Rostock MAX METAL - First Build (Updates: Ugly Boat Machine)

[geneb]

Is the power supply under the Onyx? Is its fan aimed up? If so, I bet those bands correspond to the fan cycling.

g.

[IMBoring25]

Do you have enough cable to move the PSU away from the bed? On the stock Max, that symptom occurred when they switched to the LED PSU because the PSU cooling fan was blowing on the bed thermistor and causing the bed to fluctuate wildly when it cycled.

[bradjshannon]

Onyx and Duet are on separate power supplies located about 18 inches away (laterally) from the bed.

[bradjshannon]

[img]http://ldhk.org/img/File_000_1CDA9210.jpg[/img]

[bradjshannon]

Any suggestions for fixing this perimeter issue?

It's hard to describe. But everywhere a given model's geometry changes, there's a shift in the perimeter/shells of the print.

[img]https://i.imgur.com/DMhly1w.jpg[/img]

direct link to img

[Xenocrates]

I think it may be for the banding, that the Onyx shifts as it gets that much power through it. What is your stack like? Is it just Onyx-Glass-coating? There doesn't seem to be a heat-spreader in there (One like Brian over at trick laser's would do nicely), so I suspect you haven't got a whole lot of thermal mass, thus why the bed would hit so regularly, and you don't have additional stiffness to damp the shifting like an aluminum disk would provide.

As far as the perimeter thing goes, I think it may be a geometry issue in the STL. Try running it through Netfabb cloud first, and then hold the two side by side and see if it helps. It could also be that the difference in size makes a notable line as the print speed shifts (If you have a high minimum layer time and max speed, with it being allowed to go slower to compensate for large parts). Try taking a look at it in layer view, with show speed one.

[Polygonhell]

The sifting at the start and end of holes is a relatively common issue, it's usually a sign of some slight backlash somewhere. The perimeter is consistent where the path is consistently loaded, as the path changes (introduction of loops) the backlash is exposed.
Tracking it down can be a lot of not fun, it's usually very slight.
The general inconsistency of the walls in Z on that part would be more concerning to me. I'd be back to printing test pieces (simple cubes and cylinders) to try and identify the cause of that, it could be the same thing or something unrelated.

[nebbian]

You could try getting your heat bed to use PID and PWM control, instead of "Bang bang". I'm not sure on the specifics of your setup, but it seems to me that this would solve the issue. If you're using a relay then you might consider using an SSR or high current rated FET for your bed.

If you're already using an SSR or FET, then try lowering the time that the controller sits there doing nothing between temperature samples. You might be using "Bang Bang every x seconds", try changing this to just plain "Bang Bang" or reduce the number of seconds to 1.

[bob64]

Make sure the screws holding your bed isn't coming loose.

[mhackney]

Very odd. I use bang-bang for my bed powered by Duet (both .6 and .8.5 versions) with a 24v MeanWell supply (and on 2 printers I only have this one supply, on my max I have a second 12V supply for the hot end, motors, etc). I am using the Auberins SSRs that I've written about on all my machines. Your problem might be related to the SSR. My Onyxes are rev 2 and 5. But the MeanWell should have enough umph to power these.

[mhackney]

I just read the last post on the previous page - shy do you believe the onyx expansion is causing that banding? No-one has ever seen an effect like that due to onyx expansion. After thinking through this a bit more, I really suspect your SSR if it is the Cydom D1D40 SSR mentioned in your signature. Also, do you now power the Onyx from 24V? The other thing to verify is that the connections are secure. Do you have a volt/ammeter? Can you measure the voltage drop on the onyx-side of the SSR and then measure the current as you are printing and post the results?

[miglo]

I had similar banding when I was using bang-bang on my smoothieboard. Switching over to PWM resolved my problem.

[3D-Print]

Banding issue caused 100% by Onyx heat bed thermal expansion. Solution unknown.

I do not have very good feelings about this product at the moment. If it's going to cause banding like that while simply holding at 50C, I don't see how it's even going to be usable.

I have seen this type of banding when the temp of the bed fluctuates by even a few degrees. There is a a thread of jfettig detailing with a micrometer the changes in bed hight as the it heats and cools....... it shows impressive variations.

I achieved a very stable bed temperature by adjusting the PID settings such that the bed only varies by 0.1-0.2 degrees C (http://forum.seemecnc.com/viewtopic.php ... 125#p88209). The SSR which runs my bed, flickers to keep it at the selected/desired temp.

A tight enclosure to avoid drafts will also help to stabilize changed in bed height.

[bradjshannon]

Thanks for the suggestions, all!

I've been printing for the last few days on blue tape or PVA glue on top of cold glass. Needing a heated bed for PLA is kind of a crutch, but obviously critical for other materials, which I'd like to play with at some point.

First I'm going to try PID and/or PWM control of the heat bed.

Next, I'm probably going to try getting an aluminum or copper heat spreader, which will add to the thermal mass and possibly help.

As for the non-heat-bed banding issues, a bit of belt tightening helped a lot. Here's a bad photo of the surface finish of my latest print. There were just 3 banding flaws on the entire 18 hour print. Printed at 0.16mm layer height @ 30mm/s:
[img]http://ldhk.org/img/IMG_4259_JPG_1CE134D6.png[/img]

Yes, I'm using this fancy newfangled machinery to print... let's go with "water pipes."

[bradjshannon]

Are these PID values completely nonsensical? They seem to work, but my testing is limited.

Code: Select all

M301 H0 P10 I0.4 D30 T0.30 S1

[3D-Print]

Glad it works!! When I did my adjusting, it was trial and error from the baseline using the RepRap details (http://reprap.org/wiki/G-code#M301:_Set_PID_parameters).

