GIANT heated beds?

Hi everyone. I'm looking on how I can make or buy, or daisy chain a heated bed. I'm looking to make a printer about 30" by 30"

I'm also thinking that I might have trouble heating that much with a RAMBO board, even with a 24V PSU. Does anyone have experience with this?

At that size forget using DC. A silicone rubber 110V AC heater and drive from a SSR that is controlled by the RAMBo. Check on Aliexpress for good prices.

[www.aliexpress.com]

Edited 1 time(s). Last edit at 06/09/2015 08:00PM by Jason_WI.

This looks perfect!

Though can you go into what a SSR is? Is there any examples online of this setup? Thanks!

SSR: Solid State Relay.

Google it and you will understand how it works. But basically is a electronic switch

As Jason says, for large heated beds the practical option is an AC mains powered bed controlled by an SSR. The company he linked to will make a custom silicone heater to your size, voltage and power specifications at a very good price. I use a 230V 350W 300mm diameter one from them in my delta printer. Choose a heater slightly smaller than the bed to leave yourself a margin (e.g. 15mm) at the edge for fixing screws and bed clips. Total heating power of 5kW per square metre is about right.

For the solid state relay, the Fotek SSR-25DA is inexpensive and should be adequate if your heated bed current is up to 10A, or SSR-40DA if it is higher. Depending on the current, the SSR may need a heatsink.

Using an SSR to control the bed will also get around a design flaw of the RAMBO, which is that the heated bed MOSFET is badly chosen - it can't be relied on to have a low Rds(on) with only 5V gate drive.

If you use an AC mains powered heater, you need to take additional safety precautions - in particular, protective grounding of the bed and any other metal parts in the vicinity. Powering the printer via an RCD would also be a sensible precaution. If the bed is moving, then you must use appropriate highly-flexible cable to connect it, and provide strain relief at both ends of the cables.

PS - I think you are taking a risk using an 8-bit controller board for such a large printer, because the software for such boards typically uses 16-bit variables for most quantities, and these may overflow when the distances get too large. Somebody building a large delta printer recently had a problem that probably has this cause. As you are spending so much money on the printer, it really makes sense to use a 32-bit board. You would also get a much faster USB link, and in most cases, an Ethernet port and a built-in web server. If you insist on using an 8-bit board, try to get confirmation that someone else has used the same firmware to control a large printer if the same configuration.

Edited 1 time(s). Last edit at 06/10/2015 03:42AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I have a question for the OP. Will you always be printing such huge items?

The reason i ask is that for the complexity and size the payoff is not proportional. For example if i were printing tons of smallish parts i would rather go with multiple printers for redundancy and parallel printing.

Or a printer with multiple print heads to print simultaneous duplicates. Curious abot the usage of such a beasty printer

Quote
dc42
As Jason says, for large heated beds the practical option is an AC mains powered bed controlled by an SSR. The company he linked to will make a custom silicone heater to your size, voltage and power specifications at a very good price. I use a 230V 350W 300mm diameter one from them in my delta printer. Choose a heater slightly smaller than the bed to leave yourself a margin (e.g. 15mm) at the edge for fixing screws and bed clips. Total heating power of 5kW per square metre is about right.

For the solid state relay, the Fotek SSR-25DA is inexpensive and should be adequate if your heated bed current is up to 10A, or SSR-40DA if it is higher. Depending on the current, the SSR may need a heatsink.

Using an SSR to control the bed will also get around a design flaw of the RAMBO, which is that the heated bed MOSFET is badly chosen - it can't be relied on to have a low Rds(on) with only 5V gate drive.

If you use an AC mains powered heater, you need to take additional safety precautions - in particular, protective grounding of the bed and any other metal parts in the vicinity. Powering the printer via an RCD would also be a sensible precaution. If the bed is moving, then you must use appropriate highly-flexible cable to connect it, and provide strain relief at both ends of the cables.

PS - I think you are taking a risk using an 8-bit controller board for such a large printer, because the software for such boards typically uses 16-bit variables for most quantities, and these may overflow when the distances get too large. Somebody building a large delta printer recently had a problem that probably has this cause. As you are spending so much money on the printer, it really makes sense to use a 32-bit board. You would also get a much faster USB link, and in most cases, an Ethernet port and a built-in web server. If you insist on using an 8-bit board, try to get confirmation that someone else has used the same firmware to control a large printer if the same configuration.

That is a lot of good information. I will make sure to ground it. Also thank you for including the model. I love guessing at this sort of thing.

Hmm that's interesting. I didn't know that the RAMBO was 8 bit. I was actually thinking about it, and was looking at the 3x and 4x logical boards as I might run a diamond or kraken hotend. Or something of the like. As it's such a beast, more print heads the better, eh :-D

I know the gigabot has a lot of "error detection" I am going to have to look into what they really mean by that.

Quote
thetazzbot
I have a question for the OP. Will you always be printing such huge items?

The reason i ask is that for the complexity and size the payoff is not proportional. For example if i were printing tons of smallish parts i would rather go with multiple printers for redundancy and parallel printing.

