DIY hobby small plastics mill/router

[RomanLini]

New PSU

A couple of weeks back I made a new PSU. One of the goals was to get automatic spindle on/off switching and the other goal was to raise the PSU voltage a bit to make my 200 inch/min rapids fully reliable, as before with my 25v PSU there was a very small chance of missing steps on a 200 inch.in rapid.

The new psu is 45v with no load (that doesn't really happen), 42v when the steppers are on "low current", and 40.5v when the stepper motors are running at full power. It's rated for 7 amps.

I have a 5 amp analog ammeter on it, and can see the current my stepper motors use. It's about 1.7 amps. Occasionally peaking to about 2.2 amps under hard accel from zero to 200 in/min. A 2.5 amp supply would have been plenty. My 3 stepper motor drivers are set to 2.6 amps each.


Resonance problem.

I was a little annoyed that the machine has some issues with a 40v supply. It makes significantly more resonance noise, vibration etc than it did at 25v which was silky smooth. It also has worse tooling marks (worse cut finish) on harder materials and a higher tendency to get melt when cutting HDPE.

I'm definitely going to lower the PSU voltage again, it's not worth living with worse machining performance just for a fraction higher rapids speed.

After installing the new PSU and speed testing, I coded new acceleration ramps into my PIC motion controller and improved accel through the resonance band quite a bit. I also re-aligned my Y axis bearing rails and leadscrew and now there is no sign of missing steps so it is unlikely that the machine needs 40v anyway.

I'll drop the PSU to 35v next time I open it up and see if that is a happy medium between the old silky smooth 25v performance and 35v confidence in full rapids speed.

Anyway I usually have my rapids set to about 120 in/min, which is plenty fast when it is moving between parts that are only an inch in size...


Auto spindle on/off switching

The new PSU has a solid state relay built in, which is connected to the white 240v mains socket on the front. The PC software tells the PIC in the machine to turn the spindle on/of and the PIC tells the PSU which controls the power to the router.

The switch for the spindle router power has 3 positions;
ON
OFF
AUTO (under software control)

Yeah I know it's a standard type feature on most people's systems but it's so nice to have! I can start a job and go and watch Mythbusters, cook food, etc and the machine just does its business, then turns off the router and goes back into low power mode when it's finished.

[Arquibaldo]

Dear RomanLini:

Congratulations on your beautiful and original machine... I'm particularly interested in the nuts:

"The leadnuts were cast from a special slippery plastic that I work with and know its shrinkage modulus so they shrink during casting to the correct tightness and have no backlash."

Would you be willing to disclose a little more: type of material and method used for the casting?

Thanks a lot

Nelson

[RomanLini]

Sorry I can't disclose that. I work with some specialised plastic compounds in my work as a designer and I'm not at liberty to give details on that.

If you search the forum there are a number of people here using plastic leadnuts, and some threads on cutting or forming leadnuts.

(edit) Sorry I just realised that sounded unfriendly. If you check the Igus range, besides polymer bearings they also have polymer leadnuts made from a special plastic and come in most standard ACME thread sizes. They advertise their polymer leadnuts as much better performance than Delrin or other "bearing" plastics.

[Arquibaldo]

I understand... Thanks a lot for your attention.

Nelson

[RomanLini]

Update on PSU and resonance

I modified the PSU today, to change the main voltage from 40v down to 33v.

The PSU has two transformers to give the main voltage; a 24v unit and a 20v multitapped unit so it was pretty easy to change the tap on the mutitap transformer and set the PSU to 33v (that is a 33v at full load, with motors at full current).

This seems to have made a good improvement! The machine is quieter now, and when moving at some slow test speeds I can feel much less vibration on the moving table or the moving part of the X axis (when touching them with a fingertip). I haven't done any tests of the cutting finish yet but will do so within the next few days.

All the high speed rapids still tested fine even at the reduced voltage. It even performs rapids reliably with the motors at 45% current in low power mode.

[Jesse B]

Would you say a 36v supply would perform significantly better than a 24v, or are they fairly similar?

[RomanLini]

Hi Jesse, it depends a lot of the motor inductance. My machine does some hard accelerations when the speed is cranked up, and the 33v outperforms the 25v PSU for sure. But then, my motors are 6v 2A motors which have a fairly high inductance.

It's a trade off between smoothness and high speed acceleration. With a higher PSU voltage the motors move much quicker between steps, but at low speeds that causes a lot of ringing on each step and can wreck your finish.

If you use a low PSU voltage your motor will run so much smoother at lowish speeds, but may skip steps on a high speed acceleration.

