Vlad Tepes

From RepRap
Revision as of 03:22, 13 May 2014 by NoobMan (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Crystal Clear action run.png

Release status: Experimental

Tepes v03.jpg
Rigidity Obsessed assembly
CAD Models
External Link

About Vlad Tepes

Vlad tepes is a historic figure from my country, sitting behind what is known as "Count Dracula". Weird how that happened. The public image about him couldnt be more wrong than that. This short introduction is, ironically, meant to do a little justice to a man who made no compromise about justice.

On short, he lived during very dark times. As time habits were, he was raised at ottoman's court as an "lever" of his father behavior on the throne, e.g. if his father would misbehave then he would of been executed. He came to throne in a context where there was no law, but corruption and criminality were ruling instead. Most of sources blame him for cruelty and executing his enemies, things which arent quite accurate either. He intended to eradicate criminality, so he executed the criminals, at start mostly by sword, like he saw in childhood, but it seemed to be a never-ending job. The choice to impale the criminals instead was meant to leave them there for public awareness, and probably came from the suggestion of one of his aids. The method proved extremely effective and in short time after, criminality was in fact eradicated. He took measures to strengthen the economy and help the working peoples. Among these, he imposed restrictions for traders coming with products from other places (like customs taxes do nowadays). For this he became hated and the outside traders actively plotted to kill him for years. If there is anything that can be said about his rule, is that he made no compromise, in justice or anything else. And who lived long enough knows that making no compromise is not a little thing. Sadly, its what sealed its fate in the end. As a ruler he was loved by the ordinary working peoples which gathered to form army to his aid more than once. His grave is not known - if would be then i suppose i would of researched genetics instead, for a way to clone the guy.

Legend says, one old man made a fountain aside a road for travelers to have fresh water. The recipient used for drinking was supposedly made of gold. One day the man goes to the fountain and the recipient is missing. He then said "Vlad is no longer on his throne".

Little Vlad Tepes knew that few centuries later the world would remember him as a guy obsessed with sucking blood out of his harem of sexy virgins ... (yummmm to that).

This is just one of the many general misconceptions i can see from where i am standing. This also should remind us that general judgments should not be passed about a person, unless we know hard facts, or we actually have been in that person's place for long enough to understand where he comes from. This is obviously not related to reprap, but it is a good example of what kind of misconceptions we can fall prey to. And at which degree it can get to.

What's really significant here, is this becoming a world-wide-knowledge in spite of being *that* far fetched from reality.

So better start thinking who knows what else we have got wrong.

Initial point of view and considerations

- This started as for Vert-X contest, but in the end i feel this is pretty much for myself and not primarily for the contest itself. So, in advance apologies if this doesnt make it in time to meet the deadline.

- Vertical axis setup is considered, with the intent of bringing something extra, in terms of features, versatility etc. In this context having a dual axis for a second toolhead is a direction to explore. Mostly because its so easy to do, e.g. rods will be in a square config, 2 at each side of the Z drive. Having a separate axis could let us have a second toolhead without increasing the weight of the first. There are alot of considrents to it, e.g. there is no current electronics and firmware which could eventually deal with a secondary independent axis. How the second axis could be justified is a question mark, but i am sure it can be put at some use in the end.

- For a number of reasons, i am obsessed with rigidity and the 8mm rods are not inspiring me confidence for the job, so 12mm is the choice, smooth rods + linear bearings. With bigger components and second axis, i think a bigger frame would be in order, at least makes more sense in the context. It probably should use 12mm for frame bars and vertices aswell, but for now will pass on that as it feels kinda much.

- Plastic economy and in general any kind of economic reasons are outside the scope of this.

Initial frame tweaks

1) Rigidity obsession pack: frame "patching" with rod clamps, from place to place, in different positions. For 8mm frame.

2) Accommodation for 12 mm Z smooth rod.

