z-axis not moving far enough during print

[Deleted]

Hi all,

I spent this afternoon tweaking the alignment of the carriage against the bed and varying the slic3r settings, but no matter what I did, I'm having a problem with what appears to be the amount of z-axis movement between layers.

The actual symptoms of a print look like this:
pancake.io/3e35fd/content/sumpod/raw/IMG_20121103_204017.jpg

If it's not obvious what's happening, the first couple of layers print okay, but after about 3 layers, the hot end is far too close to the print, and ends up buried in it (that's the circular indent you see on the 3 prints), and so first will mush up the layer, and if given the chance, will end up moving the print around the platform rather than building it up.

Googling around turns up this post which seems to describe the issue I'm having, right down to correct movement with Printrun, but incorrect movement on an actual print: forums.reprap.org/read.php?262,157372

However it's not clear to me what exactly the fix for this is. I was under the impression that the step-sizes encoded in the firmware were correct (reinforced by the successes I've seen from lesj on the main sumpod forum), so changing them doesn't seem like it's the right thing to do.

I'm not really sure what to do next; I don't know how to tell slic3r to move a bit further between layers to stop this happening. Any suggestions?

Cheers,
- Piete.

[highcooley]

Hi Piete,

Hmm, doesn't look very nice...definitely looks more like a pancake(.io) than a print :-)
Have you already been able to solve this problem?
It is important that you a) set the correct step-size in the firmware (eg. 106.76) corresponding to b) the correct microstepping (eg. 1/16), set by the jumpers on the RAMPS board. Unfortunately, the Sumpod wiki is currently down and I am not sitting in front of my printer to check what microstepping value is set by default. However, to check, move your z axis 20mm with your software and measure. If you get 10mm, your microstepping setting doesn't correspond to your configured step-size. You either have to double the step-size in the firmware or half the microstepps (1/16 -> 1/8) on the RAMPS.

As you mentioned on the official forum, it could also be too much current on the driver. However, you could see the movement being interrupted when moving the axis over a longer distance by the software.

Good luck!

Cheers,
Andy

[Deleted]

Hey Andy,

Yeah, if you need a large quantity of filament turned into a lot of molten PLA, I'm ya man!

When I use printrun and ask it to move 1mm, I can measure that change with some calipers, which is why I hadn't looked into the microstepping of the board.

The driver is stepping okay, it's not jerky or problematic like the overheating one I had on the x-axis.

I had a message from Richard about the extrusion rate being set too high in the firmware, but I haven't taken a crack at this yet. I'm doing some reading now about calibration so I can work it through and get some measurements. I shall resist the temptation to just bang in the new number and see what happens (for now) ...

Will keep you informed!

[highcooley]

All right, looking forward to hearing your news. I rate it sort of a bad habit of Richard just answering any questions with "use my firmware" or "set it like this and it works" which doesn't help a lot to understand the actual problem and the right solution. Good luck!

[hindessm]

If the Z axis movement is correct then the extruder is a good thing to check. You should be able to calibrate the steps for the extruder fairly accurately with a process similar to the one described at:

richrap.blogspot.co.uk/2012/01/slic3r-is-nicer-part-1-settings-and.html

You can test changes by modifying the extrusion multiplier in slic3r if you don't want to change the firmware right away but I recommend correcting it in firmware as described in the above post once you have it calibrated correctly.

I wish Richard would reply to people in public instead of in private so we can all learn from his answers.

Regards,
Mark

[Deleted]

Mark: thanks for the link; I'd come across it, but had neglected to actually go through the steps to test it ... so no more neglected steps

I ran through the process: marked 50mm of filament, extruded 15 and measured 26.35mm actually extruded. The current extruder steps should've been set to 102.0 steps per mm, so it took 1530 steps to extrude 26.35mm, so the steps per mm should be 1530/26.35 = 58.06.

This coincides pretty well with Richard's suggestion that 102.0 should actually read 60.

For completeness (and those reading this later), the numbers being discussed come from the Configuration.h file of marlin firmware, and now (in my case) read thus:

#define DEFAULT_AXIS_STEPS_PER_UNIT {84.88264,84.88264,800.00,58.06} // default steps per unit//[SUMPOD specific]

Those numbers represent the steps per mm of the X, Y, Z and extruder stepper motors, respectively. (see: www.matterhackers.com/news/3d-printer-firmware-settings-stepper-motor-configuration)

So the good news is that I'm now getting more or less the correct amount of filament pushed through. 2/6 test prints seem to suggest that this has solved the issue where the hot end was getting buried in the model. The other 4 test prints were failures due to not-starting properly.

