so what causes this?

[printbus]

Do you see ringing on the inside of the box as well? That's what I would expect if it had something to do with the extruder.

On the holes and notches in thin sidewalls, yes it is on both the inside and the outside. Whether it is on a corner, a hole, a notch, outside or inside, it's always on the right hand side looking at an upright surface. At least on my small test print, I know that's the exit direction of movement.

In other words, it's not that the same ripple is showing up on both sides of an extrusion. I should point out that with at least the test print I'm currently using for this, "thin wall" is still fairly thick - it's 1.75mm. I'm seeing Simplify3D use three passes for this 1.75mm wall.

[gmay3]

From my experience, I'd say you know the belts are tight enough when you tighten each belt by pulling on the motor as hard as you can with one hand and use the other for tightening the screws to keep it in place. I wouldn't use tools or clamps since they can apply exponentially much more force than you'd expect and you can't get a good feel for how tight things are compared to your hand.

[DanCope]

Just to add some more data to this. I am experiencing a similar issue to OP. I found this nice test object by ocrinus (http://www.thingiverse.com/thing:277394) which will show up both X and Y axis ripple. The X and Y letters are aligned to their respective axis.

I am printing with a Prusa i3v, 0.1 layer height. First print is with default settings.



Second print is with Jerk lowered to 15 (from 20) for the bottom half and then Jerk 10, plus AMaxX/Y down to 500 on the top half.



It is more noticeable in person, although the lower speeds do make it less visible. My guess would be that it is caused by rocking in the motors when they stop? It always ripples in the direction of movement, as printbus says above. It also looks fairly consistent between both axes so I wouldn't have thought it had much to do with give in the framing or vibration to the print head (but hey, what do I know ).

I am in the process of tightening the belts some more to see if that helps at all.

[TopJimmyCooks]

I would love to know what effect those astrosyn motor dampers would have on this.

[printbus]

DanCope - welcome to the board.

As time permits, I've also been running a pile of prints, changing one parameter at a time. I'm not complete, but I've seen some results similar to yours. I dropped xyjerk all the way down to 2. While there was some improvement in the ripple, it certainly didn't make it go away as I expected it to.

One observation on your settngs - you mention lowering Amax X and Amax Y to 500 mm/sec per second. I assume you are referring to what were the MakerFarm default values of 1000 either on the LCD or in DEFAULT_MAX_ACCELERATION? Note that there is another acceleration term Accel on the LCD and DEFAULT_ACCELERATION in configuration.h, which has a MakerFarm default of 500. As I understand it, the acceleration one gets is actually the lowest of either the axis specific acceleration or the global acceleration setting. With Accel set to 500, changing Amax X and Amax Y from 1000 to 500 actually had no effect since Marlin was already using that as the upper limit.

I'm far from being complete in my test prints, but the most effective scheme I've seen so far to eliminate the ripple is to reduce the acceleration (either the axis specific or the global term) down to 200 or below. I don't believe this eliminates the ripple - it may just compress it down to a shorter distance so that it isn't as apparent.

EDIT: Note that I am not suggesting people go lower their acceleration settings to 200 - I'm just sharing what I've observed. Slowing acceleration and lowering jerk can hurt print quality since the nozzle is moving a lot slower through direction changes like corners, leading to softening/blobbing in those areas.

[DanCope]

One observation on your settngs - you mention lowering Amax X and Amax Y to 500 mm/sec per second. I assume you are referring to what were the MakerFarm default values of 1000 either on the LCD or in DEFAULT_MAX_ACCELERATION? Note that there is another acceleration term Accel on the LCD and DEFAULT_ACCELERATION in configuration.h, which has a MakerFarm default of 500. As I understand it, the acceleration one gets is actually the lowest of either the axis specific acceleration or the global acceleration setting. With Accel set to 500, changing Amax X and Amax Y from 1000 to 500 actually had no effect since Marlin was already using that as the upper limit.

Yes sorry I should have clarified, all settings where done through the LCD and defaults were as you described. I thought it odd that Accel was lower than the specific Amax X and Amax Y, but thought I would drop them both anyway. Ignore that setting change then.

Is there a setting to control how the printer decelerates towards an edge? Lowering the Jerk and the Accel are only going to go so far to hide the vibration when I imagine the real cause is when a motor suddenly halts on an axis - surely you would have to lower the overall speed to make it less of a sudden stop?

I would love to know what effect those astrosyn motor dampers would have on this.

I would love to have some of these, but I would be surprised if general motor vibrations were causing this. Unless it's like a resonance thing? Not sure.

[truly_bent]

I've been following this thread with some interest and have been doing a little digging on the web. The general opinion seems to be that it's a mechanical problem and, as has been mentioned here, one that can be somewhat improved by slowing down travel speeds and tweaking the accel/jerk parameters.

It also seems to me that this problem is specific to Cartesian coordinate systems. I haven't yet found any complaints about "ringing" on delta type machines. Can anyone confirm that delta machines are immune to this ringing problem?

[printbus]

Is there a setting to control how the printer decelerates towards an edge?

The acceleration settings apply to both the acceleration and deceleration phases of a movement.

[printbus]

I've been following this thread with some interest and have been doing a little digging on the web. The general opinion seems to be that it's a mechanical problem and, as has been mentioned here, one that can be somewhat improved by slowing down travel speeds and tweaking the accel/jerk parameters.

It also seems to me that this problem is specific to Cartesian coordinate systems. I haven't yet found any complaints about "ringing" on delta type machines. Can anyone confirm that delta machines are immune to this ringing problem?

I continue to have questions on whether this issue is strictly mechanical.

Assuming it is primarily a mechanical problem, the extent of the rippling will be driven by the amount of weight being manipulated during printing. True, deltas are typically moving less weight, but someone could possibly have a poor design for a delta with a printhead that is excessively heavy. If so, that might still exhibit some of the same issues. EDIT: Thinking about it, that printer might have other print quality issues than the rippling... I really don't know.

[DanCope]

I've been following this thread with some interest and have been doing a little digging on the web. The general opinion seems to be that it's a mechanical problem and, as has been mentioned here, one that can be somewhat improved by slowing down travel speeds and tweaking the accel/jerk parameters.

It also seems to me that this problem is specific to Cartesian coordinate systems. I haven't yet found any complaints about "ringing" on delta type machines. Can anyone confirm that delta machines are immune to this ringing problem?

I think you are right, I have also found more information about stepper motors in general now that I had a better idea of what to look for. The 'Vibration Problem' section on this page seems to describe the problem we have. It kind of seems it's just a limitation of these motors.

I imagine delta printers don't show this issue because of the way it drives the motors. It is probably less likely to bring a motor to a sudden stop and any vibration would be absorbed differently into a print than straight across an axis.

The acceleration settings apply to both the acceleration and deceleration phases of a movement.

That makes sense. I would guess there is a lower limit to how slow it could go before it stops the motor completely so even decelerating slower might not help in terms of a sudden stop.