PeterE builds a 3d Printer

[PeterE]

OK, so let's start with this one. It will be a slow build as it has to be fitted into the too small time for hobbies and there are more projects than this.

Anyhow, I found this useful left-over piece of counter-topafter we updated our washing room the other year.

It has a black laminate topping and is ready laminated at top, back side and the right edge. The left edge have the cut edge which is still not sealed by laminate strip - but it will be. It is about 40mm thick which I think should be sturdy enough for this little printer.

All the box frame pieces have been sawn out from a piece of 12mm furniture plywood (7-layer type). I think you can also see that I have doubled the thickness of the fron  by gluing a second piece of ply to it. I want good thickness for the screws securing the Z and X travel parts..


And finally a trial assembly to see what needs to be adjusted and to make sure that the front of the box frame is absolutely perpendicular to the base.

I find it an advantage to make trial assemblies to get a hint on where minor adjustments are necessary and then only have to glue and screw once.  The marking tape with a line shows the centerline of the baseboard as a help to get the fram front in right angle to the base in X.

Will add to this when suitable bits arrive.

/Peter

[bertie_bassett]

looks like a solid start

[nrml]

I will be following this thread keenly. I have just finished building a prusa i3 from a kit (still yet to be calibrated and commissioned). I am not very impressed with the rigidity of the frame or the amount of slop between the linear bearings and smooth rods. Yours looks like it will be a lot more rigid. Are you going to use smooth rods or vee slot extrusions?

[PeterE]

Thanks bertie_b, we will see where it goes.

nrml, the reason for chosing the box frame was that I was uncertain about the rigidity between the Y-axis motion and the rest in the plate version. Only a couple of nuts securing the upright plate to the table motion seems optimistic to me. The box frame has a rigidity built-in so to say and the Y-axis motion is secured very differently as well.

I will start using smooth rods. Have found drawn stainless rounds to h9 tolerance which is the same as for the ground ones. These should have quite little sloppiness with the linear bearings. I tried one on and couldn't feel any slop so far. We will see where it goes. There are other linear motion solutions to look at if this does not work.

BR

/Peter

[Jonny]

Look forward to this one Peter, keep a tab on the costs.

[PeterE]

Will do.

[nrml]

The problem I found with my kit was that the combination of the poor tolerances of cheap LM8UU bearings and the use of just 3 bearings for the bed plate mean that the bed is not well supported on the smooth rods. I bought replacement bearings and these improved things slightly, but even a fraction of millimetre increase in tolerance translates to a significant wobble once the bed is fully assembled. There was very little slop when running the replacement bearing by itself on the rod, but it can rock side to side a bit after bed assembly.

The X and Z axis haven't been a problem because the use of two closely spaced bearings on each rod provides a much greater area of contact. If you haven't already considered it, using two sets of bearings together or the long version of the linear bearings on the Y axis might save you a lot of frustration . Again, if you haven't already considered it, using 10mm smooth rods  and trapezoidal leadscrews (dirt cheap on ebay) would be a big improvement over the average prusa i3 kit.

[PeterE]

nrml,

Hmmm, see what you mean. The bed on the Y motion only bears on three points and the lonely point is of course on one rail.

But I think I have seen someone make another type of bearing holder for the lonley side which is longer and have two bearings. That will make it work more like on the X and Z motions and then probably be stiffer and less prone to wobble.

I must surf around and see where I saw that. I'll be back.

/Peter

[nrml]

Within the bearing, the area of contact between the balls and the rod is miniscule. Poor quality bearings and unevenly ground areas on the rods make the contact area even less. I think the 4 bearing kit supplier (if I remember correctly) is a spanish company selling steel framed Prusa i3 kits.

[PeterE]

Yep, it's a Spanish Co. I got all the bearings and the toothed belts and gears in one go at a fairly good price. Cannot complain about that.

When assembling I have heard that the linear bearings are sensitive to forces trying to flatten them and that may be a reason for them to get out of order. I have been thinking about using a piece of round bar stuck into the bearing while assembling. That should limit the ovaling forces. But yet again, I might very well make a couple of aluminium bearing holders instead. That will totally avoid risk for out of round. If I do that I will also look into the possibility to have two LM8 bearings close together instead of one lonely at one side. That should also stiffen things up.

But we will see, I'll start with the basic originals to begin with.

/Peter

[PeterE]

OK, the test fit turned out good and now all things are square in all desired directions.  The box frame is therefore glued up properly and after some filling and sanding the first of a suitable no of coats of paint is on.  (Sorry for the crappy cut-out but I didn't want to scare you with the mess taking the view away from the printer frame.)

