I got a ‘kit’ of goodies from buildyourcnc which got me going.  The kit included the electronics, and the code required to cut out the wooden parts (called gcode).  It also came with all the nuts/bolts/chain/sprockets etc needed to assemble.  It took about 5 hours of milling to get the parts cut out of 3/4 plywood, but it was entrancing to watch!  Below is a pic of all the parts drying after having poly applied.  Probably wasn’t necessary, but as it will live in the garage I didn’t want the humidity cycles to affect it.

From the bottom up here is how those parts come together:

• 2×4 frame on locking roller wheels from Lowes.  This was pretty basic, but I did do lap joints in all the corners for stability.
• On top of that is a 2′ x 6′ torsion box made out of MDF.  There are lots of good guides on building torsion boxes, and it was suprisingly easy with just my portable tablesaw.  The advantage of the torsion box is it should stay dead flat over time.  I also put 2 coats of poly on this as MDF loves to suck up water!
• On top of the torsion goes the actual bits that make up the CNC.  The X-axis rails go on the torsion, which the gantry rides on.  Y and Z axis are rails on the gantry.  Its a pretty amazing design, v-groove bearings ride on basic aluminum L rails.

 In the gantry are three stepper motors that are controlled by the electronics below.  Pretty basic wiring for the most part.  PC parallel port goes to the breakout board bottom-right.  This controls the 3 axis via the black boxes top right.  Power supply for the 36v needed on the top left.  Bottom left are two electrical boxes for outlets.  The left most one will have a circuit to turn the router on/off.. but that is a post for later, as I intend to mill the circuit board with the cnc.

There are a lot of steps not documented here, but you can see very detailed build videos on buildyourcnc.com if interested.  I just wanted to share my beginnings on this project, and will post more detail as I go off the beaten path and start to create my own parts.  Below you will see the first milling operations I completed – the manufacture of the sacrificial top for the machine.  This top is intended to be replaceable as it will get chewed up over time.  The T-Slots allow mounting of wood for milling out of parts.

The bubbles coming off the plates was immense!  You couldn’t even see in as the water was so frothy with them.  That lasted for about 10 seconds until my 20amp shop breaker tripped.  The water, electrical leads, and alligator clips were pretty hot too.  So, with that I’d call the trial run a success – but I really hope I can bring down the current draw using distilled water and (limited) baking soda or I will have to resort to other measures to limit the current…

The picture below is from when it was all over, I wasn’t brave enough to stand close on the first run.

For the electrolysis I used stainless steel outlet covers on stainless bolts. I kept them seperated by nylon spacers. Two sizes if spacers kept them apart as well as protect the bolts from touching the plates they went through.

To mount them I purchased some sweet stainless bolts, nuts, washers as well as rubber washers.  I didn’t want any rust to form and I generally have a tendency to do things ‘right’ – even if it takes longer and costs more.  You’ll soon see how I fell off that straight and narrow path…  So my plan was run a bolt through the bracket then go fender washer | rubber washer | grill | rubber washer | fender washer| bolt.  The problem I quickly ran into was it is IMPOSSIBLE to reach around behind the lower grill.  The intercooler is right up in there.  I even had a few failed attempts at installing them with needle nose pliers using a ship-in-the-bottle approach.  I read up on removing the lower grill, but that seemed too involved for this project – I’ll save that for the rally light install.

So abandoning that approach, I screwed in the left one as the holes are not cut out way on the left.  The right one is mounted a bit more ‘GHETTO’.

Yup, that’s right.  Zip ties.  How embarrassing.  They actually hold pretty well all things said.

At this point I just need to order up a nice black powder-coat license frame to finish it up.  Otherwise good to go, until those rally lights go in.

First, I had to get the mini naked.  The factory stripes and invisible bra had to go.

The heat gun on low heat made quick work of stripes.  I was actually pretty amazed at how readily they pulled off, with how ‘stuck’ they seem during normal car-washing etc.  They didn’t leave too much film on the car either.  The invisible bra was a bit tougher.. it pulled off OK with the heat gun, but left a terrible mess of adhesive behind.  After pulling everything off I went and hand washed it prior to using goo-gone to remove the left behind adhesive.  I didn’t want any grit to get rubbed in when i was scrubbing away at the adhesive.

Doesn’t look too bad naked.  I had to pull off the mini logo which took a bit of finessing as well.  I didn’t take pictures of the front removal, but have some from the rear.

I read somewhere the dental floss works pretty well cutting the adhesive, and it does.  I went through quite a bit as it breaks a lot.  There are two posts behind the round part, and I sawed all the way around with floss.  It was still pretty stuck at that point, so I worked two zip ties in and used them as handles of sorts to pull it of.  It was troubling becuase you have to pull HARD.

