Heated Bed Design Concept

[Misha]

Here is my idea for the heated bed, would love some of your feedback.  Based on Ezra's design.

It is 3mm boro, with 3mm aluminum plate with the kapton heater attached underneath the aluminum.  The whole assembly sits on a sheet of G10 which is attached to the slide.  I am modeling a larger than standard table (290x290) with the standard TL Kapton heater (~200x250) attached for my sized up MM.

I have run some heating simulations and at least according to the computer the aluminum plate distributes the heat very well despite not being covered completely by the heater. The extra aluminum plate adds around 200 g of weight.

[Jim Morris]

Looks interesting, how do you level it?

[Ezra Zygmuntowicz]

Misha-

Here is the first prototype of my idea for using an aluminum heat spreader and sticking the kapton heater to that instead of the glass. This allows the aluminum to spread the heat all the way to the edges of the glass and it also makes it so the glass does not bow when current runs in the kapton heater as the kapton isnt stuck directly to glass so if it expands it doesnt cause the non expanding glass to bow. 

But I didn't want to loose the beautiful look of the kapton heater under the glass by putting the kapton under the aluminum so I cut a 1" x 2" relief hole in the center of a thin 1mm thick aluminum plate same size as the glass. I stuck the kapton heater to the top of the aluminum carefully removing the thermistor first and then pushing the red power wires through the tiniy 5mm hole in the center of the kapton heater so they go down through the hole in the aluminum.

Then i slide the aluminum and glass together into the tslot slit and use two small bulldog clamps on each side to make sure the aluminum and glass have good contact. I will p[robably add some fiberwool insulation under the aluminum to keep the heat in and to push the alu up against the glass more .

But I really like this solution as it fixes the known issue yet still allows the awesome look of the kapton heater under the glass.

Hewre is a photo gallery of me assembling this new prototype so you can follow along to see what I am talking about >>

Aluminum Heat Spreader Kapton Heater Doesnt Bend Boro Print Bed

I will be extensively testing this setup over the next few days and will report back with my results. It seems very very promising so far and repeatedly heating and cooling the heat bed has not caused the glass to bow a single time yet.

Also one more thing to note that I just discovered. These kapton heaters have a built in thermistor, the tip of the thermistor is completely covered with a PTFE tub. THis is bad and wrong. THe PTFE tubing should only cover the wires leading up to the thermistor tip but the thermistor tip should not be covered by PTFE. I carefully uses an xacto knife to remove the PTFE just from the thermistor tip and I placed the thermistor on one side of the print bed in the middle of the Y axis length but 2mm in from the left side of the x axis *on top* of the glass and covered it with kapton tape. This way I will be measuring the temperature of the top surface of the glass and not the temperature of the center of the heater.

THe PTFE covering the thermistor tip I think has been causing a lot of these issues as when pronterface reports my bed is 100C my IR thermometer reports that my bed is 179C!!!

Remove the PTFE from the tip of your thermistor for much better temp readings and I will report back in a day or two after I print a ton of stuff with this new bed and let you know how it fares. I am very hopeful this is a good solution.

Thanks

-Ezra

[Misha]

Ideally this should not need leveling, if built and installed very carefully, but if becomes a problem I will add something between the G10 and the extrusions.

[Misha]

Looks great Ezra,

I will try your arrangement and run some heating simulations.  I am only worried that since my bed will be considerably larger than the heater the glass would heat too unevenly if it is touching the heater directly.

But I like the idea of using thinner aluminum and adding some insulation underneath, I'll run that simulation as well.

[verohomie]

Misha -

Nice work.  I like this as it uses existing parts.

Do you plan on using shims for leveling?

Brad

[Caanon]

Looks good Misha! Exactly what I plan to use. 

I think having the kapton heater underneath the aluminum is the way to go, since ChrisUK has done some experiments where the glass directly above the tracts was heated correctly but in between the copper traces there were cool spots.  The aluminum acts as a good heat spreader.

In the other thread, Alex brought up an interesting point about the extrusions on the end acting as a heatsink for the aluminum.  Was this factored into your calculations?  The simplest solution - should the extrusions cause too much cooling - would be to use binder or bulldog clips instead.  Would also marginally reduce weight, though I doubt that it would be an issue.

Where do you plan on getting your aluminum from?  Know of any places that will cut it to specific dimensions before they ship it out?

