Striping on prints

#1

So, I have a question of how to interpret a test I just ran. I print (at least so far) mainly miniature figurines. Since moving my printer outside in its own enclosure, after multiple other issues including a “burnt” mirror, I have been getting good prints unless I print more than one figurine at a time. I would get fine stripes, about every 1/2-1 mm that went all around the printer, but worse on one axis, and it has been taking a lot of extra effort to file every print down, and even with that I am usually not getting all of the striping off (good enough, though).

The odd thing is that if I print a SINGLE figurine at a time, I get very little or sometimes no striping noticeable at all. I thought this was likely my screwing up the galvos which I had to clean and move due to a burn on the mirror.

Then I started to read a few threads about people getting striping due to having a heater installed. SO I made a test print. It has a vertical tube like cylinder surrounded by several cones (hollow), which started 5mm wide and tapered to .5mm wide at the top. The idea was to make each layer print progressively more and more quickly in a predictable manner. My hypothesis was that if there is something turning off and on and this is happening on a fairly regular schedule, that the lines should space out and get progressively wider and wider. That is, indeed, what happened. Below is the picture.

IMG_4234

I am currently heating a cabinet with 2x 100 watt infrared lightbulbs. The bulbs are about 5 inches from the right side of the moai and slightly above. They are powered by a reptile heater controller located on the left side of the cabinet, about 6 inches from the side of the moai. The power between them runs over the top of the cabinet. I have a few questions:

1). Is there another interpretation of the results of this experiment… can someone think of another cause for this striping given the picture above, given the increasing spacing between the lines?

2). If it is indeed my heating setup, WHY would it cause this striping? How could the interference cause stripes?

3). If it is indeed the heater, what is the best procedure to avoid the interference? The machine is outside, and I find that the high temp resin seems to print better at higher temperature than low, and it HAS to have some sort of heater.

4). Is the interference most likely cause by the lightbulbs, the current running through the wire, or from the thermostat?

Thanks very much for your time.

Klaus

0 Likes

#2

Which vat and build plate are you using? Is the cabinet always heated or do you start heating prior to printing?

The pattern reminds me of ringing artefact in FDM printing. It is usually caused by oscillations in the moving steppers due to sudden changes in direction. It may be that what you are experiencing is a form of ringing in the z-axis stepper caused by the stronger adhesion of multiple models. Try reducing the peel speed and adjusting the z-follows.

0 Likes

#3

Same thing happens to me. I don’t think I did much printing without a heater so I have no frame of reference as to what happens without one.

I’m building a tripod designed by @GIDEONkjv for the z-axis to try and address it: Project Tripod. 1.0

If you observe the z-axis during a peel operation, the thing moves all over the place (for large parts with high peel force).

0 Likes

#4

I had a very similar issue and your method of testing is better than mine was. I simply had the printer run for about 75% of a print job one time, then shut it off (figuring the resin would still be warm enough to not cause issues), and the ‘ringing’ was completely gone for the last portion of the print (see below).

image

I have since moved the heater from the lower chamber to the upper (having cut into the acrylic and printed some mounts) I have not had any distortion like this.

image

Also, I would not get bent too out of shape about the ‘why’ in this scenario; the galvos are very sensitive to EMI it seems and it’s enough to basically pull them out of position part of the time (when the heater is running) and then they return to normal afterward. That was my conclusion, and thankfully moving the heater itself (not the power transformer as that stayed mostly where it was) helped the issue.

Where is the position of your heater currently?

0 Likes

#5

This is not a heater issue, this is a Z-axis issue. Project Tripod WILL fix this issue… promise!

Look familiar?? :wink:

0 Likes

#6

Not sure how you would know that without seeing the OP’s setup. I’ve had no issues since moving my heater away from the laser assembly into the upper chamber. We need more input from @kshuler

0 Likes

#7

I have my heater in the upper chamber and thought for the longest time that it was emi issues. If you go read the tripod thread you will see a lot of people with the same issue. Heater interference is generally a slight strip issue, not “wavy” like the op. Yours looks like a very sudden movement so maybe the heater was your issue.

0 Likes

#8

I am using a standard peopoly PDMS vat, using grey hi temp resin (but this happens with other resins, too). The entire chamber is treated and stays heated 24/7.