I’ll admit I was excited to get started, but like more things you are better off getting all the prep work done.  In this case I wanted to show you some problems i had with the stripes I purchased off ebay.  They came in three pieces – Bonnet, Roof, and Boot.  In hindsight i should have verified they were all exactly the same, because as you can see in the photos they weren’t.  In addition to checking out your supplies make sure there are NO bumps in the finish.  I noticed a sap spot on the hood and didn’t take the time to clean it off properly.  Trust me, the vinyl will only exacerbate any bumps in your finish.

I started on the boot as I figured that would be the least visible place to start.  Don’t do that.  The compound curves are probably the most challenging part on the whole car.  I would start on the roof, then hood, then boot.  I didn’t take pictures of my approach to centering the stripes.  But basically I would put masking tape on the trim, window, plastic or whatnot bracketing both ends of the area I was working on.  I would then measure in from the sides of the hood, trunk, roof to find the center.  I would mark that on the masking tape then measure out 1/2 the width of my stripe each way and mark the edges.  I would then align the stripe to these marks.  The centering was surprisingly easy.

Applying the vinyl was easy too, you can read any number of posts to describe the soapy water mix etc.  What took me an hour and a half of frustration to learn was the proper squeegee technique.  I used an old medical insurance card (might as well get some use!) but had a hell of a time getting the vinyl to actually stick.  I would squeegee out the water and it would just go ‘loose’ again.  I finally figured out to wrap the card in paper towels.  This would suck up the water that squeezed out from under the vinyl and make it stick.  Without the paper towel the water would just zip back under and make you want to pull your hair out.

Once you have it centered and squegeed you have to trim it.  I couldn’t find ANY information on how to do this anywhere.  I saw some people would show cutting full body wraps with an x-acto knife on the door lines, but no tips on how to cut to size.  So this could be the wrong way, but it’s how I did it.  Most of the ends of the stripe were wrap around (ie around bottom of bumper or bonnet) or on the sections of glass roof panels.  These were easy, as i just cut them long with scissors for the wrap or cut on the glass with the x-acto.  What was stressful was trimming the stripe on the paint with the x-acto.  If you cut too hard its right into your finish!  This felt somewhat like cooking bacon with no pants on, very dangerous.

Here is an example of doing that.  I made a template of paper, transferred it to cardboard from a cereal box and used it to trip the Stripes.  I think it worked pretty well, and hopefully i didn’t damage the finish.  Have a better way?  Please let me know!

Overall the project worked out pretty good, although it took me a LOT longer than I had anticipated.  I probably spent 5 hours on the application, not counting the removal which I had done a few weeks prior.  I left the mini logos off as i thought it looked cleaner.  The stripes cover up the holes the badges had left behind pretty well.  I also moved my license plate as I couldn’t bring myself to put it back in its original spot.  I’ll make another post showing how I went about that.

int relayPin1 = 6;             // choose the pin for the RELAY
int relayPin2 = 12;             // choose the pin for the RELAY
int relayPin3 = 11;             // choose the pin for the RELAY
int relayPin4 = 10;             // choose the pin for the RELAY
int relayPin5 = 9;             // choose the pin for the RELAY
int relayPin6 = 8;             // choose the pin for the RELAY
int inputPin = 7;               // choose the input pin for receiver
int val = 0;
int val2 = 0;                   // variable for reading the pin status
int onToggle = 0;               // variable so delays don’t occur every cycle
int offToggle = 0;              // variable so delays don’t occur every cycle
int buttonState;

void setup() {
pinMode(relayPin1, OUTPUT);   // declare relay as output
pinMode(relayPin2, OUTPUT);   // declare relay as output
pinMode(relayPin3, OUTPUT);   // declare relay as output
pinMode(relayPin4, OUTPUT);   // declare relay as output
pinMode(relayPin5, OUTPUT);   // declare relay as output
pinMode(relayPin6, OUTPUT);   // declare relay as output
pinMode(inputPin, INPUT);     // declare receiver as input
buttonState = digitalRead(inputPin);   // read the initial state
}

