Alternative bed proposal

I have started constructing a new bed which I hope will address the flatness issue.

Proposal 5mm float glass, but machined to be the top supporting surface. This will be assembled similar to the manual levelling assembly.

The advantages of this method are:-
1. The top glass surface is fixed and this becomes the reference surface, i.e. not sitting on the aluminium heat spreader and clamped in the corners, causing distortion.

2. Related to 1 above, no clips are required.

The disadvantages are (and I value you critical comments and suggestions) presented along with my questions:
1. The glass has to be shaped and drilled. I propose to use a combination of glass cutting, grinding and drilling as used for my stained glass work, so this is not an issue.

2. The holes are too close to the edge and will be weak points. Yes this is an issue, I propose to make the surface 5 to 10mm wider and 5 mm deeper at the front to move these holes further away from the edge at the front, but this is not possible at the rear near the z axis.

3. With the glass fixed, will I be able to remove the printed articles without damaging the glass or the bed mount or drive parts? My parts have come away relatively easily with the glass left in place, so is this an issue?

4. Do we require the aluminium heat spreader? From a diagram I found the tracks appeared quite evenly and closely spaced, so by the time the heat has transferred through the (poor thermal conductivity) glass it should be 'spread' as far as the print is concerned?

5. If we remove the aluminium plate, can the PCB heater be placed directly onto the glass using heatsink compound?

6. Related to 4 and 5 above is the aluminium plate a heat spreader to protect and allow even distribution of removal of the heat from the PCB (prevent local hot spots) or is it to benefit the 3D print?

7. Looking at my bed, it is the MDF 'flat' parts that are not flat and bow upwards in the middle, or I guess downwards at the edges, so am I 'wasting my time' (love that phrase from some old movies!) and trying to fix a problem in the wrong area, where a piece of 5mm glass clamped as per the current design would be just as good.

My issue is I do not have a machine shop, so any new parts have to made by hand or purchased and so far all the great improvements that I see to the bed are not commercially available, but correct me if I'm wrong.

Thanks for taking the time in considering this.

---

Ormerod #007 (shaken but not stirred!)

I think there are two main problems in the current bed design to be addressed. The major one is the instability of the MDF you referred to in (7). I don't have a machine shop either, but to address this issue I constructed this triangular aluminium replacement bed support from 3mm aluminium using a centre punch, power drill and jigsaw.



Someone else posted that he used square section carbon fibre tube to strengthen the existing MDF.

The minor issue is that the glass bows slightly because it is clamped at the corners. My rough measurements suggest the centre is about 0.1mm higher than the corners. I'm addressing this in a couple of ways. First, I've replaced the 9mm photo frame clips I was using by 11mm clips, which are compressed less and so should provide a lower clamping force. Second, I'm bought some 4mm glass plates. Since stiffness of a plate varies as the cube of thickness, this will be 2.37 times stiffer than the 3mm glass, so the deformation should reduce by a similar factor. 5mm glass would be even stiffer, but I'm concerned about the additional weight. I hope to do some detailed measurements on both the 3mm and 4mm glass this weekend.

If you want to support the glass directly, instead of drilling it, why not bond threaded metal pillars to it? I'm thinking of those glass-top tables in which the top is bonded to the legs. A quick search came up with UV-curing adhesives for glass-to-metal bonding, for example [www.loctite.co.uk].

Edited 1 time(s). Last edit at 01/31/2014 05:57PM 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'm impressed that you used hand tools for that result dc42, you've got some very neat cutting - in the photo it looks like there may be some bends in the sheet - is this an artefact of lighting, or intentional bending for stiffening or meeting the mounting height of the bearings?

treth - to go a point at a time;

(proposal) 1 are you intending to support the glass direct from the bearings or subframe? If so you'll have to reach around the heater and spreader (if you're keeping the spreader that is). You'll also have to hang the heater (at least) from the glass
(proposal) 2 - definite advantage

"disadvantages":
1 - not really -if you extend the subframe by replacing the existing (plywood) Y rib, with a more workable material than glass, the glass top plate could stay pretty much as a slab sitting on the subframe (presuming that the glass is larger than the current piece). Wedges could be used for distancing between the subframe and the glass (double wedges allow adjustment of height by pulling them together or pushing apart btw, and you could print these), and they could be countersprung (ie glue a fastener to the bottom of the plate and to the subframe that pulls against the adjuster)
2 no need for holes if you use another height adjustment (eg wedges above)

