PeterE builds a 3d Printer

[nrml]

Bed movement does not seem to be an issue if you don't crank the print speed too much. Mine has the aluminium base plate, plywood insulation, MK2 heated bed, glass plate, all the clips/fittings etc and some very poorly fitting linear bearings and I still get pretty decent printouts at slow speeds (30mm/s). Rather than spending lots of time and effort in refining the Prusa, it might be better to put the money and effort into your second build after learning the lessons the first build teaches.

I am collecting the components for my next printer build, which will be an extra rigid all metal corexy printer with no printed components, a 30 X 30 X 30 cm build volume and an enclosed chamber.

[picclock]

@nmrl
The one major advantage of the core xy type design is that it minimises the stepper load as the only the head moves in all axis. Clearly eliminating the issue of moving the printer bed weight and allowing higher speed/better accuracy. However the price to be paid for this is increased mechanical complexity and poorer head/bed access for maintenance/blockages or alterations. For larger machines this is clearly the way to go as the problems of moving weight around increase with size. I will be interested to see how your core xy build progresses, and what advantages accrue. Will you use V slot mechanics? I have no experience of them but it seems to be a better system than smooth rods and linear bearings. Perhaps you could do a build log and keep us all up to date.

Best Regards

picclock     

[nrml]

I intend do a build log once I get started. The reason I decided to go for a corexy mechanism is that it can easily be incorporated into a very rigid printer. Besides, it is a very elegant mechanism. I love the V-slot concept, but I am not sure if wear will be an issue and how rigid a gantry running on wheels can be made without binding being an issue. Also, if you are buying readymade gantry hardware the costs rise very steeply. I am planning on sticking with linear rails and oilite bushes - I have had very bad experiences with cheap linear bearings so far and Igus polymer bearings are a bit too expensive. I am going for 32 bit electronics (duet). It is going to be a challenge for me, as I don't have an engineering background, but I am sure it will be an enjoyable one.

I don't want to sidetrack Peter's thread, so I'll start a separate one in the next week or two once enough stuff arrives for me to get started. They are on the slow boat from China.

[PeterE]

Just a short update. Of course my cameras battery decided to be empty, so no pictures today.   

I fixed the naming to an " OK it is not an altar" level. This printer is for test and learning to begin with. The fixes went OK, but could have been much better.

This means that the portal and carrier for the X and Z axis is finally secured to the base.

All structural plastic parts are cleaned up and just waiting for assembly.

Next up is to cut the threaded and smooth rods for the Y axis motion and begin assembly, and to cut out the ply for the Y carrier. I will use the heatbed as template this first try before introducing any changes.


picclock,

Interesting things you mention about the Y axis table weight. I will weigh my setup once compiled all together, I will use a 12mm ply board instead of your alu plate.  As said earlier, I also look into making a few lightening holes in the ply board as it is mainly used to carry the heatbed and glass plate.


nrml,

Interesting to read about your experiences and the way you will go for the nest build. I notice that you are very suspicious about the cheap linear LM8UU bearings, and smooth rods as a concept. On my part I have had experiences with that type of movements way back in time when I was working with packaging machinery. There was no problem with linear bearings and rods then - but I suppose they were of another quality. Will get back to that later.

When it comes to lubrication I will use the same silicon grease as I have used for my lathe and mill, namely ordinary car wheel bearing grease, good specs and copes well with both heat and speed.

As for the Z motion using M5 threaded rod, I think there is quite a difference in precision between the ordinary galvanized stuff and the stainless ones. The stainless are much nicer in both finish and movement as far as I can tell, and nearly no sloppiness either.

Will be interesting to read your build log once your bits have arrived.

No problem with side-tracking my thread.

BR

/Peter

[PeterE]

OK, so a couple of pictures, whereof one i have to re-take as it got blurred 

The first shows the naming (very blurred) will make a new tomorrow.


The second shows all the printed parts for the Y axis motion. Well not quite, I notice that I lack the four hold-down parts to lock the Y axis to the portal bottom. Those will come shortly.


... and finally, a couple of 8mm threaded rods in two of the corner feet for the Y axis.


I noticed when cutting them to size that stainless rods are actually quite a lot better material from both precision and stiffness point of view compared with ordinary mild steel or galvanized rods.

The small bit on the upper rod is the toothed belt tensioner in its correct position. The motor bracket is fitted in the same way on the other end. Time did not allow for all bits to come together tonight so will have to wait until tomorrow.

