3D printing

[awemawson]

There are some really quite exciting materials being offered as 3d printing filaments. So long as the hot end is capable of the correct temperature regulation you can print metallic filled materials, carbon fibre, wax investment, and even conductive materials. It's this capability and the dual heads that attracted me to the 'BigBox' as it ticks all my boxes. No doubt a step learning curve though.

[vtsteam]

Moldlay is designed specifically for the use of creating positives for investment casting molds. Moldlay is a wax-like material that can be printed on a normal desktop 3D printer, at temperatures of 170-180ºC.

That's just what I was suggesting..

Now we need a diy heavy square steel tube base with rigid ways (NOT end supported 1" dia rod) , and a gantry of epoxy coated box section birch ply for low specific modulus, aircraft style, and a commercial head.

Then we're talking making adequate patterns for intricate lost medium casting with a DIY machine at reasonable cost.

[efrench]

What do you mean by a commercial head?  Do you have any links to one?  Cast iron or epoxy granite might make a better base than steel. 

[vtsteam]

A head that is purchased. I don't have links, but I haven't looked for any -- I'm not interested in building a 3D printer.

In speaking of a DIY machine I'd admire of low cost, the suggestion of welded steel square tubing is generally within the range of capabilities and budget of folks here. If you cast iron in pours of, say 100 lbs, or better, or mine and surface granite, I'd say go with those. Personally I'd likely want steel tube anyway above the granite.

[Swarfing]

I recently built an i3 Prusa for a friend, first out of steel box because I thought it would be much better. Had so many problems trying to tweak it, ended up rebuilding it out of laminate 10mm birch ply to a nice thickness  x3 (30mm) and it turned out much better. I fixed this down onto some 50mm mdf for the base making the whole thing very stable rather than the thin sheets of material they recommend in the build plans. Had to cnc the bed out of aluminium so there are some thing that are not available to everybody. The parts are available for the build off ebay at quite reasonable prices from others replicating the parts.

[efrench]

I suspect most DIY printers have commercial heads using that definition The E3d v6 is probably the most popular.  It sells for around $80. Clones from China are anywhere from $10 to $30. 

Quality control is probably the biggest problem with most of the kits on the market today.

I'm in the finishing stages of building a delta printer with a 350mm build circle.  It's all aluminum with a torsion box base.  I expect that it will be more than rigid enough for the forces being generated.

[vtsteam]

I suspect most DIY printers have commercial heads using that definition

Those which don't would have non-commercial heads. Besides the adjective, there's also a noun, commerce that describes the convention of buying and selling. No matter that say, barter, is less common, the word still finds a use. Sorry for any confusion.

[picclock]

@swarfing
>> I recently built an i3 Prusa for a friend, first out of steel box because I thought it would be much better. Had so many problems trying to tweak it, ended up rebuilding it out of laminate 10mm birch ply to a nice thickness  x3 (30mm) and it turned out much better.

As this is the machine I am thinking of purchasing could you expand on the problems you encountered ?

@ EFrench 

The E3D-v6 online is the only head I have found to do the higher temperatures needed for nylon and polycarbonate. It is also self assembly.
See : -   http://wiki.e3d-online.com/wiki/E3D-v6_Assembly   M6 nozzles are available on ebay for not a lot.

Will be interested to see how your printer performs. My info is that the delta ones are much trickier to get properly set up, calibrated and working well. Good luck with it !! Could you post some pictures when done ?
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Having looked at the rest I am fairly certain the the Prusa I3 will be a good starting place for me. I am thinking of using a granite base and fixing the parts to it. May swap the plastic rod holders in the Y axis for aluminium ones. Not sure why both sides are fixed like that. IMHO you need one rod to keep the axis true and the other one merely to keep the platform level. By fixing both rods you are setting up forces working against one another. Also the linear bearings should be clamped and fixed by a narrow ring in the centre rather that along the whole length as this would allow them to find their true correct orientation (there are two bearings on one side and one on the other). Will build it as is and take measurements, then fix as needed. Will post build with pictures if anyone interestted.

