Taig CNC to 3D printer build

I've got my ABS on order, so I'll limit my parts to decorational stuff for now. I did post my ball chain sprockets on Thingiverse yesterday if you want to try them out. Also, if you ever see a posted part that you would like the orginal CAD files for, just email me.

 

My thingiverse name is RCENGR. I bought the ball chain in the lighting aisle at my local Walmart. I'll measure it and post the information with the thingi, something I should have done from the first. I found 100' reels of stainless steel ball chain available, not exactly cheap but it would provide a lifetime supply of rudder linkages.

 

The #6 is the chain I used for my sockets. I'll modify the file for a 5mm shaft and you can print one to see how it works. With the way I'm currently printing inside holes (not well) I'm not sure that I would end up with a sprocket that didn't have any wobble. Hopefully you can do better. I'll also beef up the hub a little, so that you can at least use a 6-32 set screw. For real durability, it may have to be modified to trap a nut for the set screw. At least you can try it out with some cheap lamp chain and if it doesn't work you won't be out much money.
When I looked up the ball chain size, I noticed they had some nice connectors with loops at the ends. They might be more durable than the printed connectors that I designed.
I'm looking forward to your pipe delta. I'm still flopping back and forth over whether I want to build a delta or Mendel style machine for my dedicated printer. The deltas are beautiful mechanically, but the Cartesian printers offer a bigger base print area.

 

What do you mean by very low speeds? I'm only printing at 20mm/s now.

 

My ABS arrived today, along with my 2560 board from Sainsmart. So now I can print some production ship parts. I ran two calibration shapes just to check out the extrusion settings. It appears to be right on, using average diameter and an extrusion multiplier of 1. The PLA I have expands too much, I have the extrusion multiplier for that down to 0.95. This is the first time that I have run the bed heater full on, and it is coming up a little short. I'm able to reach about 90 degrees C, about 20 degrees too low. I'll put some better insulation under it and hope it does the trick. For now I'll accept the corners lifting on my Golo dual motor mount.

 

I'm familiar with brim, I had to use it to keep the rudder on my lifeboat from lifting. The part was scrapped anyway. I forgot that the power I'm using for the bed has a timer on it. It was set for 90 minutes, so it turned off about 80% the way through the print. The part broke free and I had a lot of printing in the air when I came down to check on it. Now I have the timer set for several hours.
I forgot to make cutouts for the motor wiring, so it was OK that I had to re-print. It's on the machine now, and printing fine.
I added a 1/8" piece of foam below the heat bed as well as some thermal paste to help transfer the heat to the glass. Those changes were enough to get me up to 100 degrees C, which looks like it will work fine.

 

The motor mount for the Golo finished this time and I'm very pleased with the result. The motors snap in place and are held firmly, yet you can remove them without tools.

 

I've been making some improvements to my Taig printer. First, I've reworked my heated bed. Now it is a layer of 1/8" aluminum for the print surface, the 8x8 heater bed, 1/8" Rohcell foam, 3mm extruded foam, and finally a 1/8" aluminum base. The whole sandwich is held together with 2-56 screws in each corner, spaced by a 1/4" high phenolic spacer. The updated bed heats in 15 minutes instead of 30 minutes and it takes about 1 amp (@24 volts) less to maintain temperature. The bottom plate is actually cool to the touch. The key to making it work is the Rohcell foam, which is capable of withstanding over 200 degrees C. While not the lightest bed, the extra weight isn't a problem for the Taig.

The next upgrade is converting to a Mega 2560/RAMPS 1.4 as the controller instead of using my DeskCNC controller. The advantage of the conversion is that I can take advantage of all the 3DP control codes built into the Marlin software and I can print from an SD card. The SD card capability is probably the most important. Right now I'm limited to 1.38 Mb files - what fits on a floppy disk - probably a unique limitation among all the 3D printers out there. From left to right is the RAMPS board, LCD controller with SD card, and the Gecko servo drivers. In the blue circle is the parallel port breakout board.

The first problem I had was getting the step and direction signals to the servo drivers. I knew from reading in the forums that I was going to have some problems. It seems that Marlin only holds the step pulse high for about 3 ms, which is not enough to work with CNC equipment. So I programed a step pulse extender using an 8 pin PicAxe controller. It didn't work. Finally I got out my Gecko servo driver paper work and saw that they would work with 1.5 ms pulses. So once I bypassed the parallel port breakout board every thing worked.
At this point I have a workable version of Marlin loaded, all three axis and the extruder steps are calibrated. I ran the PID auto tune tonight so the temperature control for the extruder is set. My power supply went a little flaky on me, so I have to fix that before I can attempt a print. The first print, once I know it works, will be cases for the components so I don't have to be worried about shorting anything.

 

I got mine from CST: http://www.cstsales.com/Rohacell_31_IG.html I've been hauling it around a decade or more, so the price has changed a little.