/Peter

[PeterE]

Evening All,

Just noticed that one of the pictures was lost, sorry. I had to exchange the old one for one with better quality and now I cannot change the address. The image name is now "3Dprinter-18_1.jpg" and if any of the moderators would be so kind to add the _1 the image will be back.

But that was only one thing to talk about.

The 3D printer project moves slowly but that's how it is. Got some time in the shop tonight to think and cut a few more pieces.

In the first picture it looks much the same as the last one, but I discovered that I should have used a 12mm threaded rod instead of the 10mm one for the long rod. So off to the store and get a length of 12mm stainless threaded rod. It is now cut to length.

It does probably not show that much, but will have a significant effect. A 12mm bar is much stiffer than a 10mm one, especially when being just short of 400mm long. With this the Y axis movement should be quite stable and not want to bend under tension.

I also came to the conclusion that if I make a sort of "nut" that the 12mm rod is threaded through and that the nut itself is screwed down into the base of the portal and into the base plate, that would make quite sure that the y axis is fixed in level and angle to the portal. Two screws in each "nut" should provide a good enough connection. Will be this weeks task to do.

Next image shows the piece of plywood I intend to use as heatbed carrier plate and it is to that one the bearaings and drive belt is attached.

I am also thinking of making some "lightning holes" in the carrier plate, both to make it lighter, but also to make heat transportation easier to lessening heat skewing in the carrier.

In the final picture I just put the heatbed on top of the carrier plate to give a view of how it will look. The heatbed will be held at a distance from the carrier plate by a spring in each corner, That should make any movements from heat possible to handle within reasonable limits.

So, this weeks task will be to make the four special nuts to tie the y axis to the base and portal.

/Peter

[picclock]

Hi Peter

Really good to see your progress. Would like to post some of my own but its my daughters wedding Friday and I'm not allowed to play. Am currently in the process of making a spring loaded idler pulley. Will post pictures when I get a chance.

Keep up the good work.

Best Regards

picclock

[nrml]

@PeterE
Looking good. Are you going to put a layer of aluminium foil between the heated bed and plywood to reflect some of the heat back into the bed?
I have no idea if it actually makes a difference or not but I did it because it cost nothing and made me feel better about reducing heat loss.

@picclock
good luck with the speech. Hope the wedding goes perfectly to plan.

[PeterE]

Good evening, me again,

@picclock; Of course, such family matters have definite precedence  Hope it will be a joyful event!

@nrml; Haven't thought of that really. An idea to look into once assembled and tested. It is an easy one to retrofit. But my thinking is that as there will also be a glass plate clamped onto the heatbed, the glass will ackumulate the heat and keep it stable. Should be enough really. Actually I am more concerned about getting the plywood carrier board to hot so will see if I can add a couple of holes for better airflow and cooling.  Well, we'll see where it goes.

Now onto next advancement. The four small threaded securing blocks are now done. Had a piece of 20x40 mm aluminium from which I cut four 20 mm pieces. Following some filing, marking, drilling and threading, here we are:

The little blocks are about the same size as the printed plastic ones, but mine are threaded using the same M12x1.75 mm thread as the stainless rod. This will limit the need for using the blocks as stop blocks to which the rods are locked using the nuts, but instead acts as positioners and positively locks down the rods to the portal and base at the same time. I think this will make the whole setup more stable.

I discovered during trial assembly that the threaded rods runs more or less right on top of the portal's base surface. So, the blocks are milled down to show the tips of the thread and the rods are correctly positioned.

Assembled it will look something like in the picture below.

Still just trial assy but this week it will be completely and finally assembled after the plywood carrier board is currently getting a bit of paint.

/Peter

[nrml]

I meant foil between the carrier board and the heated bed to keep radiant heat away from the plywood board and reflect it back to the heated bed and glass plate.

[PeterE]

Hi nrml,

Yes, I understood, but just fixing the foil to the plywood carrier will not do the full job I think. It ought to have some sort of temperature insulation/isolation between the board and the foil to be fully effective - or? 

BR

/Peter

[tom osselton]

I wouldn't worry about the heat wood warms up good and it is well below the flashpoint, one of the things affecting it more is cross breezes cooling it down during printing.

[picclock]

@PeterE  NMrl

Thanks for your kind wishes. The wedding went off very well thx, even the speech seemed to go down OK so thats a big thing out of the way.