Its an interesting idea about the z-stepper motor having ringing a ringing artifact. Certainly I note that the more models on the plate the more of this striping effect I get. The thought of it being something to do with the models adhering tighter to the PDMS, so more power is required to peel it, had occurred to me. In fact, if you examine carefully the models I print, there is more of this striping lower in the model, and once you get above the supports, where the models start to taper down, this striping artifact pretty much goes away. That was my initial thought.

But then I started to wonder… if. THAT were the case, then shouldn’t the striping patter be RANDOM. Instead, I have very regular lines, regular as the stripes on the US flag. And after looking closer, they seem to get slightly less frequent the higher up the model you go. I couldn’t explain this regularity by any mechanism that involved random forces, so I tossed this theory out as a possibility.

But perhaps I am misunderstanding the issue… is it possible to get a predictable, even pattern of stripes in the model, where the striping frequency and height is a function of an individual layer’s total surface area?

0 Likes

#9

Wow-- thanks for the info about project tripod. That’s really great and may be worth a try. Nice name, BTW. My current Sorcerer is a servant of Kelemvor.

0 Likes

#10

Your post gives me an idea for the next test. I’ll pull a bulb out of the heater and crank it as high as it will go, so its will be on full time at half strength. If this makes the pattern better, then we have our culprit. Otherwise z-axis, etc has to be looked at.

1 Like

#11

Hi Gideonkjv-

I’m very impressed by your tripod project and will probably build that when I get a chance anyway. I don’t know yet if that is my issue or not, as I never had these striping issues when I had the printer indoors, and now need a powerful heater (200W) to keep the chamber heated as SWMBO didn’t want the smell of resin in the house. So something either mechanically happened to the printer when it was moved, when I messed with the galvos, or this could be caused by the fact that I had to install a strong heater.

I’ll print that test print one more time (But WHAT A PAIN to peel those things off the build plate-- Sheesh!) and see what happens.

1 Like

#12

Here is the setup I have, for people that have asked. The chamber is heated by 2x 100w IR bulbs and the air is circulated with a continuous fan. The area is insulated (but some has pulled off the left wall) and all electrical connections are inside the chamber… there Is a 6 way hub that everything (including the moai) all plug in to. This is another possible cause… when the 200 watt power spike comes on from the heater turning on, might the Moai feel a sudden short dip in power? Here is a picture of the inside of the chamber.

print%20chamber

0 Likes

#13

I think the striping would be random if it was being driven by something turning on and off. Unless, it was on a timer of some sort. Ringing, as I understand it is caused by hysteresis in the steppers which builds up and damps out over distance. I am guessing it is the stepper controller hunting for the proper set point. The difference here is that the Z-axis is not in constant motion.

An alternative would be that something in the system is loading and releasing regularly as tension builds during the print. Maybe look at the linear rails and the coupler to the z-steppper.

To test if this is really adhesion dependent, replace the conical prisms with the cylinder at the center of the build plate. If the issue is adhesion, the pattern should remain consistent all the way up.

0 Likes

#14

OK, the test print is done. Without thew normal temperature control, things did not go well, and the print essentially failed, but enough of it printed to compare… and the stripes appear to be gone.

Here are the prints BEFORE changing the heating setup.
IMG_4233

And now that I had the 1/2 strength constantly on heater:


Even though the print was a failure this time, you can see that the striping is gone. So the question is… what to do about it? You can see my setup in previous posts. How can I shield the moai from the electromagnetic field changes? Our do you think it is more due to power changes from things being on the same circuit? What is the next step? Build a faraday cage over the machine?

Klaus

0 Likes

#15

usually the reason for those lines is caused by the z-axis or the temperature
in your case, you should try installing the heater first, it’s a fast and quick fix
from your set up it’s the temperature OUTSIDE that was heated up, what we need is the temperature INSIDE, using light to heat is not recommended because the resin is sensitive to it and the cover panels of the Moai is designed to absorb light, so the whole set up does not have much affect

0 Likes

#16

Not sure I understand. The light I am using is an IR lamp. The Peopoly resin is NOT sensitive to IR. And besides, as you say, the panels are designed to absorb light, anyway. Surely you don’t think that anything that doesn’t get touched by the light is not heated up… the thermometer that is used is not detecting light, it is detecting temperature. The air heats up inside it, and believe me, when you open the moai and pull the build plate out, it is just was warm as everything else.

