(3DP) tuning the Reprappro Ormerod 3d printer

After several posts in the U-Boat thread dedicated to my printer problems, I have decided to start my own thread. I recently finished assembling a RepRapPro Ormerod 3d printer. There is a lot I need to learn about this machine, and 3d printing in general, so this thread is dedicated to my learning process and all the questions and issues that I encounter.

my laptop and printer. On top of the laptop is a heat shield, which helps the print bed heat up faster.

My first print after initial calibration. A moustache and monocle. Keeping it classy

starting in the back: an ABS x-axis runner. Middle row, left: ABS XZ brace, ABS and PLA wall thickness test pieces, Y belt holder. Front row: ABS extruder, PLA Y-axis tensioner, Z axis gears, other side Y belt holder.
Right now, the printer is able to print 3d objects from ABS plastic. They are not, however, very precise prints. Every object I have printed so far that is supposed to fit inside another object has not actually fit. While this is fine for a comedic moustache, it isn't good for gears, bearing mounts, and other high-precision parts. I am not sure exactly what is wrong yet, so I need to come up with some tests to measure specific variables to identify the problem and resolve it. A few possible causes:
1) incorrect steps-per-mm for the given axes. To check this, I need to print parts of several different known lengths in each axis, and check if each printed dimension is off by a consistent percentage.
2) axis skipping steps. Unlikely, since I have not observed any symptoms since correcting an issue with the Z axis several days ago. The motors make a unique sound when skipping steps, and the offset is usually distinctly visible in the printed parts. I have neither heard nor seen any signs of skipping steps on any axis since I oiled the Z-axis threaded rod several days ago.
3) over-extruding or under-extruding. To check this, I need to print parts with a wall thickness equal to the nozzle diameter, and measure the thickness. I also need to check how well it prints flat surfaces.
4) incorrect temperature. To test this, I will print one part several times, with a 5 degree difference in temperature between each one, and see which one gives the best results.
5) insufficient belt tension. If my test for (1) reveals an identical linear error for each length, rather than identical percentage error, it is probable that my drive belts are not sufficiently tightened. If that is the case, I will install my upgrade parts with adjustable belt tension.
Another issue that recently cropped up is a probable nozzle jam. Now when I try to extrude, the motor feeds plastic into the nozzle for several seconds before any plastic actually flows from the nozzle. It then flows in a large spurt before steadying out. If it stops feeding for a moment, ie when retracting to move to a different position on the print, it again takes several seconds before releasing another large spurt of plastic. I will be soaking the nozzle in acetone this afternoon to try and clear the blockage. Once that is done, I will re-assemble the printer with some of the upgrade parts I printed.

 

I was sorta hoping to show it to you today, so you could take a look at my current belt tension, etc. But I've got to fix a partial nozzle jam first. It's soaking in acetone right now, and should be ready to print again by this evening. I'll be installing a few other upgrades at the same time. We'll see then if belt tension was the issue.

 

Keep in mind that undersized holes is a known problem with the slicer software, Slic3r in particular. I tend to treat the part like it is a casting - close to final dimensions but holes that need to be precise need to be drilled or reamed out.

 

It's not just inside diameters, it's outsides as well. Every dimension is about 0.3mm less than it is supposed to be. I ran a few tests yesterday and it was off by the same amount no matter how large or small the part was supposed to be. Very odd. I tried measuring my plastic, but found out that it is extremely inconsistent. It isn't even round! I measured 1.73mm and 1.76mm in the exact same spot, just by rotating the calipers! The other spots I checked were just as inaccurate, if not more so. It also appears to be contaminated with water. Out of curiosity I tested my other roll of grey, and found it was just as bad. I'm not sure if I should test my other colors, as well...

Well, I just finished clearing the plugged nozzle and installing tension-adjusting mechanisms. The X and Y axes are now much tighter, but still moving freely. I can't wait to try it out tomorrow.

 

So it turns out I got a bad batch of grey ABS. I ordered two 1kg rolls of the stuff, and both had severe problems with moisture straight from the box, despite being shipped in a sealed bag with dessicant. I compared batch numbers, and they were indeed from the same batch. I then checked my roll of black ABS, and it was a different batch. So I took a chance and tried it. Much better! I need a higher temperature to print the black (currently running at 245C) but no problems with moisture or clogging. Here's a picture of the first calibration print, to test all the new upgrades I installed:


A few upgrades I've installed:
adjustable-tension X and Y axes. You'll notice the white parts on the end of the X arm, and barely visible on the far side of the machine. Both axes are now much tighter, which should allow me to run higher speeds without skipping steps. While I was at it, I installed an ABS X runner (not visible) and extruder body. Last and perhaps most significant, I re-mounted the printing bed with adjustable mounts, and spent most of this afternoon levelling the bed and improving the Z-axis sensor. It is now level to within 0.1mm over the entire 200mm travel for both X and Y axes. No more crashing the print head if I forget to run the auto-level program.
I still have other upgrades to do. I've got a solid aluminum X-arm and printing bed support on their way. But those are coming from the UK, so it'll be a while until they get here. Until then, I have a submarine to print...