Peter you seem to have made very rapid progress, and its looking really good.  Tom Osseltons comment is quite valid. I had looked at two strategies to deal with the issue. The first to surround the build area with a wall of cardboard perhaps an inch or so higher than the printed object. The other was to 3D print a throwaway wall around the print being made. May well stick a sheet of cork under the bed heater to improve insulation, though it limits the bed temp to about 120C (more than I will need ATM  )

Great to see your progress and pictures. Will post more on my log when I get a chance. Just picked up a large mirror on Gumtree for £5 which should make at least 6 printer beds (depending on how many I break cutting it   )

Hope to be getting on with the build soon.

Best Regards

picclock

 

[PeterE]

Good Evening folks,

@tom osselton; Thanks for that info. Then I can skip some work with the heatbed carrier. Will keep that in mind.

@picclock; The cardboard fence idea seems interesting, but having it come up higher than the finished piece may be an issue as the cooling fan for the extruder might be below that fence at least in the beginning and then cause a cooling airflow that is unwanted.

Anyway, got some time in the shop this evening and took the opportunity to assemble most of the Y axis bits. I am following Dragonfly's (John Ridley) instructions to make it easy on my part. The instruction is here:  http://www.dragonflydiy.com/2013/10/building-prusa-i3-printer-y-axis.html.

The image shows the assembled and secured mechanics.
.
It is very close to square all over already now by just tightening. It only remains to do the last part when securing the whole thing to the portal and baseplate. This will be done when the Carrier board and the driveline is in place.

So, time for the carrier board and I am using the mesaurements from the instruction. There shall be a bracket for the toothed belt fixed to the carrier. I found that it should be placed so that the bracket will not pass the corner of the stepper motor, and centered between the bearings, see below.


In the other end at the tension point, the bracket must not reach the fitting to avoid un-necessary bending forces.


Now we turn the whole thing onto its "feet" for a try in place in the portal. The carrier board and motor is placed towards the operator,:


... and with the carrier at the other end away from the operator it looks as follows.


As can be seen, I haven't fitted any bearing in the tensioner yet as the ones I originally ordered were too big :-( One of the mishaps in ordering. As soon as I have the correct size I will continue assembly.

/Peter

[PeterE]

Eveni'n

Got a few minutes for a quick addition. Started to assemble the Z movement. The pictures show the right side and the left will be assembled in the same way.

I placed a piece of 3 mm cork (that I normally use for track bed for my H0 model railroad) as a spacing between the base plate and the stepper motor. This gives the right measurement. The black bracket is secured both from "round the corner" and straight into the portal behind the motor.

The vertical smooth rod is placed in its hole with some light press fit. This is necessary for stability. I used a reamer to get the hole size correct.

Then the upper end is held by yet another little bracket fitted in the same way  really.

The picture shows the setup - sorry for the tall picture.

Before going further I have to get a new X-end idler as the printed part was not quite OK. Will get a new one during next week. (This does not hinder anything as I am on a business trip at the time).

TIP! I noticed that the brackets have been designed for countersunk screws. Be careful not to overtighten, the ABS plastic needs careful handling. In my view cheesehead screws would have been a better choice - where possible of course.

BR

/Peter

[PeterE]

Me again,

Got a good day in the shop and made some more trial assemblies of the 3D printer.

First picture shows an overview of the mechanical assy as it ought to look when finalized.

Both the X and Z axis are more or less done, but as can be seen, there is a faulty component in the form of a broken X end idler (to the left). It turned out that my friend didn't find the correct settings so the part was un-necessarily brittle and broke as I aimed to push in the nut for the Z travel control. A quick look at the broken surface show34ed that there was no fusing between the layers so too cold during printing. Otherwise it looks as I expect.

A note of caution here for anyone trying to make a printer of the same style. As I started to put things together I was looking at the material list for info on smooth rods length. It said 350 and 370 mm, BUT, Do measure both the X and the Z distances before cutting! They turned out to be 430 and 470 mm respectively.

Next picture is a close-up of the mechanics.

A quick check showed that the setup is actually surprisingly stable. No back-lash or slop in either bearings or other things. I think this is going to work fine.

Then a closer look at the extruder.

The herringbone gears are also surprisingly slop-free! It feels like it is very well thought through so far.

One thing that I didn't like though is that I had to pack a larger amount of washers behind the screw head to locate the hobbed bolt for the filament feed correctly. There are obviously quite a few versions of feed stuff and I happened to pick the wrong one.

Well, I already found a solution :-) I will cut off the head, thread the end and use a nyloc nut together with Loctite to create a better positioned "screw head" using just one washer appropriately.  That will be tomorrows job.

[picclock]

Hi Peter

Great build thread, and pictures. Good to see how quickly you have got it all together. Thanks for sharing.

