PeejBotBlog

Status: on pause.

2009-05-13T22:51:00.003-04:00

All progress is relative: in this case, it's lateral. On saturday, my printer died; given that I prefer to possess paper printouts of parts and plans (thppppt!), that's an impediment - when you factor in that I needed to reprint a few of the McWire layouts on self-adhesive labels, that became a larger issue. And funding is, as always, a little tight. Fortunately, a combination of store sale and timely coupon combo shaved a hundred off the next generation model of my previous laser printer: the new one is sleek, quiet, and wireless - just sits in the corner of my office and spits out art and schematics without complaint. So far, I'm happy. Come a long way from the StyleWriter II. (That said, I wish I still had my StyleWriter. A nice big stepper motor in those, along with the control electronics.. ;)

Not dead!

2009-05-06T09:54:00.003-04:00

Just resting. Had some PC issues, illness, funding failures and various other distractions. Not stopped.. just going slower than I'd hoped, and that'll probably be the case for the next month or so at least.

Still to do, organizing my thoughts:
- Cut replacement X axis c-channel to replace the one I drilled incorrectly
- Mill UHMW plastic to PTFE dimensions and install on XYZ stages
- Consider redrilling and tapping the mounting holes in the floor flange and vertical stage to use 1/2" bolts for strength
- Finish threadlocking and reassembling pipe base of McWire
- Order some solder paste-inna-syringe and experiment with the hotplate to get the reflow station working right.

I really can't exaggerate how handy having a modestly complete metric/imperial tap-and-die set is for mechanical construction. I grew up with a neat Meccano-knockoff set, lots of metal struts with predrilled holes and 10-32 screws and hexnuts to hold it all together. Handy for mockups and quick structure work; but! If you've got a drill, some bits, and a half-decent range of taps and dies.. everything becomes an Meccano set in the making. n.n

2009-04-21T12:48:00.004-04:00

Got my control boards in the mail today! Some assembly required.

When I first got into electronics in my teens, I was pretty good with discrete digital and ok with analog, but there were two areas that were just beyond my skillsets - one was working with the miniature SMT stuff, 'surface mount' components which were rapidly becoming all the rage - and the other was microcontroller circuits, which were also the cat's pyjamas. But you needed reflow equipment for the first, which was out of my range, and you needed to learn assembly language for the latter, and I was more eager to make things that worked 'out of the box'.

And here I am, assembling the kit to make SMT boards which will be run by microcontrollers. Funny how life keeps presenting the same lessons to you until you learn them, isn't it...

For scale, that coin is a Canadian penny. These parts are 'big' by SMT standards, btw. Got to start somewhere.. n.n

2009-04-19T15:14:00.004-04:00

As I've worked on the bot infrastructure, I've realized that I'm going to need somewhere to put it. Ideally on wheels, so I can roll it out of the way. But given that the bot will have lots of wires and cables leading to discrete control and microprocessor boards, and a pc, moving it will be very time-consuming and complicated.

Or I could just make the whole thing self-contained and mobile, so that there's only one or two external cables - say, power and ethernet - and can move it at whim. Even while it's in use! But that would be madness.

Heh.

The cabinet is 19.5"x19", about 30" tall - should be closer to 33" with the casters installed. The three lower 'drawers' are all 4U in height and 18" wide, able to hold a standard or rackmount PC, ATX power supplies repurposed for general power requirements, etc. I am planning to install modules for all the servo drivers, extruder controllers, endstop sensors and such on rails in one of those bays, like an old-school analog synth patchbay. The top 'drawer' will hold two drawers - one shallow for tooling and parts, one deeper for an LCD monitor, keyboard and mouse.

It's overkill, plain and simple: its 3/4" MDF held together by 36 quick-connect quarter-inch bolts. That said, it's sturdy stuff, and should put up with several years of abuse.

Ok, was productive.

2009-04-16T16:01:00.003-04:00

I decided to update the plan sticker, rather than just print it out and mark it up accordingly. So, thicker lines, center-marks on all holes, dimensions for thrus and countersinks, tapping instructions.

Probably should have a Creative Commons mark on it, but I'm not sure of the status of the copyright on the original document. Since this is just personal use, I'm cool. If I distribute, I'll check first.

Since I'm out of transparent 1/4" plexi, I'll have to use half my stock of translucent red 1/4" plexi. I'll still have plenty for the project I bought it for, though.

Mildly annoyed

2009-04-16T13:14:00.003-04:00

I'm thinking when I'm done this, I should fork and document a set of McWire instructions that work. That's just the frustration talking, of course.

1. I drilled the countersinks too deep on the Z-stage - because I followed the instructions on the sticker sheet, which are noted as wrong elsewhere in the instructions, but not corrected.
2. The parts list gives 5/16"x3/4" pan-head bolts: you may also require 5/16"x1" pan-heads to fully encompass the thickness of certain parts of the bearing arm assemblies.
3. The bearing arm nuts extend down too far - with a 1/8" PTFE bearing glide in place, the bolts will collide with the arms of the pipe infrastructure on the X-stage. You might want to have a handful of 'jam nuts', which are - not too surprisingly - thinner nuts. Personally, I'm going to part them to size on my lathe, because I'm getting annoyed enough to become a little bloodyminded. I went to some trouble to procure these nuts, and by golly, I'll make they should fit. (cue the lightning and spooky music.... now!)