Angle Bars dia. Origins Comments File1 File2
90 deg 8 + 8 Mendel mendel y bar clamp M8 to M8 y-bar-clamp.stl openscad file
spacer ~ Mendel repurposed to fit the above, as a 5mm spacer between some of the bottom rods circuit-board-spacer_5mm.stl openscad file
90 deg 8 + 12 Massive Mendel y bar clamp M8 to M12; two files (~lower and ~upper) y-bar-clamp ~ see here
Zrod clamp 12mm Zrod Mendel repository old z rod clamp; tweak: to fit M12 Z motor mount; 16mm one for z-motor; 2x20mm ones for underneath clamp. z-rod-top-clamp ~M12 ~16/20 openscad file
Zrod clamp 12mm Zrod Mendel repository old z rod clamp; as before but for M8 rod z-rod-top-clamp ~M8 ~16/20 openscad file
60 deg 8 + 8 none clamp M8 at 60 deg  ? identical
45 deg 8 + 8 none clamp M8 at 45 deg  ? identical
30 deg 8 + 8 none clamp M8 at 60 deg  ? identical

The above files are grouped here: Tepes_frame_pack.zip

Notes on the initial parts:

- the M8-angle pieces are to add reinforcements from place to place, creatively, i added some wherever i could, and then some more, and then some more again, philosophy being simple: when in doubt, add more M8 threaded bars ...  :-)

Design data

Posting some of the design data maybe it will prove of some help. For the nominals i printed sets of geometry pieces in small increments and selected the best fits against real parts i had on hand. However some parts i printed with 0.4 nozzle and some with 0.7 nozzle, with 0.56, 0.4, 0.2 layers, mostly with extreme settings, so these should be regarded suspicious but probably can make a starting point.

Nominal sizes:

M size Type and print direction Relevant data Description & Notes
M22 cylinder horizontal half r=11.6 half horizontal cylinder for LME12UU cage
M17 cylinder horizontal r=9 z actuator rod bearing fix
M13 cylinder horizontal r=? x idler bearing fix
M12 cylinder vertical r=6.3 for half of diameter, on z block (zb12~pieces)
M12 cylinder horizontal r=6.0-6.1 for half of diameter, on clamp piece
M8 hexagon vertical r=8 for nut trap (Prusa similar size = 8.2-8.5)
M6 cylinder horizontal r=3.5 z-block M6 passthrough
M6 hexagon vertical r=6 for nut trap
M5 cylinder vertical r=3.0 M5 thread impaling
M5 cylinder horizontal teardrop r1=2.75 r2=2 horizontal teardrop for M5 (lower circle r1, upper circle r2); r1=2.70 makes tight fit
M5 hexagon vertical r=5.0-5.2 for nut trap
M4 cylinder vertical r=2.3 M4 thread impaling
M4 cylinder horizontal teardrop r1=2.2 r2=1-1.5 horizontal teardrop for M4 (lower circle r1, upper circle r2)
M4 hexagon vertical r=4.3-4.4 for nut trap


- horizontal cylinders teardrop: main cylinder(r1) + teardrop peak point at half of r1 outside circle, peak flattened by circle arc tangent to both teardrop sides; result is the "classical" teardrop with the peak flattened by the smaller circle(r2) and teardrop sides are tangent to both circles.

- hexagon vertical for nut trap to be made with upper side (bridge area) raised with 0.5-1mm, or to the upper tangent point on circle hexagon; r value is for the outer circle, coincides with hexagon side length.

Small distances for printing - table for quick check and reference purposes:

Layer height perim 1.8w/t loop 1.5w/t perim+1L perim+2L perim+3L perim+4L 2perim no loop 2perim+1L 2perim+2L 2perim+3L 2perim+4L
0.56 1.01 0.84 1.85 2.69 3.53 4.37 2.02 2.86 3.7 4.54 5.38
0.4 0.72 0.6 1.32 1.92 2.52 3.12 1.44 2.04 2.64 3.24 3.84
0.2 0.36 0.3 0.66 0.96 1.26 1.56 0.72 1.02 1.32 1.62 1.92


- preferred print settings, for rigidity purpose, notably include the following (SF): Fill: extra layers 4+ (each), solid surface layers 6+

- 2 perimeters data are for e.g. cases of a vertical hole close to the outside perimeter, which is the generic issue like

- bold values have been considered in some parts; the 0.56 layer height ofc needs a >= 0.7 nozzle

- although the table was there to check, i actually printed mostly with 0.56 & 0.5 layer heights and both W/T equal to each other to ~1.8 & 2.0; and didnt got best results either in some places with some small dimensions issues - who would of figured;

Z axis block assembly

Fixes the rectangle placement configuration of smooth rods. I thought of a chain is as strong as the weakest link. In our case the weakest link is made out of plastic and feels like its this piece in particular (e.g. between x assembly and frame). Also this piece is the cornerstone of a x axis redesign. Impacts on all other parts, motor, belts, carriage, extruder, etc.

