Why does my print have tiny zebra stripes?

[DrR1pper 4]

 

So it does have our attention.

 

Great.  Thanks for the update.  

I wonder if there are other printers that suffer from this or is it just the Ultimaker?  As I had mentioned before my Wanhao Duplicator does'nt.  It might be interesting to ask the Creative Tools folks if they have seen the same artifact elsewhere.  If they did you might have a shorter list of possible combinations anyway.  

In any case, good luck!

 

Sorry to resurrect an old thread but i recently purchased a Wanhao Duplicator 5S mini and have this exact same issue as you can see here:

http://i597.photobucket.com/albums/tt54/DrR1pper/IMG_1923_zps9tk6gvys.jpg

The only OTHER common denominator that i can think of between my 5S mini and your UM is the bowden design.

I will be asking Wanhao Duplicator 6 owners if they have the same issue or not with the benchy.

[gr5 2,224]

What kind of stepper drivers does the wanhao dup 6 use? I'm pretty sure we concluded above that the problem is how the stepper drivers are hooked up (the various resister values and so on).

[valcrow 145]

I think Chrii has a hack to correct this. We chat on chat about it sometimes. pinging @chri

[MartijnvG 16]

didn't he fry his last board?

[neotko 1,417]

I'm like john snow since I know nothing but..

Wouldn't it help to use GT2 40tooth 25.5mm dia on the 8mm shafts and ofc a bit longer short belts. This would allow to change the microstepping to 1/8 without losing x/y resolution. Ofc on um2 that could be more problematic since the gt2 it's double sided, but to avoid the space restrictions the gt2 could be one like this, and the belt from motor to pulley could just be a 4mm wide.

[martin-bienz 36]

I'm like john snow since I know nothing but..

 

You know nuthing John Snow.... but hey... SPOILER ALERT, I hope you are well, as he is not sooooo extremly well at the moment... just saying.

[DrR1pper 4]

What kind of stepper drivers does the wanhao dup 6 use?  I'm pretty sure we concluded above that the problem is how the stepper drivers are hooked up (the various resister values and so on).

 

I think the D6 maybe using the exact same motors. Will double check with D6 owners.

Whilst i wait for a reply back from D6 owners on what motor it uses and whether they also suffer zebra strips on a benchy print, this is the D5S mini stepper motor:

http://s597.photobucket.com/user/DrR1pper/media/IMG_1927_zpsebg5epfr.jpg

Edited April 29, 2016 by Guest

[gr5 2,224]

I think the D6 maybe using the exact same motors.

I think it's the stepper *driver* and not the motor that causes this. There's pages of math that comes with the driver that explains how *not* to get these and it involves picking the right resistors (and capacitors) for your particular application depending on all kinds of properties of your power supply and the stepper motors and possibly inductance of the motor and so on. If you don't get it just right you get uneven steps (smaller than desired steps for 15 microsteps and then larger than desired for 1 microstep, repeat). I read it all and still didn't understand it fully - it really needs 20 hours of careful reading/study and I don't think it can be explained in a few minutes. Trying to picture how offset AC sine waves passing through 2 different coils at slightly different phases even gets the motor to move a sub step baffles me already let alone this detail of why it substeps too far occasionally... I don't quite get the issue in the first place - just that it exists in the industry and that there is a way to ameliorate it mostly (but maybe never fully?).

Really these printers should all be using servos and not steppers.

[DrR1pper 4]

Right, gotcha, thanks!

 

Really these printers should all be using servos and not steppers.

 

+ 100

I wonder who will be the first to mass produce servo based fdm printers.

[DrR1pper 4]

Does this mean the driver has been poorly made/configured/paired with the motors used on the UM2/2+ and other printers (e.g. including my d5s mini) also experiencing this issue if there are some printers out there that don't experience this issue at all?

Is the UM2+ still worth it despite such issues?

I've been contemplating on buying a duplicator 6 and now thinking about a um2+ but can't quite decide. I know what some must be thinking....not even in the same league. Ofc it's a lot of you get what you pay for but i'm also thinking in terms of price to performance ratio which is better for me.

Edited May 1, 2016 by Guest

[gr5 2,224]

Is the UM2+ still worth it despite such issues?

Oh yes. Of course. But we are biased here. The quality is the best - I've been to many shows and I think the UM printers have the best quality of all fdm printers. But is print quality really the most important thing?

The driver/motor pairing issue - there are also resistors and other parts that are involved in this marriage. Some combination is a bit off on some (all um2?) printers. But it's really subtle and you need obscure prints like a wall that is not square to any axis but *almost* square to 2 axes to get this. This is very rare in any print. But Benchy has this rare situation. None of my thousands of other prints have this situation I believe.