void loop(){
val = digitalRead(inputPin);  // read input value
delay(10);
val2 = digitalRead(inputPin);  // read input value
if (val == val2) {
if (val !=buttonState) {
if (val == HIGH) {            // check if the input is HIGH
if (onToggle == 0) {
onToggle = 1;
delay(200);
digitalWrite(relayPin1, HIGH);  // turn relay ON
delay(200);
digitalWrite(relayPin2, HIGH);  // turn relay ON
delay(200);
digitalWrite(relayPin3, HIGH);  // turn relay ON
delay(200);
digitalWrite(relayPin4, HIGH);  // turn relay ON
delay(200);
digitalWrite(relayPin5, HIGH);  // turn relay ON
delay(200);
digitalWrite(relayPin6, HIGH);  // turn relay ON
}
offToggle = 0;
}
else {
if (offToggle == 0) {
offToggle = 1;
delay(200);
digitalWrite(relayPin6, LOW);  // turn relay OFF
delay(200);
digitalWrite(relayPin5, LOW);  // turn relay OFF
delay(200);
digitalWrite(relayPin4, LOW);  // turn relay OFF
delay(200);
digitalWrite(relayPin3, LOW);  // turn relay OFF
delay(200);
digitalWrite(relayPin2, LOW);  // turn relay OFF
delay(200);
digitalWrite(relayPin1, LOW);  // turn relay OFF
}
onToggle = 0;
}
}
buttonState=val;
}

}

My approach was to take the 12v ‘trigger’ output of the pre/pro and trigger relays in a power strip turning the amplifiers on.  Where things got tricky is I didn’t want all the amps to turn on at once.. I was looking for a more satisfying *click* *click* *click* as they turned on/off sequentially.

So, I picked up a power strip at a Big Box retailer and removed two of the outlets making room for a power supply, arduino mini, and a circuit to operate the relays.  I used a piece of copper I JB-welded in place to cover up the outlet holes.  The copper came from a roll of roofing flashing I got at the same big box many years ago.. super cheap and handy to have around!

Covering the removed outlet holes

My original intent was to use a 555 chip for each relay/outlet, allowing each one to turn on and off one second apart.  I muddled through the design for that, ordered the parts and perf boarded up the first two modules.  At this point, although they were functional I realized a few things:

• Point-to-point wiring takes some advanced planning
• I stink at ‘bridging’ pads on purpose as required on the perf board I used
• Repeating this module 4 more times was not my idea of a good time
• I would not have enough room for all 6!

So I bailed on that design and ordered my first Arduino board to continue the project.  Picture documenting the 555 dead end:

555 on perf board

For anyone who hasn’t used an Arduino yet you are missing out on some good fun.  Having a bread board circuit with the proper delay being generated by the Arduino in under an hour was impressive.  I still had to do some perf board work to get the transistors in place to drive the relays and provide a home for the opto-isolator which keeps the pre/pro electronics isolated from everything else.  For those observant folks looking for the diodes they are attached directly to each relay, but in hindsight it would have been easier to have them on the perf board as well.

Transistor driver board

Arduino wired to transistors

This all buttoned up pretty nice in the power strip chassis.  I cut up a water bottle and put plastic around the electronics to keep any inadvertent grounding from occurring.  I did cheat on the power supply a bit and instead of making one, I took one from an old external HDD chassis as it slipped inside the power strip snugly.  I just wired the 120v in the strip, as there was plenty around!

Electronics all buttoned up

And an admittedly crummy video of it in action..

All the new online small-batch circuit manufacturing was intriguing and thus I took the desire for a fan controller, coupled it with wanting to hold in my hand something I designed on a computer and this is where I ended up with:

Circuit Closeup

This is actually two circuits on the same board.  The ‘cheap’ bundle order was based on this size boards so i drew them up in ExpressPCB using the default size and found I could fit two per board.  The schematic and layouts are here.  The brains of the operation is a TC649 from Microchip.  It is designed for use in servers and other electronics as a fan controller and it has all the goodies:

• PWM Based
• Speed detection – using only a two wire fan!
• Stall protection

What the above means is the chip can sense the speed of the fan and can adjust it’s speed based on the temperature detected by temperature probes (thermistors) I have remotely in the top of my cabinet.  What sometimes happens when the fan is going too slow (low temperature) it will ‘stall’, or stop turning.  This chip can detect that and bring the power up to 100% for just long enough to get it spinning again.  Pretty handy chip.. and you can get them for free for onesey-twosey type stuff.  They offer samples on their site!

This particular one is in the cabinet for my home theater equipment driving two 8″ fans – one in the top of the cabinet and one in the bottom.  The wires in the center carry over the power (12v/5v/ground) from the left side to the right, so I only used one regulator.  In the future I would probably put traces there for that job that could be cut if the board was separated.  One update I had to make for the very large fans is moving to larger power transistors.  I scavenged some from a broken power supply, complete with heat sink!  The picture below shows the configuration which has been running for over a year.

Fan controller with upgraded transistors

And the remote thermistor probes at the top of the cabinet shown below.  Wired with cat5 of course

Thermistor probes

Circuit Diagram: