Was shipped a convex-shaped heat bed. Same with the replacement

[notarat]

I received a convex-shaped heat bed with my RoMax V2 a couple years ago. (I never got around to building it until last month because I have 4 other printers that are working fine.)

My thought was, "Fine. That's on me for not examining it when it arrived and was in warranty". I'll just buy a replacement.

The replacement arrived today and it's as bad if not worse than the original one.

Anyone have some tips on a method for flattening it that doesn't cause damage but permanently fixes the warpage?

[626Pilot]

By heat bed, do you mean just the black Onyx PCB, or do you mean the glass disc that's clamped to it?

[Mac The Knife]

If it doesn't flatten out when you clamp the glass plate into place, I'm sure a few heat cycles will do the trick.

[notarat]

If it doesn't flatten out when you clamp the glass plate into place, I'm sure a few heat cycles will do the trick.

Well, if by a "few" you mean 15-20 it's not the case.

[626Pilot]

Can you show us a photo?

[ccavanaugh]

Search the old posts, but the convex bed is generally a result of not letting the bed float a bit. The screws holding the bed need to allow it to float just a bit.

The simplest thing to do is to loosen the bed screws, bring the bed up to temperature and then just tighten until they don't move. You are fighting with thermal expansion of the heated bed and it's too constrained, so it bows in the center.

[Qdeathstar]

i'm not sure if it's that big a deal... the glass, which is what you print on is much thicker than the heatbed pcb and will flatten the heatbed. Have you had any issues with printing?

[notarat]

By heat bed, do you mean just the black Onyx PCB, or do you mean the glass disc that's clamped to it?

My bad for not checking my spelling...Sorry about that.

Heat(ed) bed = the Onyx, not the glass build plate.

This seems to be the singular instance where having "ed" is beneficial.

I have some experience with 3D printing so I have already tried the following:

Loosened up all the mounting screws and heat-cycled the original bed 10-20 times to ~65°C (never tried 110°C)

After installing the new Version 7 Onyx I tried heat-cycling to 110c° but it times out before reaching 110°C (so I used 85°C after installing the new replacement Version 7 Onyx) That's using an upgraded Enermax 750w Power Supply, upgraded 16 gauge wiring to the Onyx, and just pre-heating the bed itself...I need to chase this timeout issue after resolving the flatness issue

I then heated it up to 85°C, placed my glass bed from my Creator X onto the Onyx, then stacked 7 dead hard drives on that to weigh it down to flatten out and, while the drives were on it and it was at 85°C, I tightened up all the mounting screws. (I tried this method several times since it seemed to help somewhat.)

That has reduced but not eliminated the convex shape. It's not as bad as it was but I estimate I still have only about 55-60% of the width/depth available.

I have not yet printed anything because, to me, it's better to invest the time up front on calibration and leveling/tramming before wasting filament.

Suggestions on a fix for this are appreciated.

[notarat]

Search the old posts, but the convex bed is generally a result of not letting the bed float a bit. The screws holding the bed need to allow it to float just a bit.

The simplest thing to do is to loosen the bed screws, bring the bed up to temperature and then just tighten until they don't move. You are fighting with thermal expansion of the heated bed and it's too constrained, so it bows in the center.

Yeah...I understand thermal expansion. I already heat cycled it with the screws loose. I even did it with weight on the bed then tightened the screws. I'm at about 55-60% usable bed surface now when calibrating, but can't seem to get any more improvement.

[DeltaCon]

I still do not understand. Are you saying that the Onyx PCB is bulging the boroglass? Can really not be that much... Either the clamps would loosen (add more clamps) or the PCB would bulge back because the glass is much stiffer. Perhaps your glass is bulged instead? If you are referring to calibration different at room temp or at 85c, that is really normal. You should calibrate at the temp you are printing at.

[IMBoring25]

Do you have a reliable reference for flat or is this analysis based on how the effector is tracking across the bed? Have you calibrated horizontal radius?

[notarat]

Do you have a reliable reference for flat or is this analysis based on how the effector is tracking across the bed? Have you calibrated horizontal radius?

Yes. I have a reliable reference for "flat". A 12" metal straight edged ruler that is "flat" as well as two 12" framing squares which are, themselves, flat.