Think of it this way… you say the panels absorb light. What do you think happens to that energy? From Newtons laws, we know that you can’t create or destroy energy. So the panels heats up. Then THEY start to radiate its energy. This energy hits the metal portions inside. These are black and also absorb energy. They then begin to radiate, as does the resin itself, which is heated by the radiation, as well as by conduction from the warmed plastic it is sitting in, and a bit by convection from warm air currents. Basically everything starts radiating infrared light, which the resin is not sensitive to, and everything heats to equilibrium. There is also a circulating fan to help blow the warm air around, as well, but frankly I don’t think it is even needed. It is mostly so that nothing directly under the bulbs gets too hot.

When you say I should try installing the heater first… what do you mean, exactly? Yes, I realize I am heating the outside of the moai up, but I also needed to keep my stock of resins from freezing so just heated the entire cabinet. There should be no difference between the inside and outside, or at worst the inside would be a bit warmer than the outside due to heating up of electrical components. No big deal, and shouldn’t be more than a degree or two.

1 Like

#17

Thank you Paradoxical_Cat. I had the same thought… why would the turning off and on be so regular. I can only think that it is a matter of thermal ballast… the thermometer reaches a point that tells the controller to turn off. But I have quite a bit of mass in there, and it will take a while for things to cool down to a point where the thermometer will tell the machine to turn on again. I THINK that it is just reflecting the time it takes for the thermostat to turn on and off. Since he area is not perturbed when I am printing, and I generally don’t open the doors, perhaps it just takes about the same time every cycle to cool off to the point of needing to turn the theaters back on?

I had also thought about the tension building. A while back I took down the machine and rebuilt it. The coupler between the z motor and the build plate had a metal tab that was slightly touching when I reassembled the Moai. I tried to bend it back into place, but it turns out the metal is a bit tougher than I expected, and without fully disassembling the thing, I was unable to bend the piece back to straight. As such, there IS a piece that slightly rubs (or looks like it might, at least). That has always been in the back of my mind, and I had been hoping that wasn’t causing my problems, as I really didn’t want to take apart my machine again. Given the consistency of the striping artifact, through, and how regular it was, I felt this was unlikely the cause. That was one of the things I was trying to test with the initial experiment. If it WAS this part rubbing, and it WAS releasing on a regular pattern, then it should have continued that regular pattern (lets say every 10 layers) independent of how strong the peeling forces were, how long a layer took to print, or how much area the layer covered. So I threw that hypothesis out.

The experiment you propose is confusing me a bit. Essentially, if I understand correctly, you want me to print 5 vertical cylinders instead of the cones. If the issue is the hysteresis in the stepper motors (OR if the issue is the loading and releasing problem in your alternative hypothesis), then the pattern should be consistent all the way up the cylinders. This is true. But if the issue ISN’T either of those, and WAS due to electrical interference, then you would STILL expect the pattern be consistent all the way up. My previous prints of models I actually wanted to print essentially bear this out… the pattern is pretty consistent on them (until you reach the top). The experiment done was trying to get at this very question… is it something that is time dependent, or is it something that is cycle dependent (depending on how many layers have been printed).

The experiment seemed to show that stripe frequency was more of a time dependent variable than a # of layers dependent variable. Of course, time is not necessarily what causes the problem… it could be total layer area, or peeling force related. But I have a hard time explaining either of those other two possiblities as a matter of FREQUENCY of the stripes. Time seems to make the most sense to me, given the data.

0 Likes

#18

I actually arrived at the alternative conclusion based on your print: the spacing between bands was surface-tension based. Simply, the spacing on the bands became wider as try conical prisms narrowed. Printing all cylindrical prisms would account for this as surface contact would remain constant across the print. If the issue is dependent on print time or some other factor, then I would expect the pattern to demonstrate the same spacing increase regardless of objects.

An interesting issue. But, probably not as much for you. Good luck.