[DrR1pper 4]

Thanks. Helpful response.

[DrR1pper 4]

I think the UM printers have the best quality of all fdm printers.

 

Quick question please. What temperature do you print PLA at on the um2/2+? Would i be right in saying that sufficient heating of the PLA just as it leaves the extruder is as important as how rapidly it cools once it leaves the extruder? Would i also be correct in saying that this is why/where the UM2/2+'s shines the most, in it's cooling fan arrangement/design?

I've just realised how important sufficient heating of the PLA and rapid cooling of the pla just as it has been laid down is.

[gr5 2,224]

It depends what you care about. If you want beautiful prints then here's the advantages of UMO,UM2,UM2+

Light weight head allows very very high acceleration and high "jerk". UM printers can do 9000 mm^2/sec (default to 5000) and most printers with feeder on the head do 1000mm^2/sec. Why does this matter? It's very difficult to speed up and slow down the flow of plastic coming out of the head. So everytime you slow down (for example) to print a corner on a cube it overextrudes the corner then as it accelerates out it underextrudes the wall because there is a delay between feeding and the pressure that accumulates in the nozzle and when the filament comes out. The delay is maybe only 20-1000 milliseconds but it's a problem. With high "jerk" (UMO through UM2+ use 20mm/sec) the print head never completely slows down on a corner. "jerk" in marlin is defined as instantaneous speed changes in the XY servos (true jerk is totally different - the derivative of acceleration).

Fans. For PLA (not higher glass temp plastics like nylon or ABS but for pla) fans are super critical - you want to cool it very fast. UM2 is much better than UMO regarding fans and the quality is slightly better. UM2+ supposedly is even better but I can't tell the difference (haven't done any tests on bridging or overhangs with my UM2+ versus UM2). Many people have printed there own UMO fan shroud getting the quality just as good as UM2 or UM2+. UMO is still an excellent printer. UM just keeps improving it every year.

As far as your original question - printing temp - please realize that the temp at the tip of the nozzle is different than where the temp sensor is so different manufacturers will tell you to print at different temps which can vary 30C. 210C is a good typical printing temp for all UM printers but people print PLA on UM printers from 180C to 240C. At 180C it's like toothpaste and lays down like wet cement - very good quality - stays where you put it. At 240C it flows like honey (you can print 6x faster) but the PLA doesn't stay exactly where you put it.

[DrR1pper 4]

Once again, thank you for the super helpful response gn5! Really appreciate it!

[Torgeir 259]

Hi Folks,

This is one of the more interesting topics I've been reading here so far -for me...

As I have exact the same zebra stripes and some more strange details appearing on this boat, I'll think this is a good thing to elaborate on further.

I'll think GR5 is right on the track here, -thank you so much!

I have some print to study, my own and some of the famous boat made on an UMO. But the 3Dboat is hard to print properly by any printer for sure...

The best reference is the 3Dboat printed by the UMO (Ultimaker Original printer).

I've got the zebra stripes as well, even with my home built printer. When I looked at the examples I've also noticed some other "error" that got my interest more than the zebra stripes. If you look at the bow of the ship frame, the low part there, close to the heath bed has sign of overheated print in that place.

When I looked to the reference I have, the UMO -it's just perfect, and I could not see any zebra stripes at first glance.

I'd really become interested in this, so I used some magnifiers in order to study the surface of the print. An important factor here was the source lighting when studying the surface of the print.

This first print of the 3Dboat, is made with Verbatim gray PLA using my printer, nozzle temp 204 deg C., Bed 60 deg C., infill 20 % and 0.2 mm layer tichness.

You can clearly see the stripes bot on the steering house and the ship frame. The other side is a mirror of "this" right side.

Next one is the UMO, do not know any parameter, except that the extruder temperature is 208 and bed is around 60 (deg C.).

Next one is the same as above, but tis time a close up of the steering house. Can you see the zebra stripes?

Another two pictures of the UMO printed 3Dboat.

Next, a picture of the 3Dboat printed with my printer with Verbatim PLA, 0.06 mm height, nozzle 204 deg C., bed 60 deg c., 60 % infill and double wall tichness.

I have some more pictures of this phenomene, the X/Y step marks you can see on your 3D print, well -on some prints.

Btw., you can observe the zebra stripes on all of the prints above.

The other interesting print to se is the extruder flow test, the 3D cylinder test.

Thanks  

Torgeir

[Torgeir 259]

Hi,

Here's to print of the "famous" 3D cylinder test. This first one is done with a brittle and water contained PLA. This occurs over time when PLA is stored unprotected in open air, as the PLA subtract water directly from the humidity in the air.

