Taig CNC to 3D printer build

ok got it

 

I finally made some more progress on this project. The new stepper motor is on the extruder and a more powerful stepper motor is waiting on the bench just in case. After some failed attempts at putting together a hot end controller with some opamps, I finally just wrote a Picaxe program to control it. The picture shows my breadboard test setup. The program works and I've been able to set the gains for decent performance. Now I just need to build a permanent board and solder everything together. I also got my RAMPS 1.4 boards today, as you can see in the middle. So if I get bored, I can start soldering components on those.

 

I could program in a delay and a temperature change, but I think I'll keep this controller basically manual. It only needs to tide me over until I buy the Arduino and start playing with Marlin. I do appreciate the hint, I'll plan on manually turning the knob down after the bottom layers are on. The one concession I made to automation was to put an enable input on the circuit. That way I can connect it to the CNC enable system and a fault will shut it down.

 

I was able to complete a circuit board for the hot end controller and solder it up tonight. And it still works, always a good thing

 

It prints!!!
I finished wiring up the extruder and after checking that the hot end temperature control worked under load, I attempted my first print. Here it is:

The calibration object is a 10mm x 10mm square, 5mm high, with a 0.25" hole in the middle. Here's a couple of the calibration cubes with some 25 man life rafts. The ball bearing is supposed to fit in the hole. Overall the positioning was right, which is the advantage of using my existing CNC machine, but too much plastic is being extruded.

I adjusted my extruder steps/mm down about 15% and it is a lot closer. More calibration on the extruder steps is still needed. With a "close enough" calibration I attempted to print an actual accessory, a 27' whaleboat.

I added the painter's tape to help keep it stuck to the glass. I had no problem with the first layers adhering to the unheated, untreated, glass, but they cooled off too fast and peeled away before the print was finished. Unfortunately, I had a servo fault about 75% into the boat print and it didn't finish. It could have been a random fault, or I may have my rapids set a little too high.

Even with the incomplete calibration, I'm pleased with the finish of the part. It was sliced with 0.329mm layers.

 

The thick layers are just to cut the print time down. I figure the .3mm is good for utilitarian parts and I will go lower for parts I want to look good.
Tugboat - are you looking at the rudder on the far left when you say that you think one axis is slipping? The outside of the rudder is curved, not straight, so I think everything is OK.
Tonight I worked more on my extruder steps. Part of my problem is that I entered an estimate of 1.65mm for the filament, and it is actually very close to 1.75 mm. Once I corrected that, I wasn't over filling as much. I also measured the filament coming out of the hot end at 0.69mm. That seems like a lot of swell for a 0.5mm nozzle, but I decided not to mess with the nozzle size in Slic3r. I also played with retraction on the extruder. First I tried 3mm, but I didn't like the results. My last print was at 0, which is not quite right either, so more work needed there. Here's the completed 27' whaler:

The steps at the bow and stern where the gunwale curves up is very evident because of the thick layers. I also have blobs at the start of each layer. Hopefully getting the retract setting right will take care of those.

 

I made some more progress calibrating the printer today. I think I have the retract settings close, as well as the extruder steps. I printed a 15x15x15 mm cube to test everything out. It was printing fine, with about 3mm to go, when I decided to turn the cooling fan on. As soon as the fan was on the part popped off the bed. Oops. Because of the slow speed I print at as well as the extruder temperature, about 200 degrees, I guess the part cools fine without any external help. This time I'm using layers of 0.2mm. Here's the calibration cube.

In the picture it looks worse than it really is... I found the super macro setting on my camera and used a low angle light. Although it looks like it tilts in the picture, it doesn't really. The corners concern me the most, since they bulge out. I'm wondering is these artifacts are just because the original shape has square corners. Maybe if I put a 0.25 fillet on the corners they would be fine.

 

I'm using Slic3r for my slicing program. I'm a couple of builds behind the latest stable release, so I'll get that updated.
I also just noticed that the corners of the bottom 1/4 look better. That's the point that I turned the temperature down to about 185. Maybe at that temperature it's warm enough to extrude, but there is a little extra tension. I'll keep the temperature above 195 and see if it makes a difference.

 

I'm not using ABS, all that I have right now is PLA. I figured I would start with the PLA and work myself up to the ABS. That way I wouldn't initially need a heated bed or fume removal. However, the lack of a heated bed has already caused me a number of problems, even with the PLA. So the heated bed is being cut by the CNC machine as I type.

 

I can do a 12" x 5.5" heated bed. I could probably squeeze out a little more than 12", but it doesn't really matter since the copper clad boards are only 12" wide. And the 5.5 is absolute max, I would probably make it slightly smaller to give a little margin. Have you considered two 6x6 or two 8x8 boards? That would give you a bigger build area, but it would be more expensive. The raw materials for the 12 x 5.5 would only be about $10.
Here's the heater finished. I have two elements, each designed to 3.6 Ohms. I run them in parallel on 12 volts and in series on 24 volts. Actual resistance for the series case came to 7.0 Ohms, so I'm very pleased with how close it came to the designed value.

 

The heat bed is installed and works great!. With the heat bed the PLA sticks great to the plain glass print surface. Even after I wait for it to cool down, I need a razor blade under one edge of the part to pop it loose. After reading pages and pages of people having difficulty getting their prints to stick and all the exotic substances they put on the surface, I had been a little worried. I knew the level bed of the mill would help, but I never expected it to be this easy.
I still have more calibration to do, but for now I'm declaring that initial operational capability has been reached. Last night I printed a motor mount for my Golo and tonight I printed a rudder linkage that uses ball chain. Now I need to get back to getting my ships ready to battle.