3 I've never had a problem popping the parts from the bed, no matter what they're stuck with (but I'm very butch) - I can't see a problem with a semi-permanent glass top (I've really only removed my glass five or six times - two of these were for y belt maintenance, two for trying different types of glass

4 the heat spreader is pretty much there (in my mind) because the copper may get hot, while the substrate of the heater and the glass don't - so it could set up hot spots that may be damaging, you can probably do without it though - or replace it with silicone, which is very heat resistant and if spread thinly enough will probably help conduct the heat better than the aluminium (air gaps between the aluminium and the heater and the glass will block the transfer and overcome many of the merits of the spreading- kinda like the way that heat tranfer compounds used on heatsinks are actually quite good insulators, but not as insulating as the air they displace)

5 probably yes - but since you're using the glass as the supported refernce surface (I think) then it would be easier to hang the heater off it with silicone

6 I think the benefits to the print are related to the above (ie the aluminium helps the glass get the heat to the bottom of the print without shattering it)

7 indeed if you're going to have a rigid thing, then hang it off a soft floppy thing it's not so useful - but if you extend the core subframe with something stiffer (by core subframe I mean the plywood parts that connect to the bearings) as far as your levelling points (three as in dc42's subframe or the original, two near the rear and one centrally at the front - they don't need to be right at the extremes, just reachable for adjustment), stand your glass plate on these three points then hang the heater underneath with adhesive (I'd go for silicone every time), with an insulating layer underneath the pcb, you should be in a much better place.

I actually do have a kind of mini machine shop, but no chance I could machine anything big enough on my equipment - I do though have a plan for fabricating from extrusion a possible solution - when I get the chance I'll be trying it out and if it works I'll post the pics (well if they look as good as dc42's I will anyhow)

Ray

@DC42 and Rayhicks,

Well (as always) you have both inspired and provided great feedback. Firstly I'm also impressed with the aluminium plate you produced dc42 and I shall definitely be going down that route now! As I typed my point 7 regarding the MDF, I thought "I have got this all wrong" and your post confirms this, I was really looking at "the minor issue" so thanks.

The calculations on the glass stiffness is also reassuring, so I shall now cut my glass down to the original designed size and use that.

Regarding it being an advantage not having the clips, from the discussion I see that this is also outweighed by the maintenance issues, so again thanks. Perhaps with the 4mm glass some sort of location and fixing device that applies pressure to the sides could be useful, locating with the two rear screws (upside down top hat shape) and clamping from the single front screw (using dc42's three point design). But there I go again on a minor point.........

Good discussions on the heater, spreader and silicon's etc., totally agree on the heat sink compound effect (experienced the insulation effect 40 odd years ago and learnt!). I also 'forgot' to consider those glass adhesives for mounting the nuts so another route for the future.

I have been addressing my lack of a workshop and have joined a "Maker" group, in the UK these have appeared in many places and I'm fortunate that one has been set up by Falmouth University in Cornwall UK [makernow.co.uk] some 30 miles away and yesterday completed my training for 3D printing (ABS only), laser cutting (no metals and acrylic up to 4mm) and CNC 3 axis (no metals!) but can do acrylic and they use on 10mm sheets quite regularly.
So I plan to print the 'hot' parts on their UP! machine, which incidentally I see their default ABS settings produce prints marginally better than the Ormerod (original defaults), but the Ormerod now far exceeds them with what you guys have shown.

So no route for aluminium sheet (probably could do the small cut outs) due to no coolant, but some great additional tools and knowledgeable team.

Thanks for the very useful help.

---

Ormerod #007 (shaken but not stirred!)

Thanks for your kind comments on my bed support. It isn't bent, except slightly at one corner where I cut it without adequate support for the piece that was about to fall off. The 6 drilled holes need to be in exactly the right place, but the position of most of the rectangular cutouts is not critical. I cut them roughly with the jigsaw (or drilled holes and then joined them with the jigsaw) and then filed them to smooth the edges and enlarge them as necessary. I used the MDF bed support as a template.

---

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 took some measurements using 3mm glass, then switched to 4mm and remeasured.

Using 3mm glass, the centre of the bed was 0.175mm higher than the average of the corners. Using 4mm glass, the centre is 0.075mm higher than the average of the corners.

This is a useful improvement. I used the same 9mm photo frame clips, so the clamping force will be a little higher using 4mm glass. I currently have 3.3mm hardboard under the glass for the clips to lock into, so I'll try using something thinner such as a piece of copper-clad board next.

---

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].

Quote
dc42
I took some measurements using 3mm glass, then switched to 4mm and remeasured.

Using 3mm glass, the centre of the bed was 0.175mm higher than the average of the corners. Using 4mm glass, the centre is 0.075mm higher than the average of the corners.

This is a useful improvement. I used the same 9mm photo frame clips, so the clamping force will be a little higher using 4mm glass. I currently have 3.3mm hardboard under the glass for the clips to lock into, so I'll try using something thinner such as a piece of copper-clad board next.

Notice two things that could influence your measurements

The force between the hot end assembly and the extruder drive assembly though the filament tube, mostly at the middle (without "the rubber band fix)

Ad to that the extruder drive assembly puts weight on the middle

If your x-axis was straight in "relaxed state" it will now have a small curve at the lower side from the force of the 9-10 mm bearing

This will echo in a higher measurement of the nozzles distance from the middle of the bed

..no it ain't perfect, but it don't seems to notice and prints just fine anyway



Erik

To make my point clear, when measuring pos. through nozzle hight readings you are adding your reference point to your measurements
Or to put it another way, if your bed is perfectly flat, then you are measuring your reference point, the flatness of the x-axis/bearing contact and the force of the filament tube

Z readings along the x-axis at Y100 aka the middle of bed, Z pos. read at nozzle touchdown on paper
Reading taken on 6 mm flat glass, only resting on silicone rubber both ends, no clips

I would have been nice to make better reading with more decimals

Notice how the force of the tube is at its hight in the middle of the bed
X25 Z0.0
X50 Z0.1
X75 Z0.1
X100 Z0.2
X125 Z0.2
X150 Z0.1
X175 Z0.05
X200 Z0.0

Removed the filament tube:
X25 Z0.0
X50 Z0.0
X75 Z0.0
X100 Z0.05
X125 Z0.05
X150 Z0.0
X175 Z0.0
X200 Z-0.05

Removed the filament assembly, notice all readings are lower, with the filament assembly removed the reference point for the measurement (the contact point of the bearing) is moved anti clockwise as seen from the open end of the x-axis

X25 Z0.0
X50 Z-0.05
X75 Z-0.05
X100 Z-0.05
X125 Z-0.05
X150 Z-0.05
X175 Z-0.1
X200 Z-0.15

So my starting point is no longer to make the bed stable and flat, it is to make the x-axis ditto
..or to rewrite spoon-boys Matrix quote:

"Then you'll see, that it is not the glass that bends, it is only your x-axis."

Erik

ive just had a little play around with the standard bed file in the master directory.

Ive never done anything like this before so what do you think?
What format does it need to be for a lasercutter to use it?

Quote
ormerod168

Notice how the force of the tube is at its hight in the middle of the bed
X25 Z0.0
X50 Z0.1
X75 Z0.1
X100 Z0.2
X125 Z0.2
X150 Z0.1
X175 Z0.05
X200 Z0.0

The visual:

X40 Y100, diff. 51 gram




X180 Y100, diff. 61 gram




X120 Y100, diff. 107 gram




So what does this variation of 51 - 107 - 61 gram do to my nozzle hight readings?
A difference at about 0.2 mm in nozzle hight from the ends to the middle of the bed

Erik

Quote
PaulHam
ive just had a little play around with the standard bed file in the master directory.

Ive never done anything like this before so what do you think?
What format does it need to be for a lasercutter to use it?

Paul, did you get you bed laser cut?
If so how good is it and do you recommend?

---

Ormerod #007 (shaken but not stirred!)

Out of curiosity what do we think about using plywood instead of MDF for the bed? 4mm thick ply is pretty rigid but can be laser cut using hobby gear.

---

RS Components Reprap Ormerod No. 481

How about plywood for the central section of the bed, extending to the back but not to the sides or the front, and square section carbon fibre tube running along the back and to the front centre where the heated pcb etc. need to be supported?

---

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].

My mod for this weekend if I can find the time, is to try topping the MDF bed with a 3mm piece of Tufnol (phenolic resin sheet). Light, very stiff and stable, rated for well over 100C, and easy to work with - it seems the ideal material unless I'm missing something?

I did start to make a new bed from an 8mm piece, but it was OTT, and I got bored trying to file out all the holes and make it look reasonable.

3mm is less to work with, and if I put it on top of the MDF, I only need cut to size and drill a few holes before I can try it out. It could be simply taped to the MDF for first tests.

If it works as well as I hope, I've got a second sheet available, so I can replace the MDF, or simply bond the Tufnol to the MDF and call it good enough.

Dave

@epninety, good idea about the Tufnol, that could be good in this situation.
I can only recall using it as a rod (nightmare to machine (not by me!)) where it had the properties we required, but I'll be interested to hear how rigid it is in sheet form.
Certainly does sound the ideal material. Certainly will be good for those in a damp humid environment.

I'll hold back on my aluminium bed again, there is always a good reason for not doing something!

---

Ormerod #007 (shaken but not stirred!)

I'm an electronics engineer, and it's very common to use Tufnol for jigs and test fixtures, which is what gave me the idea.

As for stiffness, I just tried to flex a 300mm square of 3mm, holding it at each side, with my thumbs in the middle. Using 'reasonable' force - enough to make my thumbs hurt - I get a small deflection, less than 5mm?

Interestingly, the non-genuine 'Tufnol' I have has a noticeable 'grain' to it, it's a bit stiffer in one direction than the other. I don't think that's true of genuine 'Whale' etc.

Dave

Quote
epninety
...I only need cut to size and drill a few holes before I can try it out. It could be simply taped to the MDF for first tests...
Dave

Or you could move your heat spreader down, would only have to drill one hole then (for the temp. probe), you don't have to cut of the corners of the MDF as I did of cause, just put the heat spreader on top of the MDF and arrest it with the screws in the four corners

Sagging MDF, the Q & D fix that anyone can do:
[forums.reprap.org]

Erik

I haven't thought about it in great detail, but I intended to drill the heated bed support holes (3 or 5) plus the (3?) countersunk holes where the MDF is screwed to the rib and bearings.
Tufnol will take a thread really nicely, so drilling and tapping the bed support holes means I can get rid of the nuts and support the heated bed on 3 springs as I currently do.

Quote
Treth

Quote
PaulHam
ive just had a little play around with the standard bed file in the master directory.

Ive never done anything like this before so what do you think?
What format does it need to be for a lasercutter to use it?

Paul, did you get you bed laser cut?
If so how good is it and do you recommend?

I got some quotes but it wasn't anything I would spend.

Made out of 3mm Alu, £250, £200 & £50

So I have moved the aluminium heat spreader down on top of the mdf bed. its better than before but stil not perfect as it still rocks on the rib.

Quote
epninety
....plus the (3?) countersunk holes where the MDF is screwed to the rib and bearings...

I tried that with the heat spreader, was not a good idea, made the bed walk all over the place, but then this was aluminium
If you (like me) want to adjust you glass on 3 points you want those 3 points to have the most direct and stable support from the bearings and that is through the ends of the y-rib and the y-cross ribs, you get that imho and experience by letting your new bed overlaid hang on these end points only

Erik

Quote
PaulHam

Quote
Treth

Quote
PaulHam
ive just had a little play around with the standard bed file in the master directory.

Ive never done anything like this before so what do you think?
What format does it need to be for a lasercutter to use it?

Paul, did you get you bed laser cut?
If so how good is it and do you recommend?

I got some quotes but it wasn't anything I would spend.

Made out of 3mm Alu, £250, £200 & £50

So I have moved the aluminium heat spreader down on top of the mdf bed. its better than before but stil not perfect as it still rocks on the rib.

Hi Paul, thanks for info on the costs, that's what I was interested in, 'too expensive'!

@ormerod168, while doing some DIY decorating, I have had time to think about this and agree with your points
The key points are:

1. The guide rails must be in the same plane, which should be OK with the laser cut ends.
2. The bearings are the key fixture points so anything from the bed must be a good secure fix to these.
3. The current design is flawed using the corners with the rear overhang and front none supported corners.
4. The ribs across the bearings in MDF are probably OK due to their alignment (vertical plane)
5. The glass for the top surface should be very flat

So that brings me back to my original thoughts, that the glass needs to be secure with bearings/ribs.
DC42's aluminium plate and the various other suggested machined plate options address the using three mounting points, but then the heater, insulation and glass are not quite addressed.
I believe the difficulties in machining the glass and that the glass should be easily removable need addressing if this area is to progress.

It may be that the aluminium bed solves the main issues, so the above is not an issue, but it be good to clamp the glass over the bearings. Perhaps the front of the aluminium bed needs to protrude slightly more with a larger glass bed to allow the glass to be clamped in line with the rib?

---

Ormerod #007 (shaken but not stirred!)

I would try a new bed construction aswell. Therefor I'm looking for the files of the original lasercut parts. Any idea where they hide?

[github.com]
There is the y-rib and the y-cross-rib - but the top?