Key points:

- meant to fit 4 bars to form 2 independent vertical axis

- could also allow fitting only 2 bars to get a horizontal (classical) axis; even at different heights;

- preferably making it in a single block, (if feels reasonable enough); if not, 2 parts to clamp z drive from either side;

- 4 rods and 2 carriages its alot of extra weight; so this piece needs alot of extra strength compared to old one;

- threaded rod impaled: these get right through the entire part from one side to the other, with nuts on each side;

- distance from the Z rod and Z motor centers - unless fits exactly might be adjustable (for the best);

** VRML - *.wrl : Avoid if unfamiliar. Can get buggy. For Cosmo browser plugin change from defaults in Graphics -> Automatic render change to Direct3D renderer

Set Size Short comment Description STL VRML2.0
0.0 100x70x50mm z-block This part was very guilty. Got impaled lots of times M4. Also too phat, 70mm horizontal. Abandoned. Media:Zb70x70.stl wrl
0.1 100x50x50mm z-block Same as guilty. Actuator thread of z is M6 (not classical M8). Spikes M4. Distance between Zrod and Zmotor axes 30mm & fixed. A way to adjust this would be nice. Part is still pretty big, 50mm horizontal but printable in pla. Added 'mouse ears' in corners for printing. Abandoned. Zb01.jpg wrl
0.1 41x37.5x16.0mm lm12 trap LME12UU trap to fit the above. This one can make the outside trap for version 0.2, because it has a tougher bottom and can resist better when tightening the nuts. Zb02lm12trap16.jpg wrl
0.1 27x40x14mm rod12 trap M12 rod trap to fit the above. Abandoned. Zb02rod12trap.jpg wrl
0.2 105x50x34mm z-block Had a better advocate - skipped some impaling. Actuator rod of z still M6. LME cage to be accomplished separately. Further slightly shorter version, 34mm height, ~6mm lost in rod clamp space, and rod clamps - M5 threads, bearing clamps - M4 threads still. This revision might be better regarding warping. Added rectangle reinforcements each sides for printing. Zb02.jpg wrl
0.2 41x37.5x13.2mm lm12 trap LME12UU trap to fit the above. From theoretical distance of 13, has +0.2 bottom so all distances fit a par number of 0.4mm layers. Two pcs traps the bearing in the middle. From edge to bearing slot distance to give ~30.2mm between actuator thread and smooth rod centers. No extra on top, the part height finishes right in the middle of M12 smooth rod. This paired with the set 0.1 one (better wall to the nuts side) should fit ok if printer is well calibrated in Z axis. Has to be tried on first with the screws tightened up properly. Only if these are too tight, there is a version with an extra layer. Zb02lm12trap16.jpg wrl
0.2 41x37.5x14.5mm lm12 trap LME12UU trap, as above, except +0.4mm higher top side, 1 extra layer, so it overpasses the M12 rod center with that distance. Also the bottom is +0.8mm thicker so it has 2 more bottom layers which would theoretically offset the distance between rod centers accordingly. Added for practical issues, as in case of misfits seems easier to sand the top surface down, than it would be to add something to that face. For example can use one 14.5 and one 13.2 and then there is one extra 0.4 layer in between. as above -
0.2 35x34x14mm rod12 trap M12 rod trap to fit the above. Zb02rod12trap.jpg wrl
0.2 56x76x18mm Z motor Remake of the z motor mount for 12mm z smooth rod and have the motor position slightly adjustable; rod clamps with the old mendel clamp tweaked at 16mm or 20mm. Zb03.jpg wrl
0.2 56x76x18mm Z motor Identical part but for normal M8 rod, as alternative for a standard Prusa to have z motor slightly adjustable too. as above -
0.2 42.4x48x13.5mm Z rod fix Lower side of Z actuator rod fixing inside a 606 bearing meant to support the entire Z+2X heavy stage. So the motors wont have to do that. Now that i can start to feel heavy entire Z+X stage is, i think Z actuators clearly should of been M8. Tried to incorporate one too many changes. I'll see if M6 is enough, but atm is the prime rank doubt. Zb02zm606.jpg wrl

Zb02 files in the archive here: Tepes_1ZB.zip

Notes & ideas for further changes:

- added 2 types bushings in zip as alternative replacement for Z bearings, but both are untested; holes need to be rectified by gripping a rod piece in a bore machine and rotating while the bushing is hold in place, 5 mins or so on cold setup; or by using a hot rod; note the vertical bushing seems to need a bigger inner passage.

- change zblock holes to SMA12 so these could be used instead;

- reverse Z blocks and add an independent nut trap part to any point on the smooth x rods;

- do a *dual* vertical x-stage with M8 smooth rods for classic size mendel/prusa?

- make zblock as a sandwich of 4 parts with bearings in the middle. Probably most "printable" setup. Probable disadvantage might be that tightening the rods might result in unwanted over-tightening of the linear bearings, so has to be tested;

X axis assembly

X motor mounts, idler and carriage are pairs of two mirrored individual components, so they clamp the vertical x bars from each side (1-back, 2-front). Belt goes right inside between the vertical smooth rods, and does 1-fold over 4 608 bearings on carriage, like in Madkite double resolution mod. Both motor mount parts and idler parts fix the belt on the lower side with 2xM3 which screw onto a small alu plate with threads in it. Also both have a set of 4xM3 holes on their top side, intended to fit some optoendstops (they dont match standard endstop tho coz not enough space). Carriage is a special issue. Atm made one out of abs (same as all x parts), and will see how it behave. But i think for the best it shall be done with SMA bearings onto alu plate with a good planar face, and no plastic parts (same as Y axis), because these will work within inner vertices (area which will become a chamber at some point). This way extruder design can be simplified down to a direct driven nema 23, simply mounted on a metal corner piece to carriage plate. Hot end all metal to be easily separated or to change nozzles.

T5 belt data:

Belt path Minimum pulley teeth Minimum idler dia
no back-bending 10 teeth 30 mm
with back-bending 15 teeth 30 mm

Design can use pulleys 10-12-14, and the end idler is made with two 608 bearings with a diameter of 22mm. Default belt path does not respect the above minimal data (not enough space), so belt flexibility and effective tensions will supposedly get affected. The 10mm width belt is stiffer than the 5mm one, but on the other hand might also be more resistant / durable, so who knows what the belt life will be. Along these lines tougher plastic parts have to be tougher than usual. If needed there can be made some alternatives to X belt path, so the "accuracy mod" can be dropped. One alternative is to drop the carriage idlers (belt returners), and use an additional part to fix belt to carriage. The other alternative is to also drop the smaller bearings idlers (belt deflectors), in which case the end idler hole needs to be repositioned (upwards - in center) and the used bearing has to be approx same with inner diameter of the pulley.

