Re: [MakerGear] Printed parts coming out shifted.

[Rick Pollack]

This in generally caused by one of two things:
1) try turning up the stepper driver for the axis that is skipping - turn it clockwise a hair (really the thickness of a hair)
or
2) you are extruding too much plastic and the nozzle is running into the plastic.

Rick

On Mon, Jul 16, 2012 at 4:01 PM, David Jasiewicz <da...@trickconcepts.com> wrote:

I am having extensive problems with my parts coming out shifted.  I am using ponterface with Slic3r 8.4.  I have checked everything mechanically on the machine and it is sound as far as I can tell and the electrical connections seem good.  The baseline is shifting multiple times in multiple directions during the print.  Any ideas would be greatly appreciated.  Just or reference, the part i the photo should be cylindrical in shape.  You can see there was a major shift after only a few layers.  thanks

[makishev]

The same problem. 

[NCBob]

I just resolved the same issue with my printer, and now it's smoking fast on parts.

For me it was two things that I did and the shifting and banding is gone.

First thing was I had my belt too tight on my x, it could also be to tight on the y, which I had already adjusted before this.

The other thing was my acceleration and jerk settings were off. I did a search on the RepRap forums and found an awesome link on how to do some tuning on your printer. Here's the relevant post http://forums.reprap.org/read.php?1,131740,131740. And here is the section for fine tuning your printer :

tbfleming [ PM ]
Re: Y axis stutter on MendelMax
May 07, 2012 06:11PM Registered: 2 years ago
Posts: 72
It sounds like you're past the hardest part (alignment). Here's the technique I used to tune acceleration (I assume you're using Marlin). I created a simple g-code file:

G1 Y100 F5400
G1 Y100.7 F5400
(repeat these two lines hundreds of times)

Your numbers will probably be different. I started with different numbers then adjusted them until I hit the worst case: the motor and bed stood still (this is a form of resonance). The goal is to reduce acceleration to the point where it can't resonate.

While it's resonating, paste commands into Pronterface to adjust the acceleration (Pronterface interjects manual commands into the running gcode )

M204 S1000.00

You want the highest value that allows it to move back and forth the full distance without overshooting. Once you've found it, use the following to permanently store the adjustment:

M500

If you still get stuttering on real gcode files then you might need to reduce it farther. When you're using Marlin and a heavy bed, the most important number isn't the top speed, it's the acceleration. Acceleration controls the forces on the machine.

Todd
Reply Quote Report

[William Bond]

 I think the problem is missing steps due to noise pickup on the end stop inputs. On my machine I shielded the motor wires with aluminum foil, replaced the end stop wires with shielded cable and removed the end stop wire runs from being in motor cable bundles.

[Dean Piper]

I hear of people with this issue still and I don't see the need for such extensive solutions. I put a good twist in all my wire groups and the crosstalk is gone. I haven't had an issue with that in ages and all my wires are bundled together. I even run 24 volts which should cause even more trouble. But I have none.

It can be a pain to do but I found the easiest way was to put the end of the wires in a drill chuck and hold the wire taught while running it slowly. It will quickly put a permanent twist in the wire and hopefully end the issue without the need for expensive shielded wires or aluminum foil.

Also, to the author of this question. Your skipping could be caused by steppers that are turned up too high as well. It will cause the motor driver to momentarily cut out and do exactly what you show in your picture. You may need to turn the amperage down. It's really a delicate balance, don't be discouraged if it takes a while. Once you get it set you shouldn't have to do it again for a very long time.

Hope that helps.

[Rick Pollack]

if the drivers are at the default level...they can be turned up (clockwise) a hair.

[TJ]

On that note, how can you determine the default level setting? Is the a set resistance value that can be measured? Or measure the current to the stepper?

TJ 

[Rick Pollack]

the default level for mosaic is where it is set when it leaves here.

[NCBob]

There are a few things that can cause the shifting, the trick is figuring out which one it is. For me it was my belt being too tight, I loosened it just a hair to match the tension on my Prusa, and the shifting disappeared without having to 'shield' or twist the wires. My Prusa has never had this problem, probably because for me it was caused by the belt, and I can't over tighten them on the Prusa because of press fit rods.