I'm sure just blind?!

Bed.DXF in [github.com]

I think

Ahh, sure, dxf

Thanks alot!

Quote
djar
I would try a new bed construction aswell. Therefor I'm looking for the files of the original lasercut parts. Any idea where they hide?

[github.com]
There is the y-rib and the y-cross-rib - but the top?

But also look at PaulHam's post [forums.reprap.org] where he provides a good dimensioned triangular bed similar to DC42s picture.

I'm looking at Front Panel Express/Schaeffer [www.schaeffer-ag.de] to make my ali parts from 3mm black anodised sheet, I have used them for front panels in the past and have been good quality. They also have good free software for developing your part, well worth a look. They provide counter sunk holes, slots etc., as required. As a panel, it is rectangular so will need to provide slots with cutoffs to finish after delivery. Doing it tjis way and pretending its a panel keeps the cost down.

As always I won't have too much time to explore this hence the link for anyone faster than me!

FYI my thoughts at the moment are to enlarge the bed to provide 3 new mounting points outside of the printing area. These points will be in line with the Y axis mounts and the single point between the heater connections. This obviously needs a new piece of glass (thicker, say 4mm or 5mm) with three holes in which guides can be glued to allow locations on screws/studs.

The new ali bed will be similar to the triangular ali beds suggested, but with extensions to mount the glass locating studs/screws depending on what I identify with nuts to set the glass level.

The current PCB heater, heat spreader and insulation will be kept and my thoughts are that this assembly will be spring loaded against the new glass bed.

This is early days so try to find flaws (but be gentle), but my thoughts are this addresses:-

1. The levelling being done on axis of the bearings.

2. Allows clips to be located outside the printing area.

3. Places the glass in direct contact with the aluminium bed, for max stability.

4. Allows the glass to still be removable.

Probably my biggest unknown is the whether the spring pressure of the heater will be a suitable solution.
Look forward to comments.....

---

Ormerod #007 (shaken but not stirred!)

Hmmmm I print on glass, and the only way I can get it to stick is to remove the glass and wash it with washing up
liquid, air dry, then mount.... Also I've found that removing the prints is easy... I put print and glass under the cold tap, and the print float off..

---

Please send me a PM if you have suggestions, or problems with Big Blue 360.
I won't see comments in threads, as I move around to much.
Working Link to Big Blue 360 Complete

Quote
KimBrown
Hmmmm I print on glass, and the only way I can get it to stick is to remove the glass and wash it with washing up
liquid, air dry, then mount.... Also I've found that removing the prints is easy... I put print and glass under the cold tap, and the print float off..

My proposal is just for a level stable bed (with removable glass (point #4)). You can use Kapton tape, pipe cement or just plane glass which ever you choose. I don't seem to have any problems with my original Kapton taped glass so have stayed with that.

---

Ormerod #007 (shaken but not stirred!)

My Tufnol bed stiffener has been fitted for a day now, and initial results are very encouraging.

I cut a 3mm Tufnol sheet to fit on top of the MDF bed, mounted on the 3 screws that fit the MDF to the ribs, and with the 3 mounting holes for the heated bed tapped to take sprung screws.
There is a single larger hole drilled for the thermistor wiring. Temporarily I don't have the cardboard insulator fitted, as I needed longer screws to fit the tufnol, and didn't have countersunk ones handy.

I can't measure flatness diagonally with the heated bed fitted, but front to back (x axis) it measures less than 0.5mm distortion along each side, which is about the same as the unused piece of Tufnol I have left over.
At a glance I didn't see any change from cold to hot, but I haven't checked it closely. I set the bed level at cold, and then got a lovely even base to the first print I tried. It does take noticeably longer to warm up with the insulator removed.

I haven't had time to test it further, but I will try to run the temperature up and down over the coming week, and check how level the bed is next weekend.

Next thing would be to drill/cut the extra holes required and replace the MDF bed entirely.

A few pics...

Looking good epninety - also looks like you should build it into the table it's stood on!

Ray

Just to follow up on this... I heated and cooled the bed a few times over the last week, and I've made several prints today, and I've not had to adjust the Tufnol bed alignment at all, it's very stable and repeatable.
The only adjustment I've had to do since installation is because I just changed the Z nut bracket again.

Very happy with that. Maybe useless info with everyone looking at the ali bed option, but worth noting it here.

BTW Well spotted Ray, the Ormerod does look a bit like a Mini-Me of my rexroth desk and workbench, maybe I should give it a matching oak bed too

Dave