In fact, when I made this first print, I've had to use one hand to add some extra force to the feeder all the time when this print was made.

Here is another picture of the same withe cylinder, but this one show the print just after printing the feeding flow.

Here you can see the little lines going upwards where the stepper change direction I'll think, as this match the actual step to step points along the circle. As the accelleration is constant (ideally) around this circle we should think that the extrudance was even all around. However, since the two steppers accellerate/decellerate from point to point all the time this will have influence on the extrude flow and even make this lines more visible. Ok., the funny thing is if you try to find the center of gravity for this cylinder, it will be right in the middle.

This next one is printed with a proper PLA, right from a sealed bag. It just come out like this. I made two in a row with same result.

Enjoy.

Thanks

Torgeir

[Torgeir 259]

Hi Folks,

As it's almost a year since this tread started, I'll like to ask; has anyone tried out / checked if there is one micro step missing? Meaning one of the micro steps (of 16) is "two steps long" instead of one, leaving visible lines on our prints under certain conditions.

This should be easy to check, -and hopefully easy correct hardware wise as well. At the moment I'm far away from my printer, so can't do this check.

Thanks

Torgeir

[Torgeir 259]

Hi Folks,

I did this "shortening of the ROSC (reference oscillator) adjusting resistor" modification, in order to avoid missing steps caused by lack of back EMF.  

This was suggested by “nope head” at the “hydraptor” web site here:

http://hydraraptor.blogspot.no/2012/04/stepstuck.html

I did this modification on all five stepper drivers, X, Y, Z, Extruder1 and Extruder2.

As I found this Airbus (A380 right?) made by Valcrow, I could not resist this “class one” object for everyone to print, thanks Valcrow.  

Anyway, yesterday I printed the tail fin of this plane as this is a typical place to see these phenomena.

And yes, our zebra stripes are very easy to see here on my print before this modification.

But not on the little piece just under, but knocked off.

Ok. after the modification there is no Zebra stripes anymore…

The stripes visible here is just the infill, it was knocked of the bed again... But the surface is nice. Yes, the left one is after modification.

However, this modification requires quite some skill with the soldering iron, and it’s not for anyone to do. (The “biggest” soldering iron for this job must have a tip of max diameter of 2 mm.)

I have a few pictures of this modification, if of any interest.

Thanks

Torgeir

Edited May 25, 2016 by Guest

[gr5 2,224]

This is fascinating. I sent referred your post to the hardware engineer who designed the PCB for the UM2.

[gr5 2,224]

Can you post the schematic changes? I'm more interested electronically in what you did than physically how to do it myself.

[Torgeir 259]

Hi gr5,

Thanks. It was a real change for sure...

I noticed a little increase in the sound while micro stepping, you will hear "this singing" but not as loud as other 3D printers. (I may say; a version 2 singing.)

The mod itself is:

X-stepper driver U8 R84 shorted. (High ref resistor R81 is originally not installed).

Y-stepper driver U10 R85 shorted. (High ref resistor R82 is originally not installed).

Z-stepper driver U11 R86 shorted. (High ref resistor R83 is originally not installed).

Extrude-stepper driver U12 R102 shorted. (High ref resistor R100 is originally not installed).

Extra-stepper driver U13 R103 shorted. (High ref resistor R101 is originally not installed).

And here is the actual wiring (Board Rev: 2.1.1. wiring page 4):

The resistors that are shorted is colored yellow.

Thanks

Torgeir

 

[neotko 1,417]

I have a few pictures of this modification, if of any interest.

Thanks

Torgeir

 

Please share them. I will like to try that on a china board. So if I fail no biggie.

[Torgeir 259]

Hi neotko,

Thanks for the flowers.

Well, here is two pictures and this first one is the "X_Y_Z_Stepper_driver.jpg".

The second is "Extrude_&_Extra_Stepper_driver.jpg"

I used two cords of (tinned copper) 0.1mm diameter and twisted. I started on the side of resistor close to the driver IC, then after soldering the first point made a loop toward the other side on the top of the resistor, parings here and soldered.

This last picture is named "the_third_hand.jpg"

Hope this will do.

Thanks.

Torgeir.

[SandervG 1,521]

Hi Torgeir,

Thank you for sharing this.

What is the home build machine you printed the boat on that also had the Zebra stripes?

Wouldn't it be weird if both machines, from different manufacturers show the same occurrence?

Could you explain to me how a micro stepper could miss steps?

Is it an overheating thing, or just an inaccuracy of the motor?

What exactly did your fix in the electronics do to avoid this?

Sorry, just trying to understand what you did