Have been checking out 3d modelling software - think I am going to go with Designspark from RS. Its fast and quite easy to learn (and free !!). Love the way you can make threads (see link - no sound or root/tip radius but could be added). 
https://www.youtube.com./watch?v=6GZXIrbvh80
Still haven't figured out how to do cones yet !!.

Best Regards

picclock

[Swarfing]

The only problems I really found was rigitity. Making it out of steel (welded) throws up problems of alignment as you can imagine with heat twist etc. Making it out of Acrylic just seems a bit week for my stomach, yet a sensible build medium if you take my bias away. Birch ply in over engineered sizes works great in my book and it can all be glued together and fettled with ease. There is also the bonus of being able to make it much taller if you wish?

Using tubing for the couplers is a poor substitute for ridged ones, yet rigid ones allow noise transference which shows up in the finish which will not be as clean. Make sure you mount the base on rubber feet, Stainless steel threaded rod as the threads you will find are much cleaner and accurate instead of plain steel which usually is rolled thread and not as accurate (As accurate as threaded rod can be which is not very anyway)

[awemawson]

Can it not be made fairly simply using a frame of standard aluminium extrusions as are popular on CNC plasma and router machines ?

[PekkaNF]

Basically you want as light as possible structure, good 3d printer head must be able to accelerate and move relatively fast and fast stepper motors are small. Nevertheless should be rigid eneough, othervice printhead shakes all over the path.

Pekka

[Swarfing]

With the i3 we are only talking about the main static frame so weight does not come in to it. Yes it could be made from extrusion of course, but these days extrusion is a very hi cost in comparison to other materials.

[picclock]

@swarfing

Prusa I3 - I had already planned to s/s allthread for the m5 z axis because it always looks like the threads are better formed. I was under the opinion that they were still rolled - but if not what process is used to thread them ? IMHO this is one of the weak areas which may prove tricky to improve on - but as its only being made in plastic with all its dimensional issues (moisture absorbtion, high coefficicient of expansion etc.) it might just be good enough. The other axis of the machine are belt driven direct from a stepper motor. I suspect the reason for change on the Z axis was to improve the resolution to allow for thinner layers and achieve smoother finishes. 

@awemason

The Velleman printer sold by Farnell and maplins uses an extruded frame - although it only uses one z motor which will likely lead to droop on the undriven side. Could probably be compensated for by bed levelling to some extent. http://www.maplin.co.uk/p/velleman-k8200-3d-printer-kit-n82qg

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

The frame (gantry) weight with the Prusa I3 design is irrelevant as it is a static part, so can be made very stiff, to allow good accuracy for head positioning in the X and Z axis. The Y axis will have to accurately achieve position with the weight of the heater bed and glass sheet plus the model weight and the cable drag. I think thats why the I3 design is a good one. Two out of three axis sorted for stiffness.

Best Regards

picclock

[S. Heslop]

With the i3 we are only talking about the main static frame so weight does not come in to it. Yes it could be made from extrusion of course, but these days extrusion is a very hi cost in comparison to other materials.

http://www.valuframe.co.uk/

These kinds of extrusions seem very cheap if you're selling kits, since you can leave alot of the assembly and fitting to the guy buying it. Wonder how much the cost stacks up against paying for water jet/ laser cut plates though.

[Swarfing]

Ha ha! yes is reasonable priced until you add delivery cost? It's all horses for courses, you choose your material of choice and go for it. Judging by the comments so far you can see people just have their preferences for different materials. I have used a lot of extrusion for my past CNC projects, in my final CNC router build I opted for welded Steel (Cheaper and stronger).

The last lot of S/S threaded rod I bought looked too good to be rolled and I was under the opinion it must have been cut? I can't really comment on processes used. In general S/S rod just seem to present much cleaner threads and by virtue gives a lot less issues. You could opt for ballscrew but really a bit over the top for a small 3D printer unless you want that extra perfection?