One day I'll probably design "the world's best stepper driver" that automatically changes the PSU voltage (or source impedance) to suit the speed the motor is operating at, but that won't be for a while as I have a lot of projects on at the moment.

Smoothness update
I've been cutting a few items now with the PSU at 33v (reduced from 40v) and the result is good, it's stopped the melting and finish is better (and of course audible vibration is reduced).

Currently the only machine resonance (I can notice) is the gantry, and it is only a small resonance. When I get some spare time I will either fill the gantry rear 100x50mm box with epoxy granite, or maybe just glue a length of timber inside it. It won't take much to dampen, at the moment I'm just avoiding the particular speed where it rings.

[RomanLini]

Well it's been about 10 weeks since I updated this build log.

Worn out the first router...

Yep. The new router I bought in Feb 2010 has done a heap of 10 and 20 minute jobs, but would have been no more than 10 (maybe 15) hours actual running time over the 5 months, and always at minimal speeds and power cutting plastics with small 1/8" cutting bits. Very light loads.

It was about 15 minutes in to a 20 min light cutting job with the speed control right down on 3 and the router started making an extremely loud squealing sound.

I popped the dust cover cap off and checked the brushes thinking it might have worn a brush down to the metal, but the brushes looked like new, full length (which is to be expected at 15 hours of low speed use).

Further inspection showed the commutator wobbling like crazy, it looks to be on a hollow spoked plastic bobbin and must have got JUST warm enough on the shaft for the bobbin to deform and the commutator to tilt. I've always checked the router temperature too and it's never got very hot at all. It looks like a nasty design defect and failing after a few hours of light use backs that up.


Router replaced.

The router was under a 3 year warranty and they replaced it with no problems. Interestingly enough, the first router model that was fresh in Feb2010 had been totally replaced in the store by a better unit only 3 months later. I assume the manufacturer found out the first model was defective.

Anyway the new one looks similar but is 850W instead of 650W, and is just a fraction (maybe 10%) larger and heavier. One problem was that the old unit mounted in a 42mm grip and when I got the new unit home discovered it mounted in a 44mm grip! This was annoying having to disassemble and re-machine my router mount to take the slightly larger unit. Also I had to re-align my machine in X and Y axis a fractional amount.

The new router is quieter, has a heavier rotor, much more low speed power and blows about 3 times the amount of air through it. All good things considering the failure mode of the first router...

Let's see how many hours this one lasts! (Note to self; Code up an hour-meter into my CNC software to log actual hours of router usage.)


Considering moving my X axis home sensor.

One thing I have noticed on some cold winter days lately is the thermal expansion of my machine is fairly large. The X axis electronic home sensor is quite sensitive and picks up a difference of as much as 0.11 mm thermal expansion between the machine when first switched on (at maybe 15'C) and later warm around body temp (maybe 35'C).

I don't have this problem with the Y axis, the sensor there is mounted at the end of the leadscrew close to the motor and doesn't record the thermal expansion to any significant amount.

At this point I'm considering moving the X axis sensor to the other end of the leadscrew. This is the opposite end to the "home" rest position but even though it will take a couple more seconds to test the X axis it will remove the thermal expansion error from the X axis reading.


Fixing the thermal expansion itself?

I haven't seriously considered this but it would not be too hard to keep the aluminium frame of the machine at a constant temp (say 40'C) to keep it's frame at a constant dimension, and then run a home switch at each end of the leadscrew and apply a software compensation.

This would be fairly cheap to retrofit and not that difficult to do, although I would need to borrow a decent DRO to do the initial calibration.

Currently I'm too busy with commercial work to do hobby machine modification but it would be an interesting mod when/if I get some spare time to put it together.

[RomanLini]

CNC Software Update;

I added the feature that logs hours/mins/secs of router usage in a text log file.

The new router has logged 2 hours 46 mins so far (over 2 weeks) but the machine has been getting a little more use than normal.

One thing surprised me... Most of my CNC cutting jobs are much shorter than I realised, they only take a couple of minutes of actual router time. I had perceived them as 10-20 mins etc because of the time to load the CAD files, set up the cutting depths, warming up and homing the machine etc.

[sarahlacy]

great info RomanLini , i don't know what to say .
i will try this in the next few days and post the results here

[eaglezsoar]

RomanLini, you mentioned James has kits available.
Could you give me the URL for the site.
Fantastic job on your build!

[RomanLini]

Thanks for the nice words. I'm very happy with the machine, it's not pretty but every part of it works very well.

The link to the SLAmStepper kits James has for sale is;

http://www.piclist.com/techref/io/st...m/SLAm_bld.htm