Here is a link to what I did if you haven't seen it yet (http://forum.seemecnc.com/viewtopic.php ... 200#p87861).

It was trial and error with manual adjustments adjusting as noted until it was quick to heat, didn't overshoot and was stable once at temp. I started with "M301 H0 P20 I0.5 D100 T0.4 S1 W180 B30" I did remove the W180 B30.

My final bed setting for the Duet and my heat chamber was: M301 H0 P24 I0.80 D105 T0.50 S1.0

For manual adjustments I used the following
If it overshoots a lot and oscillates, either the integral gain needs to be increased or all gains should be reduced
Too much overshoot? Increase D, decrease P.
Response too damped? Increase P.
Ramps up quickly to a value below target temperature (0-160 fast) and then slows down as it approaches target (160-170 slow, 170-180 really slow, etc) temperature? Try increasing the I constant.

[bradjshannon]

Glad it works!! When I did my adjusting, it was trial and error from the baseline using the RepRap details (http://reprap.org/wiki/G-code#M301:_Set_PID_parameters).

Here is a link to what I did if you haven't seen it yet (http://forum.seemecnc.com/viewtopic.php ... 200#p87861).

It was trial and error with manual adjustments adjusting as noted until it was quick to heat, didn't overshoot and was stable once at temp. I started with "M301 H0 P20 I0.5 D100 T0.4 S1 W180 B30" I did remove the W180 B30.

My final bed setting for the Duet and my heat chamber was: M301 H0 P24 I0.80 D105 T0.50 S1.0

For manual adjustments I used the following
If it overshoots a lot and oscillates, either the integral gain needs to be increased or all gains should be reduced
Too much overshoot? Increase D, decrease P.
Response too damped? Increase P.
Ramps up quickly to a value below target temperature (0-160 fast) and then slows down as it approaches target (160-170 slow, 170-180 really slow, etc) temperature? Try increasing the I constant.

Thanks, I've been using your thread as a reference this whole time . Same with that wiki site.

I wish there was info about the min/max values for each M301 parameter, or what scale they each use, or whatever. I feel completely in the dark, not knowing the range of reasonable values.

Currently plugging away at Z-N. Spent all day reading on this crap! Argh! Hurry up and oscillate, damn thing!

[bradjshannon]

Z-N results failed miserably. Using no-overshoot method, got massive overshoot (78C with a setpoint of 70C).

Trying this and still getting significant overshoot:

M301 H0 P3 I0.82 D345 T0.50 S0.4 B15 W200

This reaches up to 73C with a setpoint of 70C.

[DeltaCon]

Highjacking this thread because of the mentioning of the heatspreader. I hear people having it machined out of a thick alu sheet, because it seems extruded alu sheets have too much of a thickness variety. Is that also going to be a problem if you put the glass on top of it? Would a standard anodized alu construction sheet do? How thick is it advised?

[bradjshannon]

Highjacking this thread because of the mentioning of the heatspreader. I hear people having it machined out of a thick alu sheet, because it seems extruded alu sheets have too much of a thickness variety. Is that also going to be a problem if you put the glass on top of it? Would a standard anodized alu construction sheet do? How thick is it advised?

I didn't know I had these questions until you asked them. Thanks!

I'm on the fence about a heat spreader. I mostly want one to cover up the SeeMe logo on the onyx. Sorry, SeeMe folks! It's just... distracting. And I always have to explain its appearance to people.

[Xenocrates]

I have an aluminum sheet heatspreader (Three actually). I haven't bothered machining it down to really precise tolerances. The one I used with glass, there was never and problems with it causing imbalance, and I did some large prints with it (Both side to side and front to back on the bed, but not at the same time). With PEI on a thicker one, I felt the need to smooth it out somewhat, largely with a disk sander. Again, few to no leaving issues. I did buy the relatively high precision stock though (MIC-6 is even better, but many stocks are ±.006 IN, which is pretty good tolerances for the bed)

[DeltaCon]

I'm on the fence about a heat spreader. I mostly want one to cover up the SeeMe logo on the onyx. Sorry, SeeMe folks! It's just... distracting. And I always have to explain its appearance to people.

That's a rather sad reason for it!

I did buy the relatively high precision stock though (MIC-6 is even better, but many stocks are ±.006 IN, which is pretty good tolerances for the bed)

Thanks, with tolerances like that I am not too worried about it. I don't think that a disc sander in my hands would make that any better
What thickness do you use? Maybe I will get one extra to try PEI some time, though I am content with the glass also.

[Xenocrates]

I ended up with both .125 and .25 aluminum, although I only ordered the .125. Worked out well, as there was enough over the nominal width on the .25 to cover the bed pretty much perfectly.

[3D-Print]

I used some copper sheeting and cut it to fit (http://forum.seemecnc.com/viewtopic.php ... 303#p59898).

It has worked well, however, I have never measured if there is any or how much variation in temp across the print bed.

[bvandiepenbos]

Highjacking this thread because of the mentioning of the heatspreader. I hear people having it machined out of a thick alu sheet, because it seems extruded alu sheets have too much of a thickness variety. Is that also going to be a problem if you put the glass on top of it? Would a standard anodized alu construction sheet do? How thick is it advised?

I found that .06" thick aluminum is sufficient mass to spread out the heat yet not take to long to heat up.
That thickness also stays flat when clamped between onyx and boro glass, it is not so stiff to cause problems if aluminum is bowed.
I have them water-jet cut from 5052 alloy aluminum sheet.
The 5052 is softer and bends easier than say 6061 alloy, I think that allows it to lay flat.