lifting function itself but because you can then come back to a tooling set up weeks or months later and re-set it all to within 0.1mm using the digital read out. In a production environment that can slice hours off a spindle moulder set up time.

Unfortunately that convenience would be unlikely to translate to a router table, because installing the router cutter will not be a super accurate and repeatable process where you're working from known reference positions. Consequently I'd think twice before committing much time or money towards developing a motorised lift on a router table.

 

[Markvk]

Motorised lifts on some spindle moulders are really useful, not because of the lifting function itself but because you can then come back to a tooling set up weeks or months later and re-set it all to within 0.1mm using the digital read out. In a production environment that can slice hours off a spindle moulder set up time.

Unfortunately that convenience would be unlikely to translate to a router table, because installing the router cutter will not be a super accurate and repeatable process where you're working from known reference positions. Consequently I'd think twice before committing much time or money towards developing a motorised lift on a router table.

i dont think it will be a problem getting accurate repeatable depth. There are 2 way that i can think of doing it, i could use a dro and zero it each time a new cutter is installed by placing a straight edge on top of the table and butting the cutter up to it and then zero the dro and lift until readout gives the depth

or

set a relay in the control circuit that stops the motor lift when energized. this can be done with a trend depth setter with a fly lead attached that grounds the circuit when the cutter touches it

anyway it will keep me from under the wifes feet :lol:

 

[Markvk]

It was this one http://www.silverlinetools.com/en-GB/Products/Power Tools/Routers/124799

HTH

Thanks Mike i did look at that one but couldnt see how it could be readily adapted like my dewalt can, maybe chem will post a pic?

 

[Markvk]

It was this one http://www.silverlinetools.com/en-GB/Products/Power Tools/Routers/124799

HTH

Thanks Mike i did look at that one but couldnt see how it could be readily adapted like my dewalt can, maybe chem will post a pic?

If i substitute the word nut for bolt on his original description, it makes sense how he converted it!