Strange ripple on only x axis

Ive been trying to dial in my prints to acceptable standards but i keep getting these strange ripples along the x-axis , gives parts an almost wood grain look as if it was cut with the grain going vertically through the part. Any idea what is causing this ??



as you can see the ripples are along only the x axis.
i know i have pretty bad z wobble but thats due to cheap threaded rod, will upgrade later.

Try to change layer height to be non-periodic to the z-leadscrew pitch.
That way you can be sure it's not the z-wobble that's spoiling your game.

The things I upgraded first on my Prusa were adjustable belt tensioners, maybe an option for you too?
-Olaf

changing z height didnt change anything so i figured maybe it was loose belts and after tightening them up to the point that i can use my printer as an instrument the issue still persists :\

Im now thinking that it may be due to my extrusion , as im printing the outer perimeter last and its probably oozing out in an S shape causing these ripples.
So i figure i have to reduce flowrate for perimeters so the hot plastic is stretched more, and maybe increase perimeter speed.

feedback is much appreciated.

Edited 1 time(s). Last edit at 01/06/2016 12:57AM by Rooski.

Rooski,
It looks like it may be a form of resonance.

- Do you know what kind of drivers you have? DRV8825 or a4988 drivers?
- Can you place the part back on the bed in the direction it was printed and provide a photo of the whole printer?
- Do other parts do the same thing on that side (say if you printed a cube)?

---

My printer: Raptosaur - Large Format Delta - [www.paulwanamaker.wordpress.com]
Can you answer questions about Calibration, Printing issues, Mechanics? Write it up and improve the Wiki!

Resonance would not repeat itself in the same XY place though; it would be variable.

There is something, at certain repeatable Y positions, which is always causing X to be moved to the right a little bit, and for just a blip. And it's the same Y positions no matter the Z height.

First thing to try is to cut the speed in half. If the ripples stay at the same Y positions, then something mechanical might be pushing the bed left (causing the nozzle to move right) as it moves. If half speed causes the ripples to double or halve their positions/frequency, then it's probably something in the drivers/motors.

Also try a double-sized part. And rotating the part 45 degrees.

what does your Printer look like?
and what is the Distance between the Ripples? is it possibly the same as the spacing of the thooths in the Belt? (X Direction only ofc)


PS new to 3d Printing myself, in fact this is my first Post here ^^ so i may not have the answer, but i think these Informations could help here

Try wiggling the extruder/carriage and the build plate. If either will move it is liable to be such movement caused by sloppy bearings or flexing guide rails/printer frame that causes the ripple. It doesn't have to move much to have an effect on the print. That pattern occurs where X (or is it Y?) is accelerating/decelerating. What is the microstepping ratio? How fast are you printing and how much acceleration are you using? What is the power supply voltage for the motors? How much motor current are you using vs what are they rated for?

Does the ripple pattern appear in a flat surface like a cube or only a curved surface?

Edited 1 time(s). Last edit at 01/12/2016 07:21AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Add diodes, its work!

That "fix" (http://cabristor.blogspot.ca/2015/02/drv8825-missing-steps.html) is specific to DRV8825 driver chips set to a specific decay mode. It is not a "general" solution to microstepping problems for every stepper driver.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

actually it turned out that my driver was dying , i replaced it with a cheap chinese DRV8825 and it works great now.

well great meaning that it doesn't produce that noticeable wobble , i still seem to be getting some kind of regular pattern on both axes, it seems to correspond to the motor vibrations when powered. it feels completely smooth but i can see it running vertically through the part, any ideas ?

Also ordered some 8mm lead-screws so really looking forward to virtually no z wobble.

What sort of extruder are you using, with what driver? On my direct drive mk8 clone, using a drv8825 driver I got a very noticeable pulsing in the extruded filament that produced ripples running up the print (with varying slant depending on the relative extrusion vs linear movement rates). In my case that was solved with the inline diode fix mentioned above.

i was using a wades style extruder with a a4988 driver

I just recently finished putting on the new leadscrews and switched all drivers to drv8825's , but havent had time to print anything yet.

... there can be some "XY-wobbling" caused by the geometry of the tooth-belt - if you turn the motor pulley, the tension in the belt changes with every tooth/gap, because it's not perfectly 'round' when mating the grooves on the pulleys ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

could someone explain the diode mod? what are the specs that will influence the number of diodes? I'm on 24V with 42shd0217-24b motors and as suspected drv8825 driver.

See #26 on the Shifted Layers / Offset Layers / Missed Steps page.
There is a link on the right side to the article.

---

My printer: Raptosaur - Large Format Delta - [www.paulwanamaker.wordpress.com]
Can you answer questions about Calibration, Printing issues, Mechanics? Write it up and improve the Wiki!

This link has chapter and verse: [cabristor.blogspot.ca]

To simplify - a design choice in the implementation of the DRV8825 driver means that it can't regulate the current for micro-steps that require voltages in the bottom 12% of the supply voltage. Instead of smoothly ramping currents you get a step function followed by a plateau which translates to non-uniform motor movement. If you use a low impedance motor with a high supply voltage you will be likely to see the problem. You have this case in spades, so you probably get a pretty good show, depending on printer type and where you use the motors. High score is on a non-geared extruder like the mk8 style. Your motors are 2.5 Ohm, 1.5A so if you run them flat out they only need 3.75V. 12% of your 24V supply is 2.88V so three quarters of your microsteps would be hit by the problem.