The red shop colour strikes again - and looks nice with the dark grey (gray?) baseplate.

More pictures coming as new parts are added.

/Peter

[picclock]

Hi PeterE

Good to see your project log. I think you started as I went on holiday. Will start mine soon once the workshop is sorted (a week or two). The graphic display is very good although I could not use software to adjust the contrast (adjusts fine with the pot on the bottom left). I think we will end up with very different machines so the comparison/similarities will be interesting. There are a number of mechanical points about this design which I do not understand, and quite frankly make no sense to me, so my version will be quite different in some ways (not necessarily better - though hopefully so). 

On a slightly different topic I am building mine with a view to using the octopi wireless interface ( http://octoprint.org/ ), so that print jobs can be started, paused, and visually monitored from any web enabled device. Might be worth a look if you have not already seen it.

Good luck with your assembly, watching with interest.

Best Regards

picclock

[jcs0001]

PeterE

I've not had anything to do with 3 d printing but am interested.  Will be following this thread to see how it goes - thanks for posting.

John.

[PeterE]

picclock,
Good to hear that you will also start a build thread, that will give a good overview over different ways forward using different techniques and other choices.  As this is my first 3D build I will follow the standard description quite closely to learn.

John/jcs0001,
Hopefully everything will go as I plan - but what ever does - so I hope this thread shows both good things and things that tend to go wrong. This is also my first try in the 3D printing area, and as I said in the first entry; this will take time as I build as I go and quite a lot of stuff is still on its way in.

Currently I am (eagerly) waiting for the printed parts. 

BR

/Peter

[nrml]

If you are ever stuck needing a spare or replacement part printed, let me know. I'll print one off and send it by post.

[PeterE]

nrml,

Thank you very much for your kind offer, but I found out that there are at least two persons at my job with their own machines and one of them is printing my parts right now. If anyone of them refuses I will take your kind offer.

BR

/Peter

[PeterE]

Time for an update.

The ordered parts are slowly arriving and the bits that stop me are the printed ones. My workmate swamped by jobs so I just have to give him some slack but I know the bits aren't that far away now.

The parts ordered via ebay have basically arrived so let's have look at the goodies:
The first picture shows the paper template I intend to use as stencil for putting a name on the printer, thought that would be kind of fun.


Next "bag" contained the electronics. An Arduino Mega, a shield card for stepper drivers, the small display, five stepper drivers of the most common type and, cables and connectors to join them all. All at a bargain price as mentioned above.


Next is a kit of mechanical parts, that is bearing of different sorts, Flex couplings, toothed belt with gears and a specially treated M8x60 machine screw used as filament feeder.


Stepper motors are a must and even a bit expensive, these came to me from Italy . The shipment also included cables but I missed adding them to the photo.


Another parcel from China contained the extruder nozzle with heater and thermo resistor. The fan was also included  as you can see.


the heatbed looks like below. It can be used for both 12 and 24 V DC. It will be mounted beneath a mineral glass plate that will be the actual build bed. Size is about 200x200mm.


last week the three end switches used to mark the end of travel for each of the axis.


The last picture for this time shows the threaded bars and smooth rods I will be using to put the movements together. From back to front. 2x1m M10 stainless threaded rod, then a 1m M5 stainless for Z-movement, and finally a bunch of stainless smooth rods 8mm dia. These smooth rods keep th h9 tolerance so should work well with the bearings.


Still to arrive are the printed parts as said, together with a large graphic display to replace the smaller one - I missed the difference from the beginning but the overall cost is OK anyway.

More pictures when the build commence.

I will of course keep track of cost during the build but prefer to wait until finished to get the tally right.

BR

/Peter

[picclock]

Hi Peter

Good to see the build log continues. Do you have the part number or info on your motors? I found it difficult to get ones with optimal characteristics, and in the end settled for nema17  LDO 42sth38 1684AC, (2.8V, 1.68A. 1.65 Ohms, 3.6Kg.cm torque - spec attached)  which I purchased from SemiU in the UK. They are pretty std 5% 200 step jobs. The voltage of the motor does not really matter as its the current which drives, and the faster you can change the winding current the faster the motor speed, hence my intention to drive these from a 24V source (its all about the inductance ).

Saw a video on youtube about a printer which layed carbon fiber yarn in with the plastic (nylon in this case) to massively improve mechanical characteristics. It sounds a lot better than than using filament with carbon fibre particles embedded.
link:

  (no affiliation)

Will be very interested to hear about any snags or issues you encounter.