Or a printer with multiple print heads to print simultaneous duplicates. Curious abot the usage of such a beasty printer

Haha yeah you see I'm really just sorta a tinkerer. Will I print giant all the time? Probably not. I just like the ability to do so. I have this weird idea that I'll 3d print a iron man costume or the like. Or that I can 3d print coffee tables and such. That would be so much fun! Really though the idea of a mega printer has been festering in my brain for about a year. Came across those actuators, and it all just started happening.

That said I have a design of something I want to make that Yeah I could make on my tiny printer.. but it would be a lot better to just batch print and get it all done in a few days. Granted more risk.. but if I actually start printing giant stuff, you betcha I'll build / get a filabot.

Hmm today I want to 3d print a giant laptop case! And with the mega printer I can do that! Hmm today I want to make a life size robotic arm. I can do that! Hmm my car bumper fell off.. I can fi.. okay maybe not that one..

Edited 2 time(s). Last edit at 06/11/2015 08:19AM by L4nce0.

Yeah it makes sense. So here are some options I've considered for the same purposes.

1. One big heated bed
2. No heated bed. Print with nylon, PLA, or PET (PET is much better than ABS anyway but more pricey) . Even with a heated bed, no gaurantees ABS wont lift, peel, banana, warp, etc. It's more about the enclosure or lack of with ABS that makes it warp. Although I've had parts warp on $3000 PolyPrinter's that are totally enclosed. PS. ABS is not the be-all-end-all of plastic. Many other options don't require heated bed.
3. Use multiple 9" or 12" aluminum plates that are individually heated. This way, with some custom firmware, only the section of the printer that is going to be printed upon needs to be heated, but you still have the option of printing across multiple beds. This sorta implies that the bed is stationary or the entire bed frame is constructed appropriately. This would also require some specialized electronics.

Edited 1 time(s). Last edit at 06/12/2015 11:03AM by thetazzbot.

The 8 bit issue is not correct, if you look into the Marlin firmware you will see that there is copious amounts of double and float declarations. This makes the poor Aurduino really slow but still allows it to support large numbers. On the aurduno the float can support up to 7 digits so it is possible to address up to 1km (1 000 000mm) without decimal points and 10m (10 000.00mm) using two decimal points. Likewise the double can address up to 1km distance as well (no decimal points).

You could skip a bed heater completely by printing on a block/sheet of PIR foam. PIR foam is cheap- about $0.50/sqft ( a 4' x 8' x 1" sheet is available from Home Depot for $15) and can be printed on multiple times. You don't have to level the bed either- just bury the extruder nozzle a mm or so in the foam for the first layer. PIR is flame retardant and doesn't break down at extruder nozzle temperatures.

Stratasys industrial machines (FDM1600 and FDM 2000) printed on polyurethane foam. One of the people at the makerspace has tested the PIR foam in the stratasys machines there and it appears to work well. Stratasys prints inside an oven heated to 70C.

I have printed on PIR foam on my machine and it works great- ABS prints stick almost too well (video here: [vimeo.com] and [vimeo.com]). If you're going to print large ABS objects, you will want to enclose the build chamber and add some heat. I have found that 40-50C is sufficient to prevent delamination. You can pick up a hair dryer from your local thrift store to provide heat for the enclosure.

If you print on PIR foam, you can cut the size of the foam piece to fit the print before you print on it, and stick it to the undercarriage with some carpet tape. That way, if you're printing small stuff most of the time, you don't need to use large pieces of foam, further reducing the cost. Large objects that don't want to come loose can be dealt with by cutting the foam away from them.

Can't transport a 4'x8' sheet? PIR foam cuts easily with a razor knife so you can cut it up at the store and take it home in your Ferrari.

With the foam it is best to print on a raft which adds to the total print time and material use, but it is a small price to pay for the convenience.

Edited 1 time(s). Last edit at 06/18/2015 09:45AM by the_digital_dentist.

Quote
atunguyd
The 8 bit issue is not correct, if you look into the Marlin firmware you will see that there is copious amounts of double and float declarations. This makes the poor Aurduino really slow but still allows it to support large numbers. On the aurduno the float can support up to 7 digits so it is possible to address up to 1km (1 000 000mm) without decimal points and 10m (10 000.00mm) using two decimal points. Likewise the double can address up to 1km distance as well (no decimal points).

Marlin uses fixed point maths for critical calculations because floating point is too slow. Likewise it uses 16 bit variables in most places because 32 big integer maths is slower on an 8 bit machine and the default size is 16 bits. So the important question is whether Marlin is using any 16 bit variables that are OK for normal size printers, but too small for a larger one.

Edited 1 time(s). Last edit at 06/18/2015 10:56AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Well I got the 600mm bed. However in the mean time I have decided that I want to do 1000mm x 1000mm. The fact that would be about big enough to prin an iron man torso, was not a factor. . .