At some point I will build another for myself and again I would go for the birch ply approach. My friend is very pleased with the results he is getting. It is the 13 Laminate variety I used so is very stable especially with a coat of lacquer. I wish I had a pic which I'm sure would speak a thousand words. The overal advice I would give is to keep it simple and don't complicate it too much. With a working machine he has already remade some of the plastic parts to a much beafier spec.

Hope that helps

[tom osselton]

Have you checked out your area for a makerspace or hackerspace for laser cutting?

[nrml]

Maplin have an offer on a bundle of parts and plans to build a complete printer for £399 at the moment http://www.maplin.co.uk/p/velleman-k8200-3d-printer-kit-n82qg. From what I have researched so far, it is gives decent performance for the price and lends itself well to improvement mods and upgrades. I would love to get one just to build and improve even though I know nothing about electronics or CAD / 3D modelling. Money is a bit tight at the moment, but it is definitely on the cards at some point.

[picclock]

@nmrl

I mentioned the vellman kit about 5 posts previous. It appears to have some issues in that the head is a (relatively) low temperature one unsuitable for nylon or other exotic plastics - but its OK for PLA / ABS. The other issue is that the z axis only has one motor. This will lead to head droop on the X axis, and although it can be sort of compensated for by aligning the bed. I don't think it has a heated print bed either, which could lead to adhesion issues.

I think it depends what you want to do with one of these machines. I want to make working mechanical parts out of nylon or other loaded materials so they will have to be hard and hard wearing, and hopefully accurate  ,  but if you just want to make decorative parts or parts with low use requirements this may suit. The Velleman model @ £400 (-1p :-) ) is a more expensive - see http://semiu.co.uk/product/prusa-i3-complete-kit/ for the prusa I3 (metal frame). I would buy without Power supply for £330 as I can get a 12v 30A psu for £15 on flea bay rather than modify a PC psu (no hot resisters).

Best Regards

picclock

[nrml]

Thanks for pointing out the issues with that one. It will make it easier for me to save my pennies and do a bit more research before taking the plunge. I've been sorely tempted by it ever since the advertising catalogue came in the post.

[picclock]

@nmrl
If you want to compare different styles of repraps there is an excellent page with over 50 links to designs and variations here :-

http://reprap.org/wiki/RepRap_Options#Models

There is even one with a welded steel frame that someone here was interested in called an eventorbot.

Best Regards

picclock

[PeterE]

Throwing myself into this thread ...

This is an interesting read! I have also started to look for a potential 3D printer for my other hobby (Model Railroading in 1:87 -H0 scale). A fair bit of items lend themselves to be made using this kind of printer.

I have found that the setup I am most appealed by is closest to a Prusa i3 Rework version with a plywood box frame.

The box frame is very stable in Y-direction and fairly stable in X-direction. This can be seen from this image where a plate is used:
http://reprap.org/wiki/File:Prusa_Box.jpg

The original Prusa up till i2 (iteration 2) was set up like a triangular frame in which the moving axes are fitted:
http://www.emakershop.com/browse/listing?l=717
This frame is more stable in X than in Y direction.

What I would like to do is to make a Prusa i3 Reworked (latest version) using a plywood box frame and brace it with diagonal bars front and back. I think that would make for a very stable setup in all directions. Then, if the whole thing is secured to a base plate of something like 25 mm ply, that would be as stable as it can get and still be possible to make in a modest home shop.

In all other parts the base i3 stuff works nicely as a start.

What are your opinions on these thoughts?

BR

/Peter

[picclock]

Hi PeterF

These are my thoughts on the Prusa i3 so far.

For me plywood is not stiff enough, nor dimensionally stable (humidity and heat are the worst offenders), hence my choice of a metal frame. The weak point in the i3 mechanical design is maintaining the right angle/positioning between the X/Z axis and the Y axis. I have a 600mm square granite kitchen worktop which I will use as a base. By mounting the Y frame directly to the base and using side supports for the gantry, also mounted to the base (as per the plywood version),  the stiffness and precision in all three axis will be maximised. 

