LaserCut Mendel Assembly Blog

From RepRapWiki
Jump to: navigation, search

Photographs of LaserCut Mendel, a laser-cut plywood version of RepRap Mendel by User:Kymberlyaandrus of Tech Zone Communications.

A call to all users of these parts, please post, add, edit, update, whatever you can contribute. We will all be better with more information. Thanks

LIFO order please, newest stuff at the head of the article.

May-24-2010 First print

Wrong 3M Blue long duration tape (60 day). The heat released the adhesive. The unheated bed and no stick = major warpage. Skeinforge's gcode contained all kinds of temperature and extruder speed changes and it jammed during the 1st printing. Hand editing the g and mcodes inside of ReplicatorG, I left it at 230 C for the ABS and watched it go. Configuring Skeinforge has yet to come.

Tried to print a small gear with grub screw and captive nut and the extruder jammed - out of adjustment way too tight. Backed out the ABS and proceeded to drop the gauge rod down into the extruder. After unplugging and shaking it upside down for a bit with no luck. I plugged the motherboard back in and turned on the extruder motor in reverse while pinching the rod with finger pressure and it slowly backed up high enough to pull out. Will try again tomorrow.

Mendel print1 all.jpg

My axis were all wrong, from my point of view. X0 Y0 was back right and that was ok, I could deal with that. The Z+ went down and I found that very difficult, so I changed things. MakerBot RepRap boards turn their own way. My working wiring took a long time. I checked the data sheet at MakerBot and knew the colors of the pairs. The order of things astounded me. I plugged one in backwarrds and it worked perfectly. It took a long time to figure out, but they're just DC motors. nophead commented on re-wiring and that led me to all the changes here.

Y, X and a blurry Z and then all. This is hours. . .

Mendel print1 wire.jpg

Copy Paste this in /Applications/replicatorg-0015-mac/machines.xml Ball park settings only. Extruder tip measurements, XY eyeballed 50 mm movements along ruler, Z feeler gauge at Z0 then climb to pass measured block at 25 and 30 mm (2 tests). This came together over several days of searching and experimentation.

The steppers seem to go nuts and lock upin resonance at higher rates

These will be refined and updated. (formatting is removed here - make it match what is in machines.xml or edit this :)

<machine> <name>RepRap Mendel</name> <geometry type="cartesian">

<axis id="x" length="300" maxfeedrate="5000" scale="9.95"/> <axis id="y" length="300" maxfeedrate="5000" scale="9.95"/> <axis id="z" length="300" maxfeedrate="50" scale="750"/> </geometry> <tools> <tool name="Pinch Wheel Extruder v1.1" type="extruder" material="abs" motor="true" floodcoolant="false" mistcoolant="false" fan="true" valve="false" collet="false" heater="true"/> </tools> <clamps></clamps> <driver name="sanguino3g">

<rate>38400</rate> <parity>8</parity>

<debuglevel>0</debuglevel> </driver> <warmup> </warmup> <cooldown> (Turn off steppers after a build.) M18 </cooldown> </machine>

May-21-2010 lost edit damn. Here's pics. back later. . . weeds. Buy 100 4" zip ties. I ran out and had to use 8" ones which barely fit the 4 mm holes.

I'm done building, 7 days. Now need to learn how to set up Mendel in Skeinforge. 1st print attempt tried to punch hole in acrylic bed, wrong machine.

Geared extruder

Bowden (tube feed) extruder

Extruder/x-carriage fan install

May-20-2010 opto ends stops all day and then re-wiring late

Mendel opto all.jpg

Mendel front.JPG

Mendel back.JPG

Mendel power1.JPG

Mendel power2.JPG

Mendel power3.JPG

Mendel power4.JPG

Mendel psplug all.jpg

Clearance problems. Need feet >3/8" high.

Mendel clearance1.jpg

May-19-2010 [End of day. MakerBot Plastruder extended 5 mm and fitted to X-axis carriage.]

Problems: clearance underneath; extruder heat and clearance; Z-axis belt adjustment bolt tip causing friction.

I got the shipment from Tech Zone 7 days ago, at that time I all ready had working RepRap electronics and extruder from MakerBot, along with fasteners, belts, rod, studding, bearings. I had also hacked together a couple of Mendel frames out of all thread and hand made oak pieces. The only parts that I had to run for were 12 #6 1-32 x 1.25" and lots of nuts and washers for the Frog Plate assembly.

