I assembled my new printer last week and found some issues with the z-axis.

The first issue was that the right z-axis motor was not staying in step with the left. if the right moved up 1cm, the left might move up 5mm. It was suggested that this might be an electrical balance problem and that I should turn the potentiometer on the Anet board to fix this. I turned the potentiometer about 90% clockwise and that seemed to do the trick. They were both moving the same distance as each other now.

The next issue happened when I tried to print a calibration cube. It came out 20mm*20mm*10mm. Only 10mm high.

After much research and questioning, I settled on three steps to take (after taking some wrong turns involving changing steps/mm):

1. check the alignment of the lead screws. The frame that comes with the printer is off by fractions of a mm in some cases, and this can cause pressure that makes it harder for the motors to turn the lead screws. I found that the left motor was off by about .5mm, so I used a drill to extend the holes in that direction and reseated the motor directly under the lead screw (basically dropped the lead screw down through the brass thing on the x-carriage and moved the motor to where it fit best).
2. turn the potentiometer a bit further. it turns out that potentiometer is not for balancing the two motors – it controls the overall current that goes to the motors. If there is too little current, then the motors can’t lift the x-carriage.
3. change the steps/mm back to the stock 100/100/400. The default settings of 100/100/400 are based on the actual hardware. Stepper motors turn through precise degrees, so it is a mathematical issue to figure out how many turns it takes to move the belts or lead screws a certain distance. If the motors, belts, pulleys, and lead screws have not changed, then the default values should be perfect in all cases.

When this was done, I printed out another calibration cube, and this time, it was perfect. See the image – the cube on the right is the latest.

The next issue to solve involves large prints. My first project is to print out a new printer. I’m printing a CoreXY printer. There are STL files available on Thingiverse for this.

I found almost immediately that the default print bed is inadequate.

1. You can’t print directly onto the aluminium, as the plastic will just slip right off. Even if you could, sometimes a print will get stuck and you might have to chip it off, damaging the print bed.
2. The Anet-recommended method is to place painters-tape (a paper tape designed to let people paint without worrying about getting the paint onto glass, etc). This is not good enough, because it involves placing tape precisely so there are no seams, making sure there are no bubbles. It’s annoying work.

The solution I settled on, and others settled on, is to use a glass overlay. I got a cheap picture frame and took the glass from it. It was a little too long in one dimension so it’s currently poking over the edge on one end.

Keeping the glass in place is a problem. The recommended method is to use clips (like from clipboards) to stick the glass to the aluminium, then remove the handles of the clips so they don’t snag on things. I don’t have any yet, so for now, I’ve settled with some teflon tape. It’s designed to not warp in high temperature, so should be fine for a while, until I can print out some proper solutions for this.

The final problem of the day- I was printing out a piece for the new printer. It initially started very well, printing onto the glass like a dream. However, after the print was about 1cm high, it suddenly came lose from the glass and the printer started printing out a “nest” instead.

After examining the print, I found the problem – the bottom had curled upwards from the bed, reducing the volume that was in contact with it, and making it more likely that the movements of the print bed would break the print free.

The curling is caused by temperature differences between the newly printed layers and already-printed layers. As the plastic cools, it shrinks. If there is even the slightest gap under the plastic when this happens, then this will cause curling.

The solution was to spray hair-spray where the print was to contact the bed. This lays down a thin layer of binding plastic on the glass that increases contact between the print and the bed, reducing the chance that it will come lose.

After printing, removing the print from the glass is a simple matter – I whack the bottom of it with a metal ruler. Instant release, and no damage.

I bought a new printer, the Anet A8, which is a Prusa i3 clone. From the prints shown on the “official” Anet A8 support group on Facebook, it’s a very good printer indeed for its price.

The Facebook community are also avid upgraders and publish a lot of posts explaining what they’re done.

My previous machine was a Makibox, which rapidly fell apart. Even when combined with the parts from a replacement Makibox, it still only lasted a few more weeks, and all prints done from it were fixes for itself.

I’ve had problems with the A8 already, but the community is very good and the printer is common enough that there are plenty of articles online about it.

Example problems I’ve had:

• Z-axis motors out of sync. There are two motors controlling the z-axis (up/down). They both run basically through the same circuit. If they don’t receive the same power to both, then one might be a bit jerky or not move at all. On the Anet motherboard, there is a potentiometer between the two Z-axis controllers (looks like a little silver circle with an X in it). I just turned that 1/4 to the right and that fixed my problem.
• I printed out a calibration cube and it was only half the height it should be – 2cm*2cm*1cm instead of 2cm*2cm*2cm. The problem here is a mixture of the slicer’s code, and the firmware of the printer itself. I think I have this sorted now – the steps per mm were off in the Z-axis. After some experimentation with the console in Pronterface (and some measurement of movements with a ruler), I think I got it right. List of gcodes are here.
• clogged nozzle. Well, this one is an old friend… Sometimes you will find that the extruder doesn’t extrude anything at all. In my case, it was a clogged nozzle – there was a lump of plastic at the top of the throat leading down to the hotend and nozzle. I had to melt it out with a tea-light candle, because it was just /not/ moving. I then poked a guitar E-string through the end of the nozzle (0.4mm – too thin for a needle) to make sure that was clean.
• cables a little too short. There is no “play” in the cables, so I found that sometimes the extruder stepper’s controller wires would literally get pulled out as the X-carriage moved. Solution – remove the clip that was holding the black plastic wrap to the back of the LCD, so that the wires could be moved.

Overall, I am /very/ happy with this printer, but I’m still wary from my experiences with the Makibox, so the first project I’m working on with this new one is to print yet another printer – the Smartcore CoreXY.