Revision Size Short comment Description jpg VRML2.0
0.0 105x72.2x13.9mm x_mount Nema23 X motor mount. For 10 teeth pulley but 12 teeth (perhaps 14) should fit aswell. My N23 were little atypical needing M4 screws instead of M5. Xv00XMOTN23.jpg wrl
0.0 100x55.8x13.9mm x_mount Nema17 X motor mount clamps. Same as above, bit smaller. Xv00XMOTN17.jpg wrl
0.0 100x~40x13.9mm x_idler End idler clamps. Default is to use two 608 bearings with 22mm dia, which means their hole is lower than motor center, as it just has to align on top side. Xv00XIDL.jpg wrl
0.0 ~116x90x13.9mm x_car Carriage, as "plate face" with provisional holes, so it could mount whatever. Hex nut traps have overhangs. Xv00XSMAplate01.jpg wrl
0.0alt 32.4x24.5x28mm x_carfix Alternative belt path. Keeping small deflector idlers, this part fixes the belt to the carriage at a taller point. Overhangs 45deg. Xv00XSMAcarFix.jpg wrl
0.0alt 32.4x16.35x28mm x_carfix Alternative belt path "NBB" (no back bending: simple pulley-carriage-end_idler setup). Small deflector idlers not mounted, so this part fixes the belt to the carriage, but on a lower point as to pulley lower side. Overhangs 45deg. Xv00XSMAcarFix.NBB.jpg wrl
0.0alt 100x34x13.9mm x_idler Alternative end idler clamps, to go with the above belt path "NBB". Unlike default part, this has hole center aligned height with motor center, to be used with bearings of approx size -1mm compared to motor pulley. Though using the default piece with smaller 624 idlers/deflectors should theoretically be better instead this option. Xv00XIDLNBB.jpg wrl
0.0alt 41.5x42x14mm SMAcage LME12 bearing cage with holes as SMA12. Same style as ZB... 2 pieces clamps each side Xv00XSMAcage1.jpg wrl
0.0alt 39x42x34.8mm SMAcage LME12 bearing cage with holes as SMA12. Jar style, with a lid / cap. Vertical print, untested, certainly more sensitive to calibration and printer issues. Xv00XSMAcagevert348.jpg wrl
0.0alt 39x42x28mm SMAcage LME12 bearing cage same as above but slimmer, to keep the faceplate as close as possible to the rods. Still untested and prone to issues. Xv00XSMAcagevert28.jpg wrl

Very lame video of x axis with initial parts: T_2XV_lamevid.mp4

Xv00 files in the archive here: Tepes_2XV.zip

Notes & ideas for further changes:

- change parts to better fit the idlers double bearings spacing(s): elevate plastic near bearings so less nuts will be required;

- make some "beefy" plastic nuts so 5mm belt could be used instead of 10mm;

- make std 20mm & 34mm diameter tool mount; E43mm with 2-3kg tool is probably too much, but worth a shot :-)

Y and metal parts

These are motor mount and idler. These are extra strength and should also fit regular mendel or prusa, since the vertices are the same.

Revision Size Short comment Description jpg VRML2.0
0.0 - mm MOT Y Motor Yd00MOTN23.jpg wrl
0.0 - mm MOT motor upper clamp Yd00IDL clampup.jpg wrl
0.0 - mm MOT motor lower clamp Yd00IDL clampdwn.jpg wrl
0.0 - mm IDL Y Idler Yd00IDLER.jpg wrl
0.0 - mm IDL idler upper clamp Yd00IDL clampup.jpg wrl
0.0 - mm IDL idler lower clamp Yd00IDL clampdwn.jpg wrl

Bed assembly uses SMA12 caged linear bearings and all metal parts. Compared to other printers this kinda feels like "cheating", but still its probably best choice. So these will be able to withstand much higher temperatures than otherwise. The belt double fold (madkite's mod) uses bearings which are affixed onto threaded rods, which in turn are affixed to the bottom plate, everything being done with threads and screws which makes the position of bearing slightly adjustable. The heated bed has to be as light as possible, and as powerful as possible. It is 5mm thick aluminium plate with radiator resistors for heating. The power resistors are thermally insulated underneath. One side has side holes to fix mdf plates with the role of being sacrificial material for pcb milling and drilling.

Yd00 files in the archive here: Tepes_3YD.zip

Toolheads, milling side

Milling fixtures that fit onto x carriage plate.

Revision Size Short comment Description jpg VRML2.0
0.0 20mm dia holder fits proxxon micromot Th 20.JPG wrl
0.0 34mm dia holder should fit other tools, untested Th 34.JPG wrl-
0.0 43mm dia holder standard Euro 43mm mount, for big & heavy milling motors. And crazy users. Th 43.JPG wrl

Double sided milling requires a slot parallel to X, and close to that, a pin with a known position. For best results these operations need to be done by the machine itself, so the parallelism of the slot and the pin position are guaranteed. The video depicts an E43 mount Suhner milling motor of 1Kw and 3Kg, and should be understood that is one-time thingy, not regular.

Very lame video for specially crazy ppls: Tepes_toolhead_crazy.mp4

4Tb00 files in the archive here: Tepes_4TH.zip

Extruder, printing side

All metal (diy) hotend and extruder. The dust from pcb milling is not nice to the extruder, so it is for the best to use the machine for either purpose. After some time milling pcbs, Tepes will stick to printing.


