Heated bed always makes thermal bands on prints.

I have prusa i3 clone called Tronxy P802MA from aliexpress. Made of acryl. [www.aliexpress.com] It has a MK2A hot bed. I coupled X and Z axis the first thing by installed anto Z wobble thing After that verticals became very streight. No wave. I print only PLA.

1) After X and Z were decoupled vertical walls became smooth but every N lays i get a layer sticking out in all direction, but that happened only when i use 2 or more shells. If i print in spiral mode (1 shell) everything is fine.

2) by accident (saw my son playing with spiral mode test print near a window) I discovered that even spiral mode printed models have problems. when look through light i see banding.

When 2 or more shells are used this band add up and eventually layer sticks out.

3) checked extruder T - it is within 2 degrees

4) checked bed temperature. When T near 78C (it is just a number, real temperature measured on top of bed is around 62-65) i see darker band in print. I did PID calibration. Did not help. If print cylinder with double radius the height of bands is 2x less and frequency 2x more. So, it is either bed or extruder or power supply.

5) I turned off the heater and printed on cold bed. No banding. So, it is bed or power supply.


6) Checked voltage when bed heater and extruder heater are on. 11.84. When just bed heater is on then 1.89. No problem with PSU. Something is wrong with bed.

7) Released all springs on the table. Maybe mechanical problem when heated table or heater expands. Frequency of bands has changed, but they are there.

8) Printed without heater on again. No banding.

9) Removed heater completely but did not disconnect it. So, it was still heating while i printed on a cold bed. No banding!!! So, it is 100% not electrical problem and related only to temperature fluctuations.

10) Put the heater back, did some thermal insulation. Did PID calibration. Did test print again. Better, but problem is still there.

11) Removed all springs. Used only nuts. Reaseembled the bed. Still banding when heater is on.

So, basically, i cannot use heated bed at all now. I can only turn it on for the first layer and keep below 30 for the rest. Any idea how to solve the problem?

This is typical PCB heater behavior. The problem is that when it heats up, it expands and it bows. The other problem is that the whole bed assembly is flexible, so when the heater bows, so does the rest of the bed structure. Most people put a piece of glass on it to try to flatten it (which trades an unflatness problem for an uneven heating problem). If you already have glass on it, either go thicker or learn to live with the problem and avoid that type of heater in your next printer. I'd avoid anything with an acrylic frame, too- acrylic is the next step up from laser cut plywood- neither should have ever been used in 3D printers.

You can fix the problem completely, but by the time you get done you will have rebuilt most of the printer. The fix involves replacing the bed with cast aluminum tooling plate, probably replacing the heater, which means probably replacing the power supply and possibly replacing the controller. I'm not sure it worth all that in an acrylic frame printer- there will be probably plenty of other problems, too.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Aha! I suspected i stumbled upon something inherently systematic for such type of printers.
I do not use glass because i don't understand how it might help. If it is not firmly coupled on top of aluminum bed then it will wiggle up and down and change inclination when bed deforms.
As i see it the best way is to decouple somehow the heater from the bed, so, when the heater bows it does not push anything. Maybe bigger holes for screws and not tighten the heater and the bed together will help. Need to think about.
Also,what about MK3 heated bed? Does it have the same problem?

Here's an example of doing it right: [www.instructables.com]

In that case, a broken glass bed was replaced by aluminum using the original heater. Bed heaters are fundamental components in a 3D printer and trying to replace one often creates more problems that lead to one upgrade after another to accommodate the original upgrade. Parts for those cheapo acrylic printers are chosen on the basis of cost and checking off feature lists (heated bed, check), without regard to performance.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

He writes
"he high temperature silicone is used to attach the heater to the new bed plate."
I can do the same, remove holes for screws from the heater and use silicone to attach it to the bed plate. I doubt that the heater can bend aluminum plate, but it can push the screws like it does now.
If i can find softer kind of silicon than it can even compensate for bending.

I thought that silicon does not conduct heat very good but not i checked the numbers and it is pretty good.
149 W·m−1·K−1 vs 237 aluminum vs <1 glass

I always puzzled me how people use glass as a bed surface. I was against my knowledge.
Now i am puzzled about those hot end silicon socks which are supported to be heat insulators.

basically, the best insulator is air Just need to trap it and make it not escape.

I was the author of that instructable.

The silicone is applied as a very thin adhesive layer. It isn't there for any other purpose. The reason cast tooling plate is works is because it is thick, rigid, flat, and thermally conductive. Gluing the heater to a thin piece of aluminum probably won't prevent the whole thing from warping when it heats up. Four point "leveling" isn't helping, either.

Putting glass on top of an unflat heater or thin aluminum heat spreader traps air between the hot surface and the glass in those places where it doesn't touch. That's why you get a flatter surface to print on by clamping glass to the heater, but also why you get uneven heating of the glass. You trade one problem for another, both of which can cause problems for bed adhesion.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Aha, it was you

Your plate is 1/4" roughly 6mm
My current aluminum plate is about 5mm.
Not much different, isn;t it?
Not sure if it is cast or not. Does it matter?

Cast plate is milled flat on both sides. Most aluminum sheet is extruded/rolled and is usually not very flat. You can check it by placing the edge of a steel ruler on it as see if you can see a gap or slide a piece of paper under it.

This is a 12" wide piece of 1/4" aluminum that I bought four years ago, before I knew about tooling plate. It was completely unacceptable:



Cast tooling plate usually has a shiny surface with milling marks- large circular arcs - visible in the surface:



Edited 1 time(s). Last edit at 02/01/2017 12:00PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Your knowledge is extremely valuable for me now.
Turns out nothing in 3d printing can be taken by default. Everything need to checked, understood and rechecked.
First, my plate is not 5mm, it is puny 3mm. Secondly it is really NOT flat. In Y axis it is bulging in the middle, in X the left and rights edges are higher then center. Crazy form. The difference i estimate between 0.1-0.2mm. Lucky i always print first layer 0.3 to even out. But this is sad.
Well, now i know what want to my bed, but this is for future. This is my first printer, so, i will try to get maximum from it and they most likely build corexy myself.
Thank you.

Problem solved. Or the cause is found, real solution is to be found yet.
I took heater from screws, turned a little, Used some stuff to hold the heater stuck to the bottom of the bed.







No banding, perfect 1 shell cylinder, perfect 2 shell cylinder.
Very stable bed temperature because when heater was bending its central part was going down from the bed. Not it is stuck to bed.