Re smooth rod lengths. The info I have is X=370, Y=350,  and Z=320. Y  bearings limit movement by -95mm so overall y is 225mm after allowing 15mm per rod each end.  Length of y rod is set by bearing spacing and required movement. Bearing end to end on my stuff is 95mm. Not sure why x length is so large.. .  Is it that yours are different due to the wooden frame construction ??

Many thanks and keep up the good work

Best Regards

picclock

[PeterE]

Hi picclock,

Thanks, knowing the forums desire for well told projects I am trying my best to write down what I do and find.

The difference between the lengths of the rods is, as you say, due to to the different sized portal. Since the Z rods are positioned by the corners of the portal they will by default have another size. Same with the X rods that then must be cut to fit the Z rods positions.

This got me worries for a while and I wondered if I cut the portal to the wrong size from the beginning, but I didn't. The advantage of this is that I get a greater height for the extruder travel which in turn equal taller objects if I want to.

Next week is business travel week so no shop activities ...

/Peter

[picclock]

Hi Peter

I have been very tempted to up the build platform/size on mine, but have resisted so far. I suspect that an object 8" on a side for me is massive - and if I needed larger two prints joined would do the job. I think if I ever needed larger the cube type that nmrl was talking about may be better as only the head moves, but I can't see myself needing something so large.

I've managed to cut the mirror bed glass for mine now after breaking loads of it  , so I hope to be able to mount the bearings to it soon. 

Bad luck about the travel - but if its like over here, eating out on the company is not too bad.

Best Regards

picclock

[PeterE]

Just a quickie this evening,

The adjusted filament feed bolt now looks as below.

I cut off the hex head and threaded the end. Used an ordinary nut and screwed the whole thing together. After matching the feed knurling with the filament hole, I unwound the nut, put a drop of Loctite on and screwed it back together again. Now waitning for the Loctite to cure completely and then I will tidy the end up flush with the nut so it looks more or less like an ordinary bolt head.  That will save me a bunch of space.

/Peter

[PeterE]

picclock,

Unfortunately I read that you have done all your glass cutting, but I thought that the following could be useful for next time. I hope you don't mind.

I noticed how a proffessional cut a smaller piece of glass without it breaking at the wrong place some time ago when need a piece. He first cut the line and then lifted the pane and put a straight thin rod - think something like an M3 all-thread - right under the cut. Then a small push-down on the waste side and the glass snapped nicely at the cut line. The rod acted like the edge on a folding rig.

/Peter

[awemawson]

Dipping the cutting wheel in a very light oil or even diesel, helps enormously cutting glass. Then rather than resting the cut on something hard, use two used match sticks, one at each end of the cut line, and gently press down either side.

[picclock]

@PeterE 
@Awemason

Thanks for the glass cutting tips. I was using cutting oil (WD40). My best efforts so far have been by tapping the glass on the opposite side of the cut line. You can visually see the crack propagate under the score line. If you propagate the crack too far the piece you are cutting separates and gravity accelerates the needed part towards the scrap bin  .

I will try the bend over a rod/matchsticks trick on some scrap and see if it is more successful. The sheet of mirror glass I am cutting from is quite large, 1.1x0.7 Metres, which makes it more difficult. However it will soon all be in much, much smaller pieces   .

Embarrassingly, my father was a glazier but the magic has not passed on down the gene line.

Best Regards

picclock

PS Just tried the break over rod trick(3mm silver steel), and it works like a charm. Many Thanks 

[nrml]

Looking very nice indeed. Those direct extruders are very user friendly. Almost zero hassle with set up and tuning.

[PeterE]

Back from business trip and into the shop.

@picclock; Good that the trick with the rod works!

@nrml; Good to read that this extruder design is user friendly and easy to tune. Suits a newbie to 3D very well 

Got the screw trimmed to keep the head as compact as possible. Put the screw in the lathe and turned the protruding threaded part down to the nut thus forming a new "head".

Re-assembled it looks as above. Used a flash this time to show the hobbed part of the bolt in line with the feed hole in the extruder body.

Time to test fit the extruder to the X carrier - but it comes very close and I have seen that there is a distance piece in some of the descriptions. A piece I don't have, but it is ever so easy to make :-)

The question pops up; How thick should that distance be? I suspect it should be so thick it gives room for the bolt head but not a lot more to avoid too much twisting forces on the two parallell X way bars.

I suppose plexi would be appropriate as replacement for a printed distance piece?

/Peter