By using a direct filament feed to the hot end the loads on the x axis are varied with the need to unwind filament from the spool, additionally the mass of the x axis is increased by having a stepper motor attached. So my choice will be to use a bowden feeder system. This may be iffy for very flexible filaments but for Nylon/abs/polycarbonate it will give better results.

The X axis tensioner is either not implemented or pulls against the vertical smooth rod. I will alter this to place the tensioning load on the two horizontal X axis parallel rods. 

Will initially assemble mine with M5 stainless studding for Z axis and test the results. If accuracy is nbg will replace with better system. Am also not happy with running two stepper motors in parallel from one driver chip so may well make an adapter to use two driver boards, one for each z motor.

Any mechanical mods I make will be probably machined out of aluminium to preserve accuracy - plastic has too high a temperature coefficient for precision stuff.

Future thoughts :-
Stepper motors, the Ramps board uses A4988 chips which are rated up to 35V. Unlike most things, turning up the voltage on a stepper driver results in a decrease in average current consumption and allows a much higher stepper speed - so this is something I may investigate. Likewise an increase in heated bed voltage will allow speedier warmup. Much better Ramps Heater FET needed for this trick. Obviously, plus side of heater bed just goes to higher voltage supply.

Hope to order my kit in the next week or so 

Best Regards

picclock

[PeterE]

Hi picclock,

Thanks for feedback! Much appreciated!

About the plywood. The box version of the i3 suggests 12 mm furniture (multilayer) ply and that I agree with, but I would double the thickness on the upright frame for both stability and better meat for screws.

Also, to improve the X-Z vs Y stiffness, I will add a diagonal bar from the table surface to the top of the frame. This of course requires a sort of table surface and I have a suitable old piece of 40 mm chipboard countertop. Cutting that in suitable dimensions and then sealing the edges will make for a good enough humidity-resistant base for the whole thing to begin with.

I think it will be sufficent to stabilize the printer and maintain accuracy in print.

The vertical Z axis studding will use stainless for me too. 5mm seems to be standard so that it will be. No ball-screws yet. I will still keep the combined drive for two stepper motors from one output to begin with.

Then when it comes to filament feed and other things, they will be dealt with as the build goes and usage experience tells. Your future thoughts sounds very interesting so I will keep looking in to this thread.

A friend is currently printing the plastic parts on his machine as we speak.

BR

/Peter

[picclock]

Hi PeterE

After some further investigation yesterday, I have decided to build from scratch rather that a kit. The kits are more expensive and still don't have the parts or support the configuration I would prefer to use. So I have started to order parts from ebay and will do a build thread in the project logs when I get started.

One of the changes I think is fairly important for a good machine is to change the stepper driver chip to the DRV8825 flavour. I have designed pcb's with the a4988 and I know from experience how ratty they can be at 1/16th step. The DRV chip will microstep to 1/32, but more importantly is a smoother driver, with a higher current rating. I ordered 6 assembled, Ramps compatible boards from ebay at £1.32 each last night.

Does your plastic printing friend take orders ? A lot of the frame parts/bearing holders I will make from Aluminium, but there are some bits like the extruder and the motor holders I will struggle with. I can get a complete set for £36 but will probably not need half of them.

Good luck with your printer project. It sounds like you have the right ideas re stiffness.

Best Regards

picclock

[PeterE]

Your suggestion on using DRV8825 stepper drivers sounds like a good choice. I will though stick with the currently common a4988 ones at the moment as I have a fairly steep learning curve to do before getting fluant here.

What was the name of the RAMPS compatible boards, they sound as if they are good valy?

I can check with my friend if he can provide some prints, but don't wait as this is done within the company I work for if you know what I mean.

Sounds like I too should make some photos too when I get started.

BR

/Peter