The MakerBot Electronics will not fit under a Mendel, even one that has been lifted an inch. The plugs do not clear the belts. The steppers boards are off towards the Power Supply. Right now this entire Mendel is on modular connections plug or screw connector nothing is soldered. Removing the MakerBot plug structures and hard wiring makes the system very compact. I'll stay plug and screw for now rather than rewiring.

I'm planning on another system (gantry cnc) running the same electronics and yanking off the connectors and hard wiring everything does not make sense yet for me. In the same vein the Z-axis ground rods are not cut because I may use them elsewhere. The 180 bearing rod may be replaced with 3 pre-cut 5/16 smooth galvanized rod freeing up 3 ground rod to be 360 bearing X or Y rods in another Mendel, Adam's rib as it were.

It is all ready to test. The Z-axis was adjusted to be dead center @ 145 mm of 290 mm frame vertex distance, tightened and hot glue 'locker'. All electronics connected and all systems function, no extrusion today. The axis work and power pots are trimmed again as I switched leads. The Z-axis had trouble at the start, now smooth. The extruder drive motor works in the correct directions. The heater and thermistor both work in al low temperature test. Everything worked in previous testing, but the heater barrel assembly has been handled quite a lot, nice to see it still get hot.

Made an acrylic bed that hangs on the Allen headed M4 cap screws and make them recessed. It is held on by clips, which seem too strong for the job.

Followed nophead's idea and jogged the Z-axis down until a 3 mm rod of PLA filament will just pass. Repeat in the other 3 corners. It was very close and needed less than one full turn to get all 4 corners the same height.

All ready.jpg

Made the extension parts during the day

Extruder2 set.jpg

May-19-2010 [Report on modifications yesterday and test fitting this morning.]

Reaming the MakerBot Plastruder insulator washer to accept the OD of the 1/2" union. Actually, it is the mounting plate. All I had at hand was a dowel and a scrap of 36 grit (bowling alley resurfacing) 3M Production Paper. The process was all by hand and took a more than an hour. The paper wore out effectively making it finer towards the end of the process. After initially getting it to jam in, I continued expanding the new hole until the gap in the copper union opened up. It is an eyeball expansion gap. I worry about the copper transmitting heat and pressure to the acrylic insulating washer. This modification puts the copper into contact with it.

Extruder fit1.JPG Extruder fit2.JPG Extruder fit3.JPG

This morning I did a test fitting of the 'necked out' extruder. Looks good. The heat in the X-axis carriage is going to be intense because the extruder completely blocks the top hole. There are gaps in the side and bottom. I'm going to add a fan and have been thinking about how to cool the extruder's upper parts in this situation.

It fits.

Extruder fit4.JPG Extruder fit5.JPG Extruder fit6.JPG

Now I need a 5 mm plug which I'll try out of a composite wood/plastic tube center, more later. Hope to have it extruding today.

[Notes: A complex cooling solution would be a water block out of a twist of soft copper and hose. The old 12 year old (?) 1 GHz motherboard in a tower just got sacrificed, a new Atom 500 or a Via C7 would run circles around it. CPU and Northbridge heatsinks and fans are going on to cool a RepRap. The motherboard just missed the garbage truck.]

May-18-2010 [2nd post]

[Shameless Plug: I've been using Google Alerts to watch for 'Hydraraptor, RepRap Mendel, RepRap OS X', so anytime there's a post an email comes. Adrian Bowyer should be added to my list, done.]

Everyone should watch this movie made by the founder and crew, Adrian Bowyer himself. It energies otherwise lackluster RepRappers and helps to explain to the newbys. and unbelievers. In fact I know some biologists I'd like to infect with Bowyer's theories.

RepRap, "Wealth Without Money. . . "

And here's the article that pointed to it, which might be nice to hand someone a hard copy to read:

Capitalism and the Age of Machines, 3.0

The best OS X RepRap software that I've seen is Pleasant3D. I'm amazed and am donating, probably several times. Shareware.

MakerBot Plastruder does not drop right in, shucks. Plastruder = 34 mm, X-axis carriage = 39 mm, 5 mm shy. The thickness of the acrylic mounting ring is 4.5 mm. The filament needs to be supported continuously otherwise it will fill the space. The long cap screws should be able to extend far enough; but the Barrel (PFE, PET(?)) will no longer be engaged in the collar that the mounting ring provided.

