Heatbed issues

Hi,
We're using printers of own design. In the past we did use the regular PCB heatbeds. Because they were very slow when using larger printbeds we decided to switch to silicone-heaters. They are very fast and very good, until now. The last set of silicone heaters do not distribute the warmth evenly over the printsurface AT ALL..!
Now we're thinking of making our own heatbeds with halogen lamps. This way we can install as many lamps we want, but more important : at which location we want under the printsurface.
Someone tried this before..?

Chris

Silicone is a thermal insulator. It will not "distribute" heat. If you don't put some thermal conductor between the heater and the print surface you're unlikely to get even heating from any heater. PCB heaters used without a plate usually warp badly when heated. Sticking a heater to a piece of glass (another thermal insulator) is another bad way to go. The best way to use any bed heater is in conjunction with a massive, thermally conductive plate, usually aluminum. By "massive" I mean thick. A sheet of aluminum foil won't do the job. For most bed sizes, 5-8 mm thick aluminum plate will work. But not just any aluminum plate. "Normal" plate is rolled or extruded and not very flat. Cast tooling plate is best because it comes milled flat on both sides. Mount the plate on three leveling screws, thoughtfully positioned, and it will be easy to tram it to the XY plane of your printer. If your printer is solidly built, it's unlikely that you'll have to retram the plate. Or use a bed sensor on the hot-end, etc.

---

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

Hey thanks for the info TDD,

I knew all this but it does not answer to the original question about the halogen-lamps.

I do build my own(designed) printers and posted about it in another section of this forum. Look for Ulti-Printer.
My printers (now V3) has many non-standard-features that helps giving us superb perimeters. Six have been built so far and used by me and a few friends. Again we're all very happy with the result.
The only weak point in the setup is the heated bed.
V1 (200x200) has a PCB bed. Followed by V2 end V3(200x300) I mounted a siliconen bed (Kenovoo). The 200x300 mm silicone heaters were very good and the heat distribution along the surface was almost perfect. The newest heaters I ordered(300x300) are way thinner, than the previous one(ordered in 2017) and after installing, well lets say 'm not pleased with it at all.

That's why we're brainstorming and looking for a final solution and solve this problem. I know that this uneven heat distribution-problem is one many printers suffers from. The borders of the building surface is always cooler then the center. Not only that but we want the bed to warm up faster then a PCB heated bed.

That is why we're thinking of using halogen lamps to heat the bed.
+20 years ago a was a technician @ Nasuatec. The hotrollers of all machines were heated by halogen lamps. It was used because it was fast and very reliable. Very few lamps were replaced in a copier-lifetime. That's why I'm thinking using them on a printbed.

The question is, has someone thought of it before and what is/was the result.

Chris

Yes its been done before [reprap.org]

And mr google knows of others, you only have to ask him.

It's really a question of radiation vs conducted heat. I would guess that using radiated heat will not be as efficient as conducted heat, and won't be as easily directed to heat just the stuff you want heated. Heating up parts other than the bed can be bad. Heating the bearing blocks, screws, guide rails, and any printed plastic parts may not be a good idea. I believe NASA did some studies using halogen lamps aimed at ABS prints as they were printing so they wouldn't warp or delaminate. They use a Taz printer. I was surprised that all the plastic parts in the machine didn't melt. Maybe they did...

Lamps take up a lot more space than a silicone bed heater, and probably cost quite a bit more. Will the lamps move up and down with the bed?

I would not expect the heating from lamps to be even unless you use them to heat a thick metal plate, but if you're going to do that, why not just do it with a silicone heater?

TLDR, I fail to see any advantage in heating a bed plate with lamps instead of a silicone heater, and can see multiple disadvantages. Warming prints with lamps is another story, but the control mechanism for that might be tricky.

---

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

I have used IR for soldering surface mount components to PCBs, both for local soldering/desoldering and for whole board soldering. I was always impressed by how evenly the heat was distributed. I would be interested to see drawings of your proposed system.

BTW, as you seem to be into making them in multiples, have you thought of making your own aluminium heated build stage with built-in heaters in the form of Nichrome wire in groves. I use this in two of my printers with very good results. The groves were cut 1.25mm deep with a 1.25mm end mill using the 3D printer as an engraving machine. Photo below of the grooves for the plate. The wire is insulated with 1.1mm OD, 0.6mm ID PTFE.



Mike

Great tip there Mike..!
But I would change the concept. I'd use two sheets of thin alu. Between these two sheets I'd glue narrow shims of alu and lay the wire in a loop between these shims. No milling work involved.

Using thin sheets of aluminium may work though I would be a little concerned about possible thermal effects. Having the thinnest possible glue-line should help, it may help to glue the sheets in a bag under vacuum to press them together.
I was quite concerned that the milled grooves on both sides (the top surface has a vacuum hold down) may give thermal problems but I have not seen any temporary or permanent warping after several years of use.

Mike

Why not just glue the insulated wires against a sheet of alu with Katon tape...?
Very good stuff..!