- Choosing overall size, probably should aim at efficient use of the most expensive parts, M12 smooth rods: Z~45cm and Y~55 cm. Then vertices can be made to fit the smooth rods. Smooth rods need to have some few extra 3-4-5 cm margin, to allow rod to go through their clamps at full length. Horizontal vertices spacing can be chosen depending on X smooth rods, e.g. up to 1m+margin (for outside ones) should be workable. The inner vertices spacing need to cover the space for motors and idlers, and the interior space gives the x size of the bed. Exact dimensions are hidden inside one's creativity corner.

- Nuts should be with blocking / plastic ring, otherwise some anti-vibration washers have to be used (not sure the english technical name). Washers are of the biggest available size - wherever there is enough room.

- Added plenty of extra rods with no other purpose than just helping stiffening things up, using rod clamp pieces from "initial frame tweaks". For best placement its a good idea to add these right after finishing all axes mounts.

- Two extra sets of vertices are used, four in total. The horizontal rods pass through these and have an extra 2-3 cm of thread left so in the end some panels could fit here, sort of to box things up. Also some other rods have extra margin for this purpose.

- Squaring things: as gauge i used a long piece of M8 thread onto witch i scotch taped 2 other small parts of same thread type, like 5-6cm long. This way i could use that as either inner gauge or outer gauge. I found this pretty useful. Also what i think is a good tip would be to make the Z smooth rods height enough as to just surpass the motors with a few mm. This way the space between them can be measured directly on the top of Z motors, which makes things fairly easy (perhaps similar for bottom fixture). The Z smooth rods are fairly hard to get right especially because the assembly cant handle much error about it - if any at all. Best to get Z smooth rods perfect before anything else: i measured sides with the thread gauge and a string across diagonals points to what needs to change.

- Z block and assembly: Z blocks are big size prints, might warp a little, heated bed and other warp techniques required. I had a big towel covering the front of the printer when printing that :). The block "feets" are just for printing, have to come off afterwards. Z motor can be adjusted 4~5mm and the bottom part can also slide same, so the two Z actuator rods can be perfectly paralleled to Z smooth rods.

- LME12UU traps for Z: these are rather hard to get right. Parts are extremely sensitive to layer height chosen when slicing and to the way a printer does first layer, top layer, and generally speaking Z calibration. Best way i could figure was to sandpaper the faces just a little and try them on a fitter bearing and see what kind of play they have against the bearing. There should be a play, but rather small, as when tightening the nuts would kind of deform the plastics just a little, and then when fitting 2 of them will take that play out when Z block is all tightened up. However too much sandpaper might ruin the part and possibly render them useless. The trap that has the nuts on it should be the 16mm one, because its bottom is thicker and can support the nuts force easier. Linear bearings are sort of finesse items, if only slightly over-tighten they dont slide properly - this can be felt at hand. Since cant tighten them too much, then also cant rely on thread itself to block the nuts in vibration environment, so this place especially would benefit of thread lock or blocking nuts or similar.

- X setup: most of idlers are made out of two bearings packed together, with washers to fill the gaps. I used all types of washers, larger ones especially M5-M6 etc. Its fairly hard to get all tighten up but they have to be tightened, so things run silent. Carriage has quite a few provisional holes and also the M5 bearing screws can be used to mount things up as i find that not all of them are really necessary. I think its like any married man with experience can put it in a gross manner: "its better to have a hole and no need to use it than the other way around" ^^ :-)

- Alu pulleys: actually these exist in lots of different standards, and its fairly important which one it is. "T"5 as letter says, its a timing belt, so even the base standard has less play than the random given pulley, but nonetheless still does have some play. This play has its own reasons but also can lead to errors / belt specific backlash. Especially when using pulleys with less than minimal number of teeth (8,10,12), these issues couples to aggravate the effect. For positional purposes / applications, usually there are actual variations of the T5 "standard" pulleys, and these ones specifically have tooth gap profile with "reduced play" or "zero play" (or some similar words). So for the best, keep an eye out if there is an option, search for the pulleys with reduced or no play in tooth gap profile. Although if there is no option like that then in the end take anything available. No promoting intended in the following link, except it may have a better explanation: illustration