0 Likes

#19

So, I will preface this by saying I have NO IDEA of how the forces work in a machine like this, and even if my theory of it being the time that heaters turn on and off is based on the concept of “and then something happens that makes a stripe.” How the stripe forms, I am unsure. All I was testing is the timing of it. I would somewhat disagree with your thought that if I printed only cylinders and the issue is relating to timing, that spacing would increase. If the heater turns on every 15 minutes, and a layer takes 1 minute to print, then whatever is causing the stripes (I’ll call it “the miracle”) will happen every 15 layers, or every 0.6mm (40µm layers). With cylinders, since each layer will take the same time to print, one would expect every 0.6mm there will be a stripe consistently for the entire printing of the model. If, however, I print things that take less and less time to print a layer as times go on (the design purpose of my cones), then “the miracle” will happen still every 15 minutes, but, while the 1st layer will take 1 minute, subsequent layers take less and less time. As such, while the 1st stripe appears after 15 layers, but subsequent 15 minutes periods give the printer time to print more and more layers, thus spacing out the striping the longer the print goes on. If at the end layer only takes 15 seconds to print, “the miracle” will happen every 60 layers, making a stripe every 2.4mm. Whether “the miracle” is an issue with the z axis, the galvos, or whatever, I don’t know, but in this theoretical construct, all I would have to do is eliminate the periodic disruption and things would be better.

If I understand your theory correctly, and I am pretty uneducated on the forces that work in these machines, the pull on the model (and thus, the force the z-motor experiences) may be responsible for the stripes. I am trying to figure out how this would work with my experiment. Mechanically, how would this work? What I imagine is that the pull is too strong for the z-motor, which gets pulled down a little bit from each peel and doesn’t recover fully. As such, the more the surface area of the print, the stronger the pull, and the more the deflection of the z-motor. The total deflection of the z-axis is related to the number of layers and the strength of the pull over time (essentially the integral of the pull forces over time). At SOME point, the deflection gets to the point where the Moai figures out that the z-axis is no longer where it is supposed to be, and then corrects to where it is supposed to be, creating the beginning of a new stripe. If it is not the Moai actively figuring this out, then it might be something getting under tension, then when a certain total deflection (in this scenario due to a buildup of tension) or tension is reached, then it springs back to the default position before any tension was applied. Is this what you imagine is happening? Sorry, I am just trying to understand the mechanics to help me with troubleshooting.

Either way, the results of my last experiment are hard to ignore, as it clearly showed that when the turning on and off of the heater is eliminated, ALL STRIPING WENT AWAY. While I still want to understand what is going on, it clearly has to do with the heater turning on and off, so I want to know what the next step would be. I have a few options:

1). Make a separate hole in my enclosure for the power cable to the heater controller. This would lessen any power spikes or dips from the sudden turning on/off of the 200 watt lamp if I could get it on a different circuit. But then I have to drill more holes and rewire everything… no fun and the hole would be permanent, even if this turns out not to be the case

2). Enclose the entire Moai in aluminum foil, making a makeshift faraday cage. Cheap and easy, but wouldn’t isolate the circuit, AND would decrease the efficiency of the heating setup, as it would somewhat decrease radiative energy transfer to the inside of the chamber.

3). Any other ideas? I like the Tripod Project and will likely pursue that at a later time, but given the last experiment, this does not seem to be my issue at least at this time.

While this issue has been a bit of a pain, it is a minor hiccup compared to previous issues and for the most part, an hour per figurine with a file and some sandpaper has been ALMOST adequate to make the stripes go away, and after things are painted it is hard to notice the stripes unless you are looking specifically for them. I now know at least what system has to be changed, but would like to understand the actual mechanics of the issue. This will help me understand this, and future, issues and be better able to figure out what tests will be best at determining causes.

Thanks to everyone for help. The Moai community has been very helpful and I am honored that people are willing to reach out and help others.

0 Likes

#20

I didn’t see the part about deactivating the heater eliminating striping. That would argue that this is heater dependent.

The Moai is dumb as to the function of the Z-axis. At startup the Z-axis drive moves the trolley to the top, which trips the z-stop sensor and sets zero. The vat distance from zero is set by the user during calibration. The firmware then calculates the number of steps required to get from zero to vat. All other motion is a calculated number of steps without any further feedback from the Z-axis regarding position. This is why the printer must be reset between prints. The stepper motor may provide some feedback to the driver, but it is probably limited.

As for how the striping appears, my guess is that 50-60hz noise throws off the steps in the z-motor or controller, resulting micro-stepping or over stepping until the transient works out. This could result in inaccurate z-resets. Or perhaps it results in mechanical binding in the system. The problem is that most motion artefacts arise from the Z-axis or build plate and the system is pretty tightly constrained.

In either case, since the consistency of the subsequent layer is set by the previous layer, I am guessing that only a couple of bad layers would be necessary to produce a band.

0 Likes