On Sunday, September 16, 2012 9:49:25 PM UTC-7, Misha wrote:

[John Losche]

in the US, onlinemetals.com was Mike's source for tables when MM 1.0 got started.

--
 
 

[Alex]

In case anyone is curious to prices.  I'm just ordered from onlinemetal this morning for a Prusa sized sheet and my scaled size sheet:

Shipping was to Northern California

PRODUCTS
1 piece @ $24.94/piece
Aluminum 3003-H14 Bare
Sheet
PVC 1 side
0.125"
Cut to: 12.5" x 17.5"
-------------------------------------------------
1 piece @ $7.56/piece
Aluminum 3003-H14 Bare
Sheet
PVC 1 side
0.125"
Cut to: 7.875" x 8.425"
-------------------------------------------------

Shipping Cost: $12.70
Cut Fee: $0.00
Tax: $0.00

[Misha]

I did not factor in the extrusions in the simulation, but do not think they will have a large effect. The contact surface area is very small.  I will probably go with onlinemetals.com for my aluminum.

[Misha]

Shims would probably work, I am trying to avoid using plastic parts because I will be unable to print ABS until I build this new bed, my current home made bed just does not get hot enough for anything except PLA.

[Jonathan Bastien-Filiatrault]

That looks very good, very nice work Ezra !

I was wondering, if I pull the kapton heater off the glass, will the adhesive become unusable ? Can I just degoop it and stick it on with kapton tape to the aluminum ? I might just wait for the upgrade kit to come along before sticking the kapton heater to anything.

[Larry Knopp]

@Jonathan
Be *very* careful in removing the kapton heater.  The PSA (glue) is tough, and the heater traces very fragile.  I would also want to source "replacement" high-temp Pressure Sensitive Adhesive to re-apply the heater...  just taping it will most certainly not allow good contact with the surface you are trying to attach it to.  This is *highly* likely to cause the heater to fail catastrophically.

@Misha, et al...
I'm not sure what results Ezraor anyone else is seeing regarding aluminum extrusions acting as heat sinks, but I will attest that it does not take much contact area at all for a cool piece of aluminum that size to suck one hell of a lot of heat.  In my own testing, I have found that four 15mm x 15mm contact areas at the corners (w/ 1/16" cork insulation) of an aluminum heat-spreader with a fairly heavy-duty aluminum support structure beneath...  within 15 minutes, the support piece had soaked enough heat through those tiny contact patches that it read 92C with an IR temp gun.

Just my 2c...
I haven't tried the extrusion arrangement personally, so I can't say for sure, but I would be trying to insulate the hell out of it.

--
 
 

[verohomie]

Ezra -

Will you need to get the heaters made with the 3M adhesive on the other side?

Brad

On Monday, September 17, 2012 7:30:19 AM UTC-4, Ezra Zygmuntowicz wrote:

[Ezra Zygmuntowicz]

Brad-

Nope this solution uses the same kapton, g10 and boro we have right now. I am testing this bed tonight and having excellent results from it. The temperature in pronterface actually matches the temperature of my IR thermometer!. Before this fix when pronterface said 100C my IR thermometer said 179C!!/

This was mainly because tyhe PTFE tub completely covers the thermistor head as well. One thing you can all do right now to improve performance is to carefully use a xacto knife to remove the PTFE tubing just from the 2mm tip of the thermistor to expose the glass tip so it can accurately measure temperature.

Once I validate this solution for the next few days of heavy printing and am completely satisfied with it we will start shipping the new Y axis upgrade kits with this new system.

But I do not like to leave anny of my customers in the dark or left behind so I am going to put together a kit that will include a new kapton heaterm a new piece of glass, one of these aluminum sheets with the reliefe hole cut into it, the two machined aluminumn ends that bolt to the G10 and 2 pieces of tslot extrudsion to con\strain the glass.

This kit will be made available to anyone who has revieved one of the older Y axis with the 4 printed corners. If you aren't having isues with your Y axis nd don't care then this is optional but if you are having the bending or bowing glass when the kapton heater has current flowing or any other issues then this solution is what you want. 

I don't know exactly what the coast will be yet but I am going to make these upgrades to the print bed available to people who already got the 4 printed corner version y table as 10% above my cost plus shipping just so you can all have the new hotness that works great.