Hope to start my build log soon.

Best Regards

picclock

[PeterE]

Hi picclock,

Yes, the build log continues, but slowly. The set of stepper motors I bought are these http://www.ebay.com/itm/301698572024?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT (scroll down to the specs.)

This is the first time I am using stepper motors so have to guess a lot (to say the least), anyway they are here now so these will be the ones that goes on. If I find them too weak I have to get other ones, but the specs looked OK as far as I could judge.

I will take smallish steps through the build and include photos on the way, and of course add things i find as awkward or odd or anything that don't feel right, as well as those things that goes well.

That video was interesting. Something to keep in ind for more complex bits - if I get that far.

Waiting for your build log as well ;-)

BR

/Peter

[picclock]

Hi Peter

I think the motors are all but identical. I am assuming 360mN.m is 0.36Kgs force at 1mm, which equates to 3.6Kg.cms. If so they should be fine. The amount a stepper deviates from its controlled position is caused by the lamination tolerance plus the load you put on the motor. Lamination tolerance is unchangeable, however by minimising load you should be able to get close to that figure. The 3.6kg.cms figure is the point at which the motor rotor will have moved so much that it will likely step to the next available position. Choosing a motor for optimum performance is a tricky thing. On the one hand you want loads of torque so that you can maintain positional accuracy, but that comes at the cost of much higher rotor inertia which limits acceleration. So at best is an educated guess. I think the Y axis is the one to watch out for, as this will accelerate the highest weight load.

I weighed the assembled x axis filament feeder with motor and attached head/fan (minus bearings, carriage and rods) and it came out to 480 gms (picture), which is a high load for the X axis. I intend to use a bowden tube setup to minimise this load, but its something to bear in mind as this will hopefully be traversing the x axis at a fair rate and may cause frame issues.

Best Regards

picclock

[awemawson]

It's a bit of a pain panning around that huge image Picclock - any chance that you could reduced it to the 640 x 480 norm 

[PeterE]

Hi picclock,

Thanks for verifying the motors, it seems I have selected something useful.

About the weights/loads. I am with you on those issues that is why I doubled the thickness of the portal board in my build. It makes it handle loads better and probably also counteract some of the vibrations together with the thick base. That will, if I think correctly, improve printer accuracy as a whole.

I also intend to look into possibilities of lightening the carrier board (200x200x12mm 7-layer plywood) as the glass plate plus the heatbed adds to the weight but also to stability if I get it right.

/Peter

[picclock]

@ awemason
Sorry bout picture size - I did not check.

@PeterE
Trouble with the stepper offset from intended value is that it is not linear, e.g, for a small load initially a large error occurs then as the load increases, error deviation is less per unit.  A complete can of worms .. . 

I wish you every success with your build - looking forward to seeing your first printed plastic. Hope to catch up with your log very soon.

At the moment I'm trying top figure out how to make flexible filament less temperature dependent (probably not possible).

Best Regards

picclock

 

[PeterE]

A small update before some work related travel must be done.

I cut out the template to paint the name onto the printer portal using a spray can. I did not succeed very well as the spray lifted the paper template at places and smeared the color, sigh. I think it shows despite having replaced the template for next go.

I did not want to redo all the painting so with the template back onto the portal I used the traditional method of dabbing color onto the template.

A small piece of washing spunge was used for dabbing and now it worked better - should have done that from start really. This time the missed places was filled with black and the misplaced black did not get worse.

When back after travel I will sit down with a smallish brush and fill in the bits that are supposed to be red and that way touch up the whole thing. I hope it will turn out OK. We'll see.

The printed parts have turned up! They are made in ABS as that seems to be the best material for structural parts like these. Looking more closely at the bits we can see some flash and that comes from wanting to create a larger surface on the build plate to secure the part better and to get a better precision.


I have started to tear the flash away, but to get it really nice I need to use both files and a knife. I want nice smooth bits to work with. As soon as all bits are cleaned up, the assembly will begin.

I'll be back!

/Peter

[picclock]

Hi Peter E

Good to see that your making progress, and have received your plastic parts. I weighed the Metal Y base and it came in at 302gms bare, the heater board at 164gms.  With 3mm glass at another 300 gms + bearings looks like an all up weight of around 800 gms. To which the weight of the plastic must be added. Let me know how your build weights work out. This is quite a mass to move about in an accurate timely fashion using a thin (stretchy ?) belt and relatively low power motor.

I am still completing other work after the house move, but hope to begin assembly on mine soon.

Best Regards

picclock