I can stand the ruler on edge and it teeter-totters on the middle of the bed. No. The thermistor is not protruding above the bed. I am placing the ruler next to, not on, the thermistor hole. (and, as I stated earlier, this is with the onyx at 85°C and the glass is clamped down with all 6 clips)

No. I'm not basing it on effector tracking. That's completely different. You know that, right?

No. I do not have wire coiled under the bed causing it to bow/bend/deform/insertwhatevertermyoucaretousehere.

I understand simple concepts like flat convex, concave, and thermal expansion, as well as many other technical, electrical, electronic, and mechanical terms.

People please stop treating me as if I'm a 4-year old. If you don't have a method for fixing the convex-shaped bed, don't respond.

[geneb]

When someone is asking you a reasonable question (at least from their point of view), it's bad form to be a dick about it if you don't like the phrasing.

g.

[IMBoring25]

That was the first time you said here how you were determining the convexity. Otherwise-intelligent people do sometimes miss or misinterpret instructions and then get their heads wrapped so tightly around what they THINK the problem is that they miss other possibilities. Someone who flipped back in the manual to check something and then accidentally missed the horizontal radius calibration step when they flipped back could easily think they had a non-planar bed. It's also an issue that people who deal primarily with newer or upgraded machines with autolevel might not immediately think about.

There was no intention to insult anyone, but it's almost more common than not that people think they've diagnosed their problem and request clarification only to have someone pipe up that they've seen the issue before and it's something else entirely, which winds up saving a lot of tail-chasing.

If you covered those issues somewhere before, sorry for repeating the diagnostics, but you didn't explicitly cover them here and it's really bad form to bite the heads off people who are trying to help.

I would expect there aren't a lot of people with a fix for this, other than attempting the temperature cycles, modifying the attachment, replacing the part again, or compensating with horizontal radius or going to grid-based autocalibration. Radio silence as requested, out...

[DeltaCon]

Pff, attitude...
I often thought my bed was bulged, but in the end it always came down to calibration faults. I am still convinced that the strength of the pcb is by far too little to make the boroglass bulge so much that it becomes unusable. It would simply break before it comes to that. And yes, no Onyx and no glass bed is 100% flat. So, we all are probably to stupid to help out, so I wish you good luck with it.

[Mac The Knife]

And we didn't even address the mdf snowflake yet,

[Xenocrates]

If the PCB is actually domed and not flattened by the glass being clipped to it, an alternative may be to take it pre-installation of the thermistor, and feed a bolt or screw through while it's on the snowflake, and use the 6 un-used holes to clamp it down once it's been drawn flat by the screw, then install it relatively normally. It might begin to bow due to thermal expansion though, since this would definitely over-constrain it.

[Dale Eason]

I could not find anywhere you said what the difference from edge to center z delta is. I could have missed it. But by how much is it not flat. I did see where you said you have not yet tried to print anything with it. Perhaps it is not enough curve to matter. I say that because both beds by your measurements are not what you think they should be. But maybe they are within spec and you are just too picky. Maybe they will work fine.

I never measured mine for flatness. I and most people just do the horizontal radius calibration. It probably takes care of any spherical shape the bed has. We don't ever notice that it may be curved. The horizontal calibration process removes the curvature.


Dale

[nebbian]

I recently thought my bed was bowed slightly, but it turned out to be my metal rule! I was using one of the arms on an angle finder, as this was the perfect size to check the bed. Difficult to check easily. Both metal 'rule' arms on the angle finder had the same non-straightness. Eventually I put it up against another longer steel rule and discovered that the angle finder arms had a bow in the straight edge, at about the same spot relative to their end. It wasn't much of a bow, but it was enough. Try putting all your metal rulers together and see if they are all completely flat relative to each other.

Just another thing to check. Good luck with it.

[626Pilot]

It would be good if you could post a photo of the Onyx. I've ordered several of them over the years and none have been bowed. It's weird that you would get two defective ones in a row. Not impossible, just statistically unlikely. Then again, if you multiply a tiny percentage chance by enough customers, the result a probability value that gets closer and closer to 1.

[Reikal]

The Onyx i have was bowed on delivery. I just assembled with it regardless; given the design of the board and its manufacturing its very likely to be bowed at least a bit. I haven't had an issue with it though, the glass plate is flat and as other have pointed out is much stronger than the Onyx is.