So an upgrade kit would include>>

1. New kapton Heater

2. New boro glass print bed

3. 1mm thick aluminum sheet with relief hole cut in the center

4. 2 pieces of machined aluminum end mounts to bolt onto the G10

5. 2 pieces of 298mm tslot 20mm extrusions for each end of the y axis

6. Vitamins to put this all together.

You will be able to resuse you G10 and you will have an extra piece of glass as well! Unfortunately if you try to remove the kapton heater you already stuck to your glass it will be ruined for sure.

So after a few more days testing I will try to get a good price together on this setup and get it up for sale for folks who want to give it a go.

*****************HUGE TIP**************************

For much better results with yoru current setup *right now*. Take an xacto knife and carefully remove the PTFE tubing that encases the tip of the thermistor. Right now these thermistors come completely encased in PTFE which I just noticed. Pronterface would report 100C and my IR thermometer would report 179C, after cutting away the PTFE encasing the thermistor pronterface says 100C and my IR thermometer says 102C.

I think this exaggerated the bowing of the glass too because the thermistor was very slow to respond to heat changes abnd made the kapton heater work way way overboard 

-Ezra

[Misha]

Here is my update design, and probably what I am going to build.  It is the following sandwich, from top to bottom:  I am using square extrusions in the model because that is what I could find a ready CAD model for online, but I will use the same quarter round that Ezra is using.

----------------

3 mm Boro

----------------

Kapton heater

----------------

.063" Al sheet from OnlineMetals.com

--------------------

.046" G10 from Amazon

-----------------------

I ran some rough heating calculations based on continuous 250W heat source from the heater, so the simulations are not a very good quantitative model but allow for some qualitative analysis.

Above is the temperature after 1000 seconds of running, the actual numbers don't really matter but most of the surface is within 8 K.  This is further backed-up by the follow graph of temperature gradient:

Again, these are really rough simulations run in Solidworks on a rough mesh, all I am using them for is some qualitative guestimation on how the heat will spread.  I have also run simulations without the aluminum sheet with just the heater on G10 and you end up with just a giant hot spot on the boro right over the heater.  The aluminum helps a lot.

[UK Chris]

Hi Misha,

I'm using most of your design here. Kapton 200 Watt/24 VDC heater
(TL), stuck to underside of 3mm ali, with 2mm window glass on top. TL
corner pieces with raised walls (3mm extra) to accommodate ali +
glass. Slight bow in ali (3mm total), but who cares. Small blob of
sealant at each corner, attaching glass. Warm up 20 degsC to 80 degs
C in approx 6 mins.

Regards

[Duncan Frazier]

any chance you could run a simulation like that with the aluminum layer replaced with a perforate aluminum sheet like can be found here.

http://www.onlinemetals.com/merchant.cfm?id=1003&step=2&top_cat=60

I was thinking with the correct hole pattern you might get very close to the same amount of heat spreading with a nice reduction in weight.  I was just gonna buy a sheet and slap it on my heated bed, but I'm curious since your using science and computers and such if there might be a more definitive answer.

thanks,

Duncan

--
 
 

[Alex]

I see you've reduced the thickness of the G-10 and Aluminum sheet.  Probably to save weight.  What's your intended Y length.  Any issues with bowing because of this?

[Misha]

I though about using perforated aluminum, but I think it will cause hot spots on the heater without saving much in weight, at this thickness the aluminum only weighs like 100g.

[Misha]

I reduced the thickness of the Al and G10 in order for it to fit in the 6mm slot in the extrusion.  I am designing for a 290x290mm bed size.  No idea about bowing, this is just theoretical for now.  We will have to wait for Ezra's testing results.

[Jonathan Bastien-Filiatrault]

Starting now, are you planning to sell the Kapton heaters with the PFTEctomy already performed ? It would suck for me to botch the surgery and have to wait for another heater to get shipped...

[Caanon]

Great stuff. 

ChrisUK: So when you say 3mm total, you mean the aluminum bows away from the glass a total of 3mm? 

Misha: love the simulations.  What toolbox did you use for it?  Quick question... if you sandwhich, where do you run the wires: a hole in the middle, or a channel out the side?

[Misha]

I am just using SolidWorks Simulation, it isn't the best but it gives a rough idea.  I don't really have any metal working tools to make the channel so I am just going to drill a hole.

[UK Chris]

Caanon,

I mean that there is a 1mm gap between the ali and glass on the left,
and 2mm on right. Have a look at the picture 821, on the 'slump'
thread, post 5. You might just be able to see it at the rear of the
table. But it matters not - the glass is flat above it.

Regards.