CNC technic

Plans Otocoup

Plans 3Dcoup

Credits - Links



CNC router 'OTOCOUP'
- How did i arrive there ? -

 In English
 Circuit L6208

In fact, this is a long story....

I was unemployed and beginning to gave up of computer programming, so i came back to more physical homework.

I already got the idea that a big sawing machine as it does exist in do-it-yourself stores might be very useful.

Also, i'm an experienced do-it-yourselfer though i am a clumsy guy, and i have always been afraid of circular saws, as they can do ravaging injuries.

So, i've done drawings of a one-axis (vertical) machine. I have made four versions, the final one being based on medium wood panels.

My aim was also (and that is nearly as important as the cutting part) to have some sort of vertical workbench, for room reasons and also ease of use.

Drawings and BOM's were finished, i've bought the saw, best price supplier for medium was found, but i get a new job(...). So, this project was set on hold.

I am not really a modeller, but i'm interested in that activity, and i have designed and built a hovercraft model based upon full size concepts (this is also a long story, but not the same... see here).

While browsing on the web, i've been on various modeller's site, and along that travel, i've gone through sites about machining for modellers. I got the idea of building such a machine. And soon, got also the idea to mix that project with the saw machine project, as it might be not really much more work to do a bigger machine. That's true for building, but it is the sake of evidence that the longest time is to make the design.

I've discovered on Aeromaniacs site (french) Guillaume Coquery's machine that i have closely studied. I've found it to be the best in terms of simplicity, stiffness and ease of building. Though considering the size of my project, not usable as is, particularly because aluminium rulers lack some stiffness. So i came back again on the computer, and have done new drawings on this concept, sending nearly all previous drawings in the bin.

For motor control, i have found the best explanations on telescope follow-up systems.
A PWM (Pulse Width Modulation) was done only by software on 486 PC (under DOS, evidently, who says Windows is slow ?). Really a remarkable work (http://www.bbastrodesigns.com/cot/cot.html)

The biggest problem is that i need powerful motors, considering moving weight (10 kg for the beam, 7 kg for the carriage fully equipped). It arise quickly while reading catalogs and Internet, that the most powerful motors easily affordable (in France) were bipolar motors (2.6 V, 2A). Though no control board of good quality for bipolar motors was existing on french market, complete cards exists only for unipolar motors.

I've searched what we can found for stepper control. My conclusion was the couple L297/L298 was the simplest and most affordables circuits. French site of Laurent Fouga was proposing printed boards based upon these circuits.
Small tracks and the imperative need for such a machine to have a manual command, have driven me to study myself new boards. It was easy to find informations about L297/L298.

I'm a beginner in electronics, but i have good technical bases, and have some knowledges about logic.

So, i've began to design circuits directly (without schematics, but based upon application notes, which are nearly complete) on Autocad. I've tried Proteus in DOS version, but found it's ergonomy not that good.

So, next, for ease of use, i've used PCBExpress, but you must get a screen bitmap copy and rework it to have small holes needed for manual drill. I've designed five boards like that, which is at the end, a very heavy manner : One base board, one board for unipolar motors, one board for bipolar motors, and two boards for manual pad.

I've ordered my printed boards, which never arrived (after 3 months). So while waiting them i've discover Eagle PCB software, and began to design new boards. I've also discover a new circuit L6208.

All these cards now works, after some tuning, mainly due to the poor quality of my soldering. I'm happy to have the control card based on L6208 working properly, as this is more original work than my others cards based upon L297/L298.

Gear reduction i have installed drive to a static load of 45 kg, with two winding supplied on the stepper. Practically we can get 10/15 kg permanently. This is sufficient, but the machine don't travel very fast. I have to use a software which manage accelerations, and can sustain sufficient steps per second.

The steppers i've used (bipolar) still maintain half of the maximum torque when used at 5000 steps per second (if driven by a chopper). Such rythm is obtained reliably on Windows by very few software. This is related to Windows latency, which is a true horror. This problem can be solved making a board with a timer which will run hardware interruptions on parallel port. This is done yet on foam cutter software of Gilles Muller, but not on router software. Also Artsoft have solved this latency problem with their software.

With DOS or Real Time Linux, this is not a problem.

I've chosen the present steppers because they are very easy to afford and they have a good torque (6.2 kg.cm). This works well with an acceptable speed with power supply 13.8V. I can have fast moving much faster by increasing the voltage of the power supply, but the global gain will not be that high.

I've in stock unipolar 12V Nema34 stepper of larger torque, which are heavier and bigger, but they need a supply (with a chopper) of around 35 V. So now, i've preffered the 'small' steppers..