Plastruder size.jpg X-carriage size.jpg

Plan to use the copper 1/2" union and slide it up into engagement with the acrylic collar, probably by gently reaming out the mount to the larger diameter (sand paper on a stick). Internally a washer of HDPE (cutting board) so that the filament guide path has no spaces. Especially on the 'high pressure' side of the extruder. (Melted plastic would flow into the gap. Hydraraptor had it happen to him.


Mendel rebuild. Yes, you can remove the Y-axis 180 Bearing from a completely assembled Mendel. The build plate was left installed for this whole process as well to maintain alignment.

Continued clearance problems with the Y-axis 180 Bearing. plasmator calculated that is was about 2 mm too long. I'd sanded it before and got about 0.050", but needed more. In preparation for wood removal I added fillets of 50/50 System III epoxy.

Y-180 gluetrim2.jpg Y-180 gluetrim1.jpg

When it had set up enough I used the simplest tool in the world, the wedge, a very sharp wedge. ~1.5 mm shaving. Most of the clearance problem was my adjustment of the Frog Plate upon the #6 10-32 x 1.25" round head machine screws that I'm using; however a Y-axis 180 Bearing that was just a hair higher would be great (2 mm).

Y-180 gluetrim6.jpg Y-180 gluetrim7.jpg

Small blob of glue from the Post Processing was touching the corner of a bearing and was easily removed by scraping with a pen knife after removing the bearing.

Y-180 gluetrim4.jpg Y-180 gluetrim5.jpg

Z-axis woes, wobbly pulleys on the lead screws. The clearance from the deck or plane beneath the Mendel had about 2-3 mm when the belts cleared and ran fair. The pulley was wobbly. I guess a way to do it would be to epoxy the pulley while bolted to a lead screw on both sides and then cut off the excess lead screw. The Z-axis drive pulley is almost touching as well.

My solution was to run the lead screw down and add a nut and washer to the bottom as well as the top of the Z-axis lead screws. 1/4" rubber feet will probably give me the clearance that I need. In the mean time I now extend below the plane and wired on some temp feet out of wooden Mendel RepStrap parts that fell to hand.

Z pulley split1.jpg Z pulley split2.jpg Z pulley split3.jpg Z pulley split4.jpg

I used stainless tie wire on all of the 5/16 studding of the frame. Good exercise in delicate application of force, hard to install. Comes completely off with a snip.

Final adjustments need to be made to the Z-axis loacation. As I have no level surfaces, checking the axis for plumb and level is problematic, I going for parallel and square. Someone posted a Mendel running upside down "Will it work in Space?" to YouTube.

Using a Kmart $5.98 glue gun to put a dot on nut/bolt interfaces and the tip of a screw driver as an applicator for the hard to reach parts.


Installed Bed Springs and plan to grind 4 M4 cap screws down to < 3 mm to get down below first layer build height. [Edit: Acrylic build plate will cover the heads.]

Bed springs 1.JPG

Measuring and inspecting.

Z-axis lead screws looked like rabbit-ear antenna. The rough cut 8 mm (5/16") ground rod in the X-axis was 'close', but not exactly the distance at the bearings. Reset the X-axis carriage ends by releasing every single nut until loose snugged into alignment and then retightened [Claiming points for no Nylocs or Loctite at this stage].

Moved X-axis bearings up by removing spacers, a more solid mounting. Thanks to plasmator for pointing out that I'd installed my belt clamps as end stops for the X-axis rods. Still not happy with the belt bearings. In fact I like the Y-axis belt drive train the best. The X-axis' tight loop into the motor adds a lot of friction through the belt's resistance to flex. Something to improve... a more flexible belt would help.

Z-axis lead screw pulleys were tipped out of being perpendicular by the force of the belt tension. The glue appears to like it tipped. This being set crooked moved the belt up and down with each revolution and moved the drive pulley on the motor shaft, overcoming the rubber cement. [Claiming points for weak glue. It failed quickly. The drive pulley moved down the shaft, then I inspected the lead screw pulleys. I can remove the pulleys and do a better alignment job.]

Thanks to plasmator for suggesting a stop nut above the pulley on the lead scew. I may try a washer as well. What I'd like to do is raise the Mendel on feet, run the Z-axis lead screw down and put a washer and a nut on the bottom. Oh and another bearing too :)

Measuring and locking. Securing all the frame fasteners and putting dots of hot glue on the stuff that may need to spun off and on. Set Z-axis heights by removing the belt, rolling the bed full forward and measuring vertically from the Y-axis rods to the Z-axis rear rod and spinning the lead screw by hand to adjust. Use any object as a gauge, make them equal. You may need to measure several times because moving one Z lead screw always affects the height all the way to the other Z screw. Then drop the toothed belt on without loosing a tooth and re-tension the belt. {Think: twin cam timing marks}

Need to order more fasteners. Going to buy M4 and M3 nut drivers and wrenches too. Using needle nose pliers for turning nuts is a waste of time.