 

Grrr.....and Tuggy hasn't printed out mine yet.

 

Well I've been looking for a good model of a 726 class with DDS, but haven't had much luck. I guess I'll go with something a little older:
http://www.thingiverse.com/thing:44975
Or even older than that:
http://www.thingiverse.com/thing:161589

 

The even older than that! Go for that one.

 

I propose a special ruling: If you build a working (submerging) RC 1:144 scale Nautilus (Nemo's, not USS) sub, then you ARE allowed to ram, since you won't have a deck gun.

 

I've seen some difficult-to-sheet ships, but man... that would be tough... Still, at 19.1" length and 2.18" beam (per Jules Verne's official dimensions from the book), it would be entirely do-able. Wait! On closer inspection, only the middle two hull sections are within 45 degrees of vertical. The rest of the hull is outside the 45 degree rule, or is superstructure. Hmmmm...

 

Work continues on improving my Ormerod. It seems I am printing new, improved parts for it almost every day now.

Today I encountered a distressing new failure mode. While printing landing gear for an RC plane, the rapid oscillations from printing the in-fill were violent enough to jiggle some contacts for the temperature sensor to the hot end. This caused the readings to drop by 100C or more, which the printer would counteract by sending more power to the heater element. End result was indicated temperature rapidly alternating between 100C and 300C+, with actual temperature hovering around 300. As soon as I saw it, I pressed the connector back together, which fixed the problem temporarily. It worked itself loose again in just a single layer, so I cancelled the print and set about printing an improved hot end that would not be so susceptible to vibration and oscillation.

I also found something else out. The plastic that I'm using, a black ABS mixture by Filamax, is extremely resistant to acetone. I tried chemically welding several parts together, but none of them would stick. I couldn't even make a slurry of acetone and scraps. Very frustrating.

 

Disaster has struck again! I think I fried my Duet control board when I crashed the printer today. To reach the bed temperature needed for ABS to properly stick, I often place a sheet of cardboard insulation on top of the bed, and remove it just before starting the print. Unfortunately I got careless and today I forgot to remove the insulation. It got through about half of the first layer before the entire thing jammed up. By the time I noticed, the printer had stopped communicating with the web interface, and a thin stream of smoke was rising from the printer head. I immediately unplugged it and raised the head enough to remove the cardboard, but the damage was done. My printer no longer responds to USB or web interface. If I plug in the power supply, both the bed and nozzle heat up without control. From what I can tell, the smoke was from the nozzle heating up too much, because I do not see any signs of burning on any of the wires or components. If I plug in the USB connection, the USB light turns on, but the other lights for the USB testing mode do not light up.

I'm not sure if there's anything I can do to salvage the control board. I will probably have to order a replacement from the UK, which means my printer will be down for at least a week

 

Ouch. If you disconnect the bed and nozzle heater can you still move the motors? Replacing those surface mount FETs would not be fun, but it can be done. Or you could add TO220 size FETs on a daughter board to replace the on-board ones.
Unfortunately, I've burned up my share of all-in-one boards. If you have a four axis controller and burn up the stepper controller on one axis, the whole board is worthless. So now my CNC has a separate controller on each axis. For the same reasons I like the RAMPs board. It's not as compact as an all-in-one board, but the stepper drivers can be easily replaced and I can replace a FET if I burn it up.

 

Unfortunately it appears that the whole board is down, it is not communicating or responding to any commands. based on the specific symptoms, The Reprap forum theorizes that the 3.3v supply burned out, although I have no idea how. I have ordered a replacement which should be here by the end of the week.

 

I was all set to connect the replacement Duet board today, when I noticed an important difference. While the old board had pin connectors, the new one has screw terminals. So all the connectors on my wiring harnesses are not compatible! What should I do!?
should I:
modify my wiring harnesses to fit the new board?
de-solder the screw terminals and replace them with the pin connectors from the old board?
make some kind of adapter (I don't know what yet)?

In other news, I believe I have spotted the cause of the old board failing. Now that it is removed, I can look much more closely at it, and I noticed what appear to be small scorch marks surrounding a tiny pinhole opening on the big square microchip in the middle of the board, the one with the blue sticker on it.