I'll get some photos of the prints that have been done up in the next day or two (just 10mm cubes). They exhibit a kind of squashed look, where the height is < 10mm, and the widths are 11.something.

So clearly I've still got work to do, but things are looking more on track now!

Thank you both, the next instalment is likely to land sometime over the weekend!

As a funny aside, I've just started a new job, and one of the guys there has built a Mendel (I think) and got it spitting out some nice prints. So while I say weekend, I may return sooner

[airtripper]

These are my settings:

#define DEFAULT_AXIS_STEPS_PER_UNIT {106.76, 106.76, 800, 67.16}

With stepper jumpers at 1/16, 1/16, 1/8, 1/8 for axis X, Y, Z, E.

[Deleted]

Thanks airtripper, I should have mentioned I'm using an Aluminium Big- I don't know if there are significant differences between the originals and alus. I shall check my stepping sizes later on!

[wills]

Which firmware are you using in the control board? I found Marlin to give better results than the 'standard' firmware.
I also have one of the 'old' MDF Sumpods , which has had a few mods but now appears to works pretty well.
Looking at your piccies in the blog (very good) you have some strange effects in your printouts. Have things improved since then?
It also look like you had a few problems with the hot end/PLA swelling in the PTFE tube. I went through this as well, and ended up changing the extruder drive to a 'Airtripper' and changing the hot end to another type, but I did get some good print from the originals, just not very reliably.
Regards
William

PS Many thanks to Mikematt for setting this up.

[Deleted]

Sorry for the hugely delayed response William, but in answer to your questions:

1. I'm using the firmware branched from Marlin, supplied by Sumpod.
2. Regarding print effects, the short answer is: no improvements. In fact, after wrestling with it for 2 months last year, I gave up trying to get it to print and moved on to other projects. I have yet to produce *any* full prints, and it's coming up on a year since I first put my order in.

I'm planning on visiting our local reprap group at the end of this month to either get it working or declare it dead!

My biggest problem thus far has been the turn-around time for testing and experimenting. In 2 hours I can usually get 1 print to stick to the blue tape, but I have spent literally days doing nothing but cleaning the bed, pushing start on the printer, watching the PLA melt to the hot end, move it around the bed a bit, pushing stop, cleaning the hot end, and then starting the whole process again. I have the parts now to construct a heated bed, which I am told solves this major issue, but good grief! What a show-stopper! Especially considering I bought a model that "included heated bed", but really only came with some of the parts and no instructions on how to assemble it or what was missing!

More news next month ... possibly ...
- Piete.

[wills]

What are you printing onto? Plain glass or blue tape ?
Does the PLA not stick to blue tape ? If you are just using hot glass , try painting a layer of watered down PVA onto it.
Is the printing starting and then going wrong after a few layers or right at the beginning?
Are you in the UK?
William

[Deleted]

I've made some progress courtesy of the guys at Southampton's RepRap user group, but let's cover the questions in order:

>> What are you printing onto? Plain glass or blue tape ?
I was using blue tape with no heated bed, and it was difficult to get it to stick. I'm now using plain glass with a heated bed, and it's much easier to get it to stick, but there are still some improvements to be made. Primary issue now is bed levelling and getting an actual flat surface to work from.

>> Does the PLA not stick to blue tape ?
Not very well, no.

>> If you are just using hot glass , try painting a layer of watered down PVA onto it.
That is a good idea, I'll give that a go next!

>> Is the printing starting and then going wrong after a few layers or right at the beginning?
Once things stick, the build tends to work okay. I've got a load more pictures to get up and a blog to update after February's meeting. The next meeting is this coming Sunday, so with luck I'll have some more success and get a chance to update the ol' blog with some more pictures and so on.

>> Are you in the UK?
Yes; Hampshire

[airtripper]

Printing on to glass has worked very well for me so far and I keep the heated bed at around 60 degrees C throughout printing. I give the glass a good clean with acetone oil free nail polish remover.

I have found that it is important to find the best temperature to extrude each particular filament you have in order to get the best adhesion to the glass. Sometimes if the filament is not hot enough it can warp and become unstuck from the glass.

I have a real of PLA that extrudes best at 190 C while another real extrudes best at 220 C. Unfortunately, a lot of people don't have hotends that are stable and reliable enough over wide temperature range.

I have the MDF Sumpod and have the issue with nuts and bolts coming loose on the X, Y and Z axis. This really upsets by bed levelling efforts. I now plan to replace the MDF axis and hopefully gain a larger printable area in the process.