Anyways I got the big sheet in, and I'm trying to decide what to do with it. It cost 25 bucks to get here, if I returned it I would be out a net of 50 bucks. I almost wonder if I do an all mental sheet at 1000mm, then the size difference wouldn't matter, or just try to make it work. Sides at these sizes we are talking a lot of juice. maybe I would be better off making an enclosure, doing a metal bed, then trying to get a bigger heater. What do you all think? I've also thought of daisy chaining a few of these. You know get a 400 by 10000 and a 400 by 600 and make it a square. Then I worry about heat being uneaven. Then I wonder about that even on one mat. Thats what makes me thing metal might help even it out.

What have I got myself into haha. the 3 e3dv6 volcanos will be in soon! Just one more order at the open rail shop and the electronics and I am good to put this together.

Another option would be build the frame 1000 to 1000 and go with a smaller bed for a while.. or not print over those unheated portions.

Edited 2 time(s). Last edit at 06/30/2015 10:41PM by L4nce0.

Ouch found their giant bed
[www.aliexpress.com]

That would be 300 bucks before it got here! (ignoring its too long) At that price point I would think I would be able to make one.

If you have a high temp heated enclosure....do you even need a heated bed. I had a few small ABS parts printed locally on a Stratasys printer and was surprised to see that the print bed plates are made from ABS. No heat bed but the enclosure is like a sealed oven. He said the plates are reusable for quite a number of prints and they only cost about $5 to replace.

Quote
simspeed
If you have a high temp heated enclosure....do you even need a heated bed. I had a few small ABS parts printed locally on a Stratasys printer and was surprised to see that the print bed plates are made from ABS. No heat bed but the enclosure is like a sealed oven. He said the plates are reusable for quite a number of prints and they only cost about $5 to replace.

Thats interesting.. I can see it. I mean, seriously I've been printing a year on a plate that wont go over 55cel. Thats an open air design too.



I did realize something.. That if I make it 1000mm , instead of 600mm .. then it wont fit on most tables. Not to mention a lot of other things.

I'm going to have to think about this. I think I'm going to mesure out some cardboard and just get a feel for what I'm considering. Saying numbers, and throwing a tape measure around is nice and all, but doesn't really give a feel. We aren't talking chump change for these size differences.. Looks like about a 500 - 1k increase in costs to go up a 300mm. (If the bed is 300 + machined metal base.) Not to mention power cost increases. I almost wonder if this is a walk before you run sorta thing. Get it to work at 600mm.. Then scale it up. I'm using open rails so I can easily add length to it later, and swap out the bed to a bigger one. Enclosure is another story.

Actually!! That brings me to a good question. I'm trying to decide What the base is going to be. I don't really think I want to use a 1000mm ^2 pane of tempered glass or a 600mm^2 for that matter. I have emailed a few machine shops, but I don't think most can even machine a sheet that size. I also don't think I'll like the price. I figure if my layer height is .5mm, then I can accept tollerances of about half of that. Which looking at metal online, is sort of a big deal.

I can think of a few solutions.. but I think I'll need to do more research. A metal bed would be best I think. Even if I have to do it in segments..

Edited 1 time(s). Last edit at 07/01/2015 02:07AM by L4nce0.

My aluminum bed is 560 mm by 760 mm for a printer I'm building. I made it out of 8020 v-slot extrusions. I used ni-chrome wire looped inside to make the heater circuit. It heats to 110c using 120v mains power. I built it on a stainless steel kitchen cart. I'm having some bearing shaft parts made to tie the dual Nema 23 Z axis motors together for the Y gantry. I just finished tying the dual Nema 23 X motors together to synchronize that axis. The Y axis uses a single Nema 17 motor. I have it all broken down waiting for the machined parts before the final assembly.

I like the way you think sir!

Though what wire are you using? I've only need nickel chromium wire in tiny tiny pricy strands.

So you essentially have the wire going through the slotted aluminum extrusions. How much wire are you doing per length? What temps are you able to get it to? I would LOVE to see what that looks like with a heat camera.

Been drooling over
[www.flir.com]
IF it had a better range of temps I would buy in a heartbeat.

Quote
simspeed
My aluminum bed is 560 mm by 760 mm for a printer I'm building. I made it out of 8020 v-slot extrusions. I used ni-chrome wire looped inside to make the heater circuit. It heats to 110c using 120v mains power. I built it on a stainless steel kitchen cart. I'm having some bearing shaft parts made to tie the dual Nema 23 Z axis motors together for the Y gantry. I just finished tying the dual Nema 23 X motors together to synchronize that axis. The Y axis uses a single Nema 17 motor. I have it all broken down waiting for the machined parts before the final assembly.

[attachment 57344 Heatbed2.JPG][attachment 57343 Heatbed5.JPG]

Hey btw you got a link to your blog? Surely you are doing something online. Wouldn't mind seeing how it turns out! (also glad to see I'm not the only one crazy enough to build a mega printer at home)

That's FLIR camera is really cool. I'll have to think about getting one of those for my Nexus phone. I used 16ga. wire total length about 450 inches. I bought a big spool online for about $50. Got a Mager SSR for the control relay. I'll run it through a smoothie like the Dentist is doing. No blog yet but I'm planning for one when I'm finished with the build.

Actually I found a better one since my post
[www.thermal.com]
It has a Huge temp range! Not 100c like the first one but 300c+!