Some calculations showed that the transmission of heat through the PTFE insulation was inadequate if it was only through one surface but O.K. through three. Thinner insulation would probably be good as would be using a thick layer of thermal epoxy. I guess I chose to put the wires in grooves because calculations showed it would be good and because I wanted to try milling on the printer.

Mike

I'd use very thin PTFE tubing.

Does the wires run extra hot at the bends?

---

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

Quote
the_digital_dentist
But not just any aluminum plate. "Normal" plate is rolled or extruded and not very flat. Cast tooling plate is best because it comes milled flat on both sides.

Not only that, but also—even if you mill rolled plate flat to the same precision as cast plate, the rolled plate still tends to warp when heated due to internal stresses, whereas cast plate does not.

Quote
mcdanlj

Quote
the_digital_dentist
But not just any aluminum plate. "Normal" plate is rolled or extruded and not very flat. Cast tooling plate is best because it comes milled flat on both sides.

Not only that, but also—even if you mill rolled plate flat to the same precision as cast plate, the rolled plate still tends to warp when heated due to internal stresses, whereas cast plate does not.

There is a great difference between believing that something won't work and trying it out. I was informed by several people on this forum that 0.25mm thick PTFE over nichrome wire would be too thermally insulating. I did the maths and then tried it out and it was good. If you have tried something and have knowledge that it won't work then say so. If you think that something won't work then say that you think it won't work.

Mike

Sometimes we get lucky and things work OK, until they don't.

In normal operation, the wire in the heater is always much hotter than the bed target temperature. I don't know how you can guarantee that the wire temperature will never exceed the chemical breakdown temperature of the PTFE (or the epoxy). The wire temperature is a function of the power applied and the thermal transfer to the bed plate. It's isn't difficult for something to go wrong. After many thermal cycles the epoxy bond with the metal plate may fail and allow air gaps to form. Air is a great thermal insulator and the wire near one of those gaps will run hot.

Here's a kapton and foil heater that was on one of my printers for years. The brown areas are hot spots where the kapton was toasted at air bubbles between the bed plate and heater.



Here's a silicone heater that developed an air bubble under it when the adhesive started to let go of the bed plate. It toasted the silicone.



I don't know what sort of fumes may come off a burning kapton or silicone heater, but I do know what comes off burning PTFE and it isn't good. See: [en.wikipedia.org]

---

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

Hey guys,

About these fumes I wouldn't worry to much because these fumes only releases from the Teflon parts at 300°C.
Heating the bed to about 75°C, is far from that 'target'. I never use other then PLA.

I also think of using two thermistors in 'cascade'. The first to prevent the wires from heating to much and the second to controle the bed for it's working temperature. There's no need to warm up the wires to say 200 or even 300°C to have the bed warmed up in 10 sec. During startup the hotend heater needs some time to warm up(it's a slow process, so why try to speed up warming the bed..?

Chris

Thank you DD for your informative posting. For information, the temperature of the wires on my heated bed with wires in milled grooves reached 168°C in the bed with a bed temperature of 110°C. With insulated wires in still air and the same mean current as on the bed, the temperature of the wire reached 285°C . The wires in the bed are held in place with a glass tape rated at 250°C continuous duty. Anticipated maximum temperature rise with a controller failed on is 120°C above ambient. These temperatures were obtained by knowing the coefficient of resistance of Nichrome and calculating the temperature rise from the change in resistance.

Returning to the original subject of this thread, I did find some problems with people dismissing what could be a useful way of heating a build stage without knowing any further details. Having said that, if the OP (chriske) gave us more details of his proposed implementation then it may be possible for this forum to give a fuller response on potential problems and possible workarounds.

Mike

Edit: You wait all day to hear more and then the OP replies at about the same time as you click on submit.

Edited 1 time(s). Last edit at 02/16/2020 07:10AM by leadinglights.

Hey Mike,

We're eager to start working on this project. We only started discussing it only last Thursday after we failed these new silicone-heaters to implement in our new set of printers. They actually do work but not to our standards.
We're hoping to find some answer here on this forum how to proceed. That's in fact the main reason of this post.
Some of your posts here has given us already lots of info how to proceed.
Most important change of concept : is dumping the halogen lamps and replacing them with wires. It's a very neat solution and far easier to do.

Thanks again for the input so far.
Chris

Quote
leadinglights

Quote
mcdanlj
Not only that, but also—even if you mill rolled plate flat to the same precision as cast plate, the rolled plate still tends to warp when heated due to internal stresses, whereas cast plate does not.

There is a great difference between believing that something won't work and trying it out. I was informed by several people on this forum that 0.25mm thick PTFE over nichrome wire would be too thermally insulating. I did the maths and then tried it out and it was good. If you have tried something and have knowledge that it won't work then say so. If you think that something won't work then say that you think it won't work.

I'm pretty much lost with respect to what that has to do with what I wrote. I've certainly experienced wrought plate warping when heated, and experienced cast plate staying flat. I'm sorry that you choose to be offended by my failing to express level of personal experience.

What that has to do with PTFE I don't know.

/unsubscribe