One thing I would highly recommend, change only one thing at a time and test it. The reason I say this is because if you introduce another issue while trying to fix this and you've made multiple changes it'll be nearly impossible to figure out which of the changes you made are causing another issue. Small steps when troubleshooting.

[makishev]

Looks like my problem was too much infill

[makishev]

Looks like there was another problem besides too much infill. 

Too much tension on the plastic filament. It requires too much force to pull it. I changed the supply system and shifting stopped.

[David Jasiewicz]

So I was away for a little while, but I am still having trouble with parts shifting.  I have spent most of the weekend fine tuning the printer.  All the mechanical parts are as smooth as possible.  It is as close to zero friction as possible.  I am a mechanical engineer, so I have this part down.  My alignment is dead on.  I tried rerouting all wires so nothing crossed.  I twisted the motor wires.  I can print a test square and it comes out beautifully.  I have been printing down to .1mm and the results are great.  The machine runs so quiet you can barely hear it.  I seem to have shifting problems primarily with circles.  The shifting usually occurs in both the x and y axis at the same time.  Also, it is somewhat repeatable.  The shifting will often happen in the same place, but not completely consistently.   I tried turning up the pots up on the motors and the problem just gets worse.  I turned them down so the motors have the minimum voltage needed to run.  I reduced the travel speed down to 60 and it did not help. The problem maintains.  I am out of ideas.  Any thoughts would be great.

Thanks,

~Jaz

[Jay Couture]

So squares work but not circles. You have tested the friction, but what about torque? How much fingertip resistance can the system withstand before it misses steps? To increase torque you need to increase the current limit on the stepper driver. Go too far and the driver will go into thermal shutdown over and over.

• Use a light tough with the screwdriver and do on driver at a time. Turn the current limit pot CCW until it stops. This is the minimum. Current pot is on the right when you do this BTW.
• Not turn it 1 quarter turn CW. You can do this with the power on, but if you tough any two points if the screwdriver, you could short something so use your best judgement.
• Issue a command in Pronterface like G1 X100 F3000 (assumes you home the X axis first). If the axis moves smoothly, then try the finger resistance. Again use you best judgement to tell if the toque is high enough. If not increase the torque and repeat. 
• Now increase the speed to 6000 (100mm/s) and try it. Still move smoothly? And the finger resistance? The faster a stepper moves the lower it's torque. If you want to go faster, then you'll need the max torque you can squeeze.
• Note, this may require active cooling of the stepper drivers (fan blowing across them)

FWIW,

Jay

[David Jasiewicz]

Thanks.  I have tried resetting the motor pots to 1/4 turn.  If I set them much higher, the machine just goes out of control and prints spaghetti.  If I set it lower, there is no change.  Changing the speed does not make any difference.  It does not seem to over heat though.  I had put to large computer fans on the control board, but it did not help the problem.  I am thinking my control board is defective, but I have no idea where to go from here.  Looking at this print, I can see there is a huge shift all at once, which is far more than can be accounted for in slip.  Also, the shift is on both the X and Y axis.  This has to be an electrical problem.  This tells me a movement is probably occurring when it should not, or commands are being missed.  Also, it appears that when a shift occurs, it is more likely to happen on the next layer to.  It can print 50 good layers and then 3 bad ones.

Thanks,

~Jaz

[Jay Couture]

Right, it looks like it missed a lot of steps when it did a large move, a fast travel maybe. The machine has no way of knowing where it is other than by counting the number of pulses it sends to the drivers. If the motors slip, then the machine is lost, and it continues printing as if it's in the right spot.

I'm surprised that there was no change when you turned the current pot up or down. Are you using a RAMPS v1.4 board and Pololu drivers? Can you take a photo pointing to which pots your were turning so I don't assume you knew which ones I meant.