[Shameless Plug: Fantastic service awards for those of my suppliers who keep up with Amazon's shipping speed. McMaster Carr, Adafruit , Jameco, Electronics, MakerBot.]

People have posted elsewhere, wondering if Lady Ada was single?  :) I'm definitely in love with her work! The tools that convinced me to build a Mendel (also CNC gantry router) came from the Adafruit Store. I bought kits: Arduino Starter Kit, ATMega, Motor Party Shield, LCD, another Arduino & Motor Shield. Proof of concept. Yes, I can do this and Yes the Arduino programming environment is a good one.

Don't forget that Lady Ada is at the very core of RepRap <3


May-16-2010 [2nd post]

All the belts were clamped and tightened. Then to get the axis' to move freely, everything was loosened again. All of the spare bearings were used to double up in the X-axis. The Fender washers were not turning and raising the friction, so now I have bearing bearings :)

MakerBot Electronics running Tech Zone Mendel YouTube Tech Zone Mendel with MakerBot electronics

Stepper motor power is too high with the factory settings. Adjusted each stepper controller board down until the motor stalls and then gave each just enough to turn. It is very much quieter and motors should do better cooler and with less power.

X-axis belt clamp mounts to the X-carriage back. You must install the fasteners before the belt :) I removed the upper 360 bearing and put 4 M3 16 mm cap screws with washer sticking out from the inside. No, you cannot see the upper 2 and yes a short arm Allen wrench will fit in the cap screw. Feel only, truly Blind-Zen-Archer country.

X-axis belt clamp.JPG

I am very happy with my no glue / weak glue start up strategy. If I had glued the upper 360 bearing, then it could not have been removed. Nylocs would have been so difficult to work with that I probably would have quit. Loctite the same. Most everything has been disassembled and reassembled at least once. I will add fixatives as needed. (I love Loctite and Nylocs and use it on stuff that doesn't require safety wire, spins 13,000 RPM and goes 100 mph when racing things with wheels and carrying my butt at speed.) I plan on taking things apart again and again and making adjustments, so I will use removable adhesives until I'm out of Beta.

Shameless plugs: Buy good tools. The Radio Shack hex keys failed in the 1st M4 cap screw that I turned. The Allen Brand hex keys from McMaster Carr have never failed yet. MakerBot suggests a $5 Walmart clear plastic tackle box, I chose a Stanley case at Target for $13. Works well and the compartments come out for work and the lid fits very well so that everything stays in its compartment when stood on edge.

Stanleya.JPG Stanleyb.JPG Stanleyc.JPG


The following assembly was done over the course of two days, 18+ hour days. It was slow, because I had only one photo to work from at the start and I'd never done it before and there were no instructions. There are many hours of re-measureing and adjusting to come. I could plug in the tested MakerBot electronics and Plastruder and make everything go right now, but the belts would fall off.

A huge thank you to Kimberly and Lambert Andrus for making this all possible and for answering my questions.

All but belt.JPG

All but tension.JPG

Someone was asking for Y-axis pic help, as I was. Almost completed Mendel will have to do. Those remaining pieces must be belt clamps :)

Deburr every cut bolt stud and smooth rod or you will bleed later.


Here's what I used initially.

I just threw these in straight from Speedy Metals 6 x 18" 8mm ground 303 SS bearing rod. ~$32 to your door. They are very sharp and should have been rounded off, cut me twice all ready.

SpeedyMetals 8mm.jpg

Initial McMaster Carr order. You'll have "... too much of some things and not enough of others."

McMasterCarr 01.jpg

I use marine epoxy and quick set epoxy, silicone, white glue, Loctite, hot glue, but Not Here! Intentionally using the weakest glue I have, Latex Fabric glue (tree sap the original rubber cement). I'll move on to firmer adhesives as needed.

Weakest glue, on purpose.

Weak glue.JPG

Bearing fabrication. I made this up, just got lucky with washers and using both metric and SAE sizes. M6 and 10-32 nuts were used as stand-offs. M4 washers nest inside of 5/16" SS flat washers and 10-32 washers don't fit through the hole. I believe that I will be rewarded for paying close attention to the washers. They are cut and so have a smooth edge and a sharp edge. Always put the smooth face towards bearing, moving or turning surfaces, sharp face to dig in. This way the belts slide against smooth edges.