The diiodes shift the operating region away from the 0V cross-over to get the motors out of the 12% area. So for an ideal case you need to shift the voltage-current curve by 2.88V with series diodes rated for at least 1.5A. 1N5404 will do the job easily, IN4001 are a bit on the tight side (but I used them for a while without any signs of trouble). To clear 2.88V you'll need three in series, and you need two sets connected back to back, so that's

-|>--|>--|>-
-<|--<|--<|-

with the ends connected together. You need two sets like that per motor, one inline with each of the coils. And if you're thinking that 12 huge 1N5404 diodes per motor is a bit daunting, then yes I'd tend to agree. You could try the smaller 1n4001 diodes and only using 2 in the series (8 per motor) and see how that works.

thank you very much! that were the info I needed!

Should some A4988 be better in this case? Will they have the same dacy problem or will they be different?

1N5400 should do the trick as well, am I right?
Will this hack work with A4988 as well?

As far as I know, the A4988 doesn't have this problem, so you should get smoother movement out of the microstepping. The down side is you lose the 1/32 micro steps and higher current handling capacity.

The 5400 has a much lower reverse voltage capability, only 35V. Given you're running 24V and motors produce large voltage spikes when you remove power, I'd probably stick to the 5404 that is rated for 280V. They're dirt cheap, so it's not like cost is an issue.

Edited 1 time(s). Last edit at 03/23/2016 07:12AM by JamesK.

I tried some A4988 yesterday and the problem is still there... seems only a little bit smoother. Even the decay mod on both the A4899 and DRV8825 wont help.
Mechanical seems to be fine, I'll excahange the PU belt to a rubber one in case this is the problem.

Edited 1 time(s). Last edit at 03/23/2016 07:11AM by Kev0.

Ah, that's a shame. Must be some other cause then - let us know when you find it!

[forums.reprap.org]

it is so damn periodical...
I first tought it would be the smootg idler pulley but after changint it to a teethed one its still the same.

Are the 5400 are rated @ 50V and the 5404 are not available now at my local store :-(

[www.produktinfo.conrad.com]

Edited 1 time(s). Last edit at 03/23/2016 07:23AM by Kev0.

What does it look like on a single wall cube? Nothing against that model, but it's hard to tell what's going on with complex shapes.

If the 5400 are easily available, give them a go and see, chances are it will be ok. The extruder is the most likely cause of that sort of problem unless you are using a delta?

edit: Oh, and a similar affect can be caused by the extruder gears being off-center, so check that too.

Edited 1 time(s). Last edit at 03/23/2016 08:16AM by JamesK.

Made the diode mod.... no changes. Must be something mechanical.

If you say of center, do you mean exentrical on extruder shaft or not aligned with the PTFE?

btw tge diodes reach easy 60-70°C... normal?

Edited 2 time(s). Last edit at 03/23/2016 03:18PM by Kev0.

Quote
JamesK
This link has chapter and verse: [cabristor.blogspot.ca]
[...]

It works because it increases impedance of the motor which reduces current build-up. I think the explanation provided in the link is not correct. The driver does a closed-loop control scheme but not a common mean-current-PWM.

To hold a specific step current the driver applies the full voltage to the motor, waits a blanking time, checks the current and turns it off if it has reached the target current. The full voltage will be applied after a specific time and checked again. See the fast ripple here. The problem is: If during the blanking time the current rises too fast which occurs if the impedance is too low or the supply voltage is too high. The result is seen here. The current rises above the limit which results in a wrong driving current shown here. Typical steppers are rated for around 5V. Using 24V increases the current too fast during the blanking time.

The solution is to: Reduce supply voltage, reduce effective voltage, increase resistance, increase inductance or activate fast decay mode. 3 Diodes may not be enough to reduce the voltage. Ideally, the supply voltage should be tunable to match the impedance of the motor. Unfortunately, RAMPS uses the same voltage (12V or 24V) for the steppers and the heatbed/hotend.

Quote
Kev0
Made the diode mod.... no changes. Must be something mechanical.

Shame, but consistent with not getting an improvement with A4988 either.

Quote

If you say of center, do you mean exentrical on extruder shaft or not aligned with the PTFE?

btw tge diodes reach easy 60-70°C... normal?

Eccentric on the shaft, it causes changes in the effective radius of the gear. That sounds hotter than I would have expected for the diodes, but I haven't checked mine to see what they run at.

If its been the extruder gear the effect would be related to the extruded distance and be constant on this, regardless of the direction. As seen its not related since the effekt doesn't take place on straight lines.

May my motor inductance is to low? Which kind of motors are working @ 24V with the hack? Would give it a try and exchange them.

Edited 1 time(s). Last edit at 03/24/2016 04:05AM by Kev0.

Quote
Kev0
As seen its not related since the effekt doesn't take place on straight lines.

Hey Kev, did you post a photo of a straight line print. or does that literally not have anything to see? It's quite difficult to tell from your torso print exactly which problem we're talking about. Perhaps you could zoom in and mark up exactly what you're looking at? There are some quantization steps in the print that look like they are probably in the original model/ Heh, I'm trying to avoid the temptation to down load it and try a print, not sure it would go well with t'missus.

I mean the "annual rings" that look linke layers in X and Y direction

Quote
the_digital_dentist
Try wiggling the extruder/carriage and the build plate. If either will move it is liable to be such movement caused by sloppy bearings or flexing guide rails/printer frame that causes the ripple. It doesn't have to move much to have an effect on the print. That pattern occurs where X (or is it Y?) is accelerating/decelerating. What is the microstepping ratio? How fast are you printing and how much acceleration are you using? What is the power supply voltage for the motors? How much motor current are you using vs what are they rated for?

Does the ripple pattern appear in a flat surface like a cube or only a curved surface?

How much movement on the carriage is acceptable? Does it have to be rock-solid. While the rods and bearings are metal, there are a lot of plastic parts carrying them and plastic is more flexible.