[David Jasiewicz]

Thanks.  I did not say the pots did not do anything, just that it did not solve the problem.  If I turn them up it gets worse.  If I turn them down, nothing changes so long as there is still enough voltage to move the motor.  The fact that it gets worse, tells me that the board is glitching on higher power.  If you set it too high, the movement becomes almost random.  I am not talking so high that it starts to make a clicking sound.  Just a little higher than 1/4 turn.  I purchases the machine used on eBay about 4 months ago, so I do not know anything about the board or drivers.  Maybe the photo I have attached will be helpful.  As a note, I did have to turn up the z axis motor a fair amount for it to be able to lift the build platform up all the way.  Also, I had a heavier build platform at one time and I thought it was the problem, so i reduced the weigh way down and there was no change.  I think i can rule out anything mechanical.

Thanks again,

~jaz

[David Jasiewicz]

[David Jasiewicz]

So I am out of ideas here on this shifting problem.  Maybe I should try upgrading the firm ware.  I have no idea how to figure out what I am currently using, how to update it or where to get an updated version.  Unfortunately, I am just trying to get it working well so I can sell it because I need a dual extrusion machine for my engineering work.  I wish the M2 had dual extrusion.  Anyone who can point me in the right direction, or is interest in the machine, let me know.  Thanks

[Ron Walker]

well. That sure looks like a ramps board. Im thinking it should have an Arduino mega underneath it?. Just by looking at that picture it sure looks wanky to me. The ramps board is sold in a kit form so i think i would unplug everything and remove the Arduino board and motor drivers from the ramps and take a look at all the solder joints on the ramps board. I would be willing to bet that was a kit board not premade at any factory. Check all the solder joints or maybe even reheat them and make sure there solid.

If you pull everything apart and check all the solder joints on the ramps board check also the wires and connectors that plug into the board. They were also put together by someone else... check for broken wires going to the steppers.

The ramps board your looking at is a really good thing. It could be much worse with some chinese boards you could have.
I bet this was a Kit board like the one i built. It most likely was soldered together by a dufus like me lol except mine both turned out to work :-)

double check the guys work (on both sides of the board) and look close at all the solder joints even on all the surface mount mini components yes the guy soldered those on too.

because the board does work. i would look close at the wire connectors that plug into the board and go to the stepper motors. They are very tiny strands and easily broken when there stripped to put the connectors on.

once you figure it out... the good news is that machine or at least the ramps board should be able to run a dual extruder setup like your wanting it to be.

Good luck.
Ron.

P.S. the good news is even an idiot like me can build a ramps board.... the bad news is even an idiot like me could have built yours lol.
 

---

Date: Tue, 2 Oct 2012 13:28:24 -0700
From: da...@trickconcepts.com
To: make...@googlegroups.com
Subject: Re: [MakerGear] Re: Printed parts coming out shifted.

[David Jasiewicz]

Thanks!!  Luckily I am very good at soldering, so I will take a look at it.  Maybe I will use it to make a larger machine from scratch (or borrowed parts from some other similar machine)  How realistic is it to use this board to run some type of amplifier to run larger motors?

[Ron Walker]

Wow.. I have no idea about an amp? but its a good board for a plastic printer the or one of the best i know of.
Glad you know soldering. I am only suggesting a couple week spots to look at.
Ramps/Arduino is all i have and for what i do its working awesome.

Please Google Ramps and look for yourself. It can be bought from ultimachine or i think even makergear but it can be bought as a kit. put together by any dingaling that comes along.

You may have bought a printer from some dingaling on ebay who made a bad board? check the board out..


I have a feeling..... its only a bad wire connection to the stepper motors.

you should look... do it over yourself if you find it questionable.


Ron



 

---

Date: Wed, 3 Oct 2012 00:41:36 -0700

[sto...@gmail.com]

Since Ron Walker suggests to review all the solder joints and they are now suspect - drag a hot soldering iron over all the joints and touch the suspect joints with solder. I am suggesting you to reflow the entire board. Cold joints sometimes can not be seen with out magnification. I use a jeweler's mag to see the tiny stuff. It is all tiny to me now. The SMD stuff I usually get detailed with. Caution with the surface mount comonents - they are tiny and you can easily fry them if you hold a hot iron on them too long. So, just touch and run with the hot iron.