Bearingfab 1.JPG

Bearingfab 2.JPG

Z-axis assembled. The Z-axis 8mm ground rod has been left 'wild'. I may leave long. I glued the toothed pulleys together on an all-thread stub and then glued them onto the bottoms of the Z-axis lead screw with latex. Both the Z-axis lead screw and the captive nuts are stainless steel which is usually more accurate than zinc plated. Mine's bent. The pulleys are set so that the upper flange is lower than the all interference. The lead screw is about 5-7 mm off the base plane.

Z-axis tensioner

Z-axis tension.JPG

Z-axis drive Note: the end stop/blocks for the X-axis rod stick out. I did not see about making sure they were flush and so used spacers to drop the belt out of interference. M6 and 10-32 nuts slip over M4 bearing support nuts and make nice standoffs.

Z-axis drive.JPG

Z-axis belt

Z-axis belt.JPG

The Y-axis is the most delicate and difficult. When you begin it may not slide at all, be patient. Once adjusted it slides very well. Thanks to Lambert and Kimberly Andrus the designers and makers of Tech Zone Laser Cut Replacement Parts for RepRap Mendel. I asked for help and they set photo set of how it works, here are my shots of their solution. Looking at the alignment of the holes on the Frog plate tells me that I've may have mine in backwards. Some that would be handy for belt tensioner mounting are "over there".

The key to the Frog is 12 @ 1-1/4" 10-32 machine screws, these are round heads. 36 nuts, get lots of washers, I used the10-32 washers in my bearing mounts too. I could not find fender washers, so I made my own - M4 inside 5/16 SS.

4mm fender.JPG

Y-axis 10-32 1.25a.JPG

Y-axis 10-32 1.25b.JPG

Move slowly, adjust the Y-rods to fit the Y-bearing. Keep everything finger tight. Help the Y-axis 180 bearings onto the rod. Then go back and adjust the bearing mounts relation to the Frog plate. The Frog may be hitting the rod. Adjustment in these pictures is to 0.5 mm clearance from Frog plate to rod.

I had a another problem. Even with my best adjustment the Y-axis 180 bearing holder brackets were hitting the all-thread in 3 locations. They were too tall.

Lapping Y180 prob.JPG

The fix was an old piece of 120 grit sand paper wrapped around the offending all-thread and the held while the entire Y-axis was moved gently back and forth across the abrasive until there was enough clearance.

Lapping Y180 fix.JPG

Here are the geared extruder parts included in the assembled kit.

Geared extruder.JPG

Photos by stevew may-13-10

Package Arrived

TZone LCut DSCN3972.JPG

Opened box

TZone LCut DSCN3973.JPG

McMaster Carr belt in pulley

TZone LCut DSCN3975.JPG

[Needs info] carriage parts and geared extruder parts

TZone LCut DSCN3976.JPG

Bed springs

TZone LCut DSCN3979.JPG

Y Axis Bearing 360 and Y Axis Bearing 180 [Need assembly info, please. The Post Processing pictures help.]

TZone LCut DSCN3980.JPG

X Axis Vertical Drive Plate 360 endt

TZone LCut DSCN3989.JPG

X Axis Vertical Drive Plate 180 end

TZone LCut DSCN3991.JPG

Z Axis Belt Tensioner and Z Axis Motor Bracket The Z Axis Motor Bracket is assembled wrong it should look more like Z Axis Leadscrewbase so you could use all 8mm holes. TZone LCut DSCN3993.JPG

X Axis Carriage

TZone LCut DSCN3997.JPG

Z Axis Leadscrew Base

TZone LCut DSCN3998.JPG

parts laid out.

TZone LCut DSCN3999.JPG

frame assembled +/- 0.5 mm, nuts tight.

1600 x 1200 TZone LCut DSCN4029.JPG File:TZone_LCut_DSCN4029big.JPG

I'm impressed with the accuracy of these parts, compared to hand made. I stripped parts from two partial wooden Mendels to assemble this much. With everything adjusted and tightened, a single threaded rod I could still be removed to add a missed nut. On my (oak) RepStrap attempts this would have meant partial disassembly of the entire frame, because of my inaccurate drilling caused everything to bind.

Right now the best guide appears to be the Post Processing photos by the designers. Hopefully there will soon be photographs posted to help with the assemblies that are unique to this version of the RepRap Mendel. A photo set of a completed one would be wonderful.