Welcome to the SeeMeCNC Community

Here is something I discovered on my new Rostock Max v2: Z-ribbing or horizontal banding in fairly regular intervals. It gets worse with the speed turned up and is also better or worse depending on which slicer is used.

After extensive testing and many theories, I discovered that the problem went away when I turned off the heated bed.

I presented these results too SeeMeCNC but they were not very receptive to my final conclusions. They dispute that the heated bed could be the cause of z-ribbing.
I would recommend others to run some tests to confirm whether you have this issue on your build, and if you do, please let them know about it.

Try running a small tall rectangular box at your default settings with and without the heated bed on during the print. Taller than the photos is better.
I was using PLA and a 60c bed but I also saw z-ribbing when I ran ABS at a higher bed temperature.

You need to shine a bright light from above to really see and photograph the z-ribbing.

AFTER MUCH INVESTIGATION, THE PROBLEM WAS SOLVED BY JJ AND THE CREW AT SEEMECNC. SEE THIS LINK FOR THE FULL EXPLANATION:
http://repables.com/r/726

I always have the bed turned on, I have never seen this before in my prints.

Which PSU do you have?

I also have a little of this in my prints, have only been printing ABS with high heat.

After reading this thread, I wonder if this can be fixed with an UPS with filtering, or is it down to aPSU with little filtering after the switcing action ?

I resolved it by adding a dedicated PSU on the heat bed rail. Now the banding is gone and the bed heats faster to higher temps and is way more stable.
I used a Mean Well NES-350-24 dialed back to 19v. No excess heat on the mosfet on my rig but use at your own risk and discretion.

http://www.amazon.com/gp/product/B00KTJ ... ge_o00_s00

Nice work. I'm really surprised with the slicer results. I can't imagine how the slicer has an impact with heater control once the print is running since that is done in firmware until another M140 or M190 is issued. The heater is completely controlled in firmware so I could see different firmwares having an impact like this.

Did yuo print your MC, Cura and S3D objects with the bed turned off to do a 1-1 comparison test with the photo above? That would be interesting. You could heat the bed up to print temperature to get the first layers to stick and then manually turn it off after a few layers to take the bed heater control out of the loop. I'd love to see the results of that test!

I suspect you've got stepper motor wires running in parallel with the bed power wires - that would explain the issue that you see and nobody else does.

I'd also try measuring the voltage of the bed during it's constant-on heating phase to see what the voltage output is - make sure your hot end is heating as well.

g.

Take a look at this thread,
http://forum.seemecnc.com/viewtopic.php ... 232#p86232

Here is a video I took showing the phenomenon. [youtube]http://www.youtube.com/watch?v=Jyan7uc3Ww4[/youtube]

PID loop needs to be tuned very well on a bed such that the power applied to the bed doesn't fluctuate wildly. at 100C my bed power sits at about 40%(24v), on a 12v bed it sits at about 90% at 90C, before I manually tuned the PID it would fluctuate wildly.

How many mm of variabilty did you have? Could read the dials very well.

mhackney wrote:Nice work. I'm really surprised with the slicer results. I can't imagine how the slicer has an impact with heater control once the print is running since that is done in firmware until another M140 or M190 is issued. The heater is completely controlled in firmware so I could see different firmwares having an impact like this.

Did yuo print your MC, Cura and S3D objects with the bed turned off to do a 1-1 comparison test with the photo above? That would be interesting. You could heat the bed up to print temperature to get the first layers to stick and then manually turn it off after a few layers to take the bed heater control out of the loop. I'd love to see the results of that test!

I was working with SeeMeCNC on this so we wanted to narrow it down to a single slicer. We used MC since and use the program defaults to eliminate variables.
The third photo compares the MC slicer both with the heat bed running during the entire print vs. heating the bed up to print temperature to get the first layers to stick and then turning it off after the first layer to take the bed heater control out of the loop.

geneb wrote:I suspect you've got stepper motor wires running in parallel with the bed power wires - that would explain the issue that you see and nobody else does.

I'd also try measuring the voltage of the bed during it's constant-on heating phase to see what the voltage output is - make sure your hot end is heating as well.

g.

Running the wires in parallel is how the assembly guide instructs it to be done so I would assume everyone else has them wired in parallel as well if they followed the instructions.

That's not exactly what he meant. I believe that he referred to the 4 wire sets which go to the stepper motors from the Rambo being too close to the wires which go to the bed from the rambo, rather than the wires which provide power from the PSU to the rambo (which will always be in parallel as the PSU doesn't have multiple rails)

Xenocrates wrote:That's not exactly what he meant. I believe that he referred to the 4 wire sets which go to the stepper motors from the Rambo being too close to the wires which go to the bed from the rambo, rather than the wires which provide power from the PSU to the rambo (which will always be in parallel as the PSU doesn't have multiple rails)

I'm not sure how close proximity those wires are, but to me proximity of wires means something completely different than running in parallel. I ran the wires according to the assembly guide and tried to keep them as neat as possible considering the tight space you have to work in.

For me, the addition of a second PSU solved 2 problems at once: one, the z-ribbing while the heat bed is one, and two, the fact that the stock PSU really struggles to get the bet up anywhere near 110c for running ABS. It heats up much faster now and stays there stably.

Are both power supplies putting out the same voltage?

Also, -10 points for telling the guy that wrote the manual, what the manual says.

g.

geneb wrote:Are both power supplies putting out the same voltage?

Also, -10 points for telling the guy that wrote the manual, what the manual says.

g.

In my earlier post, I stated:

"I used a Mean Well NES-350-24 dialed back to 19v."

-10 points for asking a question that was answered already.

Dial it back to 12v and re-test. I'd be curious to see what your results are....Hufflepuff. *laughs uproariously*



g.

SOLVED!
JJ and the rest of the crew at SeeMeCNC did an amazing job tracking down this issue.
The issue is real, it doesn't happen in all circumstances and may depend on the environment of the machine. It is confined to machines with the 12VDC 29A power supplies they began shipping a few months back.

To summarize, when there is enough power draw for the PSU fan to kick in, that fan blows directly through a hole under the bed and onto the thermistor. This immediately drops the thermistor temperature, causing it to request more heat, and the result is an amplified cycling of the heat bed on and off as the PSU fan cycles on and off in a fairly regular pattern. The resulting cycling causes expansion and contraction of the bed which results in the z-ribbing I encountered. This only happens when the heated bed is used.

Barring adding a second PSU as I did, the simple fix is just to tape over that hole under the bed. They also devised a plug for the job. See JJ's full report and download the plug from here:

http://repables.com/r/726

Congratulations to SeeMeCNC support for figuring that one out.
That had to be a tough dog to diagnose and fix.

Too funny, I spotted the bed temps dropping when the PSU fan came on and reasoned that it must be an airflow coming through the hole under the bed. I taped up the two holes last night. With any luck I've headed that one off at the pass.

I had to do the same on my V1 after I added a fan blowing on my Rambo board to cool the stepper drivers down. The problem was however not intermittent, and did not cause any printing artifacts... just that my bed temp readings were not realistic.