I have a different machine - Makerbot ToM and that is why I haven't been piping in but, on my machine when things shift one of my axes are dragging. That is my machine's clue to lube the rods and bearings. Then check for binding, realign and tighten loose fittings. In your case you might need to loosen a few items and give some physical twist with two hands to reseat or realign. I just finished assemblying a B9 (yes, slight brag.) and it was binding in the Z axis. I loosened a few screws, ran the Z axis into the trouble area then tightened and lubed. Good to go. My trick: I use STP engine oil treatment (any one).

Hang in there. Yours is causing grief but, you are learning more than many. When it works you'll be even more amazed with the thing and readily able to fix it. :-)

[Dean Piper]

You can use any stepper motor drivers you want, there would be no amplifier or anything that could be used here, it will just have to be interfaced from the RAMPS board. You could drive a massive CNC if you wanted, really anything that needed Cartesian style control. So to use bigger motors you would remove the pololu drivers and interface your new stepper driver with the corresponding sockets on the RAMPS boards. You would of course be supplying a  separate power to each individual stepper/driver so it may be more complex. But realistically there is no size limit or current limit other than what you can buy or design.

-Dean

[NCBob]

One other thing to check is the belt tension, either too tight or too loose can cause this to happen as well.

[sto...@gmail.com]

You should have check list. NCBob is Right! The belts! Super important to check. I completely forgot. You will start doing some things without thinking and forget others. Thus, write a checklist for your personal machine operations.

From the results of postings for your problem there have been a lot of great pointers to the root causes. Great experience shared. Hopefully, you'll be able to share your successful solution soon.

----- Reply message -----
From: "NCBob" <ncbo...@gmail.com>
Date: Wed, Oct 3, 2012 11:44 am
Subject: [MakerGear] Re: Printed parts coming out shifted.
To: <make...@googlegroups.com>

[Rick Pollack]

The only issues we've encountered that cause skipping:

1) too much plastic extruded and the nozzle hits it
2) stepper driver pot level
3) belts too tight
4) issue with the slicing program or model

Have you tried switching the x and y axis (swap plugs) - it will just rotate the print.

We'll be announcing a dual extruder for M2 within the next 60 days or so.

Rick

--
Rick Pollack
MakerGear
23632 Mercantile Rd.
Unit D-Rear
Beachwood, OH 44122

[Matt]

Rick: On side note, will the dual extruder be upgradeable for the existing M2. Something we can swap and replace ?

[David Jasiewicz]

Thanks to everyone for all the help.  I am fairly certain i do not have a mechanical problem.  Everything is lubed up with Teflon lube and smooth as silk.  I thing it might be a solder joint.  I don't think it is a motor joint though because I tend to shift in 2 motors at a time.  I will reflow the whole board.

I think I will get another RAMPS board and look into making a large dual extrusion machine.  I am thinking 24" x 24" x 12".  I may run several motors on the Z and X axis to get the power up and keep the motor inertia low.  I will probably have to make the moving parts out of carbon fiber to keep the weight down.  I know I can build the machine, but i get a little out of my comfort zone when I start messing with the electronics, but i think I know enough to figure it out.  Maybe if I figure it out I could offer a kit or work with MakerGear in the future if they take an interest.

[Ron Walker]

I had a similar shifting problem when i built my printer. It ended up being a bad connection where i over crimped a connector piece and it broke the wire almost clear threw. I think it was making contact untell it got just the rite wiggle and jiggle. It was hard to find because all the neat tidy wiring i was at first trying to do.

My comment as to looking closely at the board was more from looking at the picture and seeing what looks to me like the mosfet being bent crooked. The mosfet on my board is much more parallel to the wiring block in front it and it also looks to be missing the heatsink.

i dunno..

---

[David Jasiewicz]

Thanks.  I resoldered every joint on my board last night and it did not help.  It could be an internally cracked resistor or something like that.  Some of these intermediate failures are just never solved.  I decided to just buy a new board.   I should have it in a few day.  We will see if that solves it.

[sto...@gmail.com]

Ack!  That's a bummer

[David Jasiewicz]

Well, I replaced the RAMPS board and no improvement.  Looks like the control board is next.

[Jay Couture]

When you replaced the RAMPS, did it come with new stepper drivers?