Anyhow. Yeah, nothing ever works right, but this is revision three of another person's work - it should be smoother than this.

Ah well. Need springs. Spring, as they say, is coming.

For now, I need to rebuild the Z-stage from scratch and dust a few mils off my nuts.

2009-04-14T21:02:00.003-04:00

Ok, I'm tossing the opto-endstops - I'll likely try mechanical microswitches instead. Heard enough about troubles with the optoboards, odd given how simple they should be. Also need to make another two bearing arms - but at least the verticals are all cut and rounded now. I've resorted to cutting the acrylic with score-and-snap, the classic way, and rounding the corners on a grinder. Overkill, but it works.

Turns out the vertical base may well need to be made of 1/2" acrylic. The docs are.. more suggestions than anything, really. I can't actually determine if anybody has made one of these from scratch before. Which is sort of cool to be doing it now.

2009-04-14T18:57:00.004-04:00

About that bearing arm: you'll notice there's a mark across the arm from side to side. That's from the jigsaw: I found I could alleviate that with clear tape across the pieces I was cutting, to help keep the sticker-labels from rising. Since I'll be cutting the two replacements manually, it's not a concern any more.

Note to self: need a couple of holddowns for the workbench.

BTW, Mr. Garner's instructions on changing your bit profiles to cut acrylic more easily? Bang on the money, work like a charm even on the big holes. Always good to have useful advice out there on the 'net.

Delays.

2009-04-14T18:41:00.003-04:00

Things have been going quite well! So it was time to encounter some issues. There is an order to such things, after all.

Acrylic sheet is lovely stuff - very clear, consistent, grinds well, and cheap. It is, however, brittle if you catch it right. You shouldn't drill it without a sacrificial piece of wood underneath it, to keep a broad bit from cracking the hole as you approach the end. You also should predrill all holes you intend to countersink: do not try to countersink them from the start -

- bad things happen. For that matter, drill the holes before you cut out thin, fragile bits, like the bearing arms. The instructions recommend a jigsaw: I recommend a coping saw. It's faster, doesn't melt the acrylic, and gives better control. (A scroll-saw or bandsaw might be ok, but I don't have either right now.)

Acrylic tends to ride up along standard metal-cutting bits, and then fracture or shatter. The 5/16" holes were all fine - excepting one arm that fractured, which I'll recut from scrap. The 1/2" holes, however, were another matter.

Ruined the Vertical Base, completely. I'm going to have to remake it from scratch. Fortunately, acrylic isn't too pricey - I may even have an appropriately-sized piece in the car. For drilling these holes, I could purchase a proper plastic-cutting 1/2" bit, but that's out of my range right now: I'm going to pick up a cheap jobber carbon-steel 1/2" bit, and modify it. You can see Jon Garner's instructions at http://www.hawkfish.org/snailman/acrdrill.htm - you want the bit to scrape the plastic, not cut into it. (I might do the same with a 5/16", while I'm at it - I've probably got a spare one in my toolkit somewhere.

Progress.

2009-04-13T21:07:00.009-04:00

The 1/4"-20 threaded rod is what moves the three stages, allowing the head to index any point in threespace (within reason). Cut it to length with a Dremel to reduce the chances of bending it, and chose the straightest lengths - the drivenuts slide along the threads smoothly for the whole length of all three pieces, so it should be good.

Trimmed the labels for the anglealuminum (angluminum?) with a knife, then took my time aligning the front edge with the leading edges of the anglecuts. The angle joins with the 3/4" channelstock to create a rigid platform for the stages. (It'll all make sense someday..)

Could use a hacksaw to cut the angle aluminum, but a jigsaw is faster and accurate enough if you take your time and use clamps to make things relatively stable. Always remember to leave some space between your cuts for kerf (the waste your blade turns into metaldust).

Finished work, all the angle aluminum cut to length, labeled, drilled and deburred. Need to finish the bearing arms now: next entry.

Unrelated, but I realized my centerpunch was misbehaving pretty quickly and decided to properly cut the tip on a 60-degree taper.. and rapidly discovered that the entire thing is unbalanced. No surprise, it's a tool that gets smacked with heavy objects constantly. Still, irksome. I used a transfer punch to finish punching the drillholes, and will make a new punch when I get around to it. (Need to make a knurler first. So many projects, such a short lifetime...)

Instructions? Why?

2009-04-10T15:49:00.003-04:00

The base is intact, if not drilled. While I'm waiting for the mounting hardware, I thought I'd cut the channel and angle aluminum to length. Two problems:

1.You can cut all the rails you need out of the 3/4" aluminum channel from two 48" long pieces, but you'll have to lay them out carefully - one 21" piece and two 13" pieces on each will take up 47" plus what you'll lose from the saw kerf. So don't make any mistakes. (No pressure. ;)

2. The parts lister declares you'll require 43" of 1.5"x1.5". The instructions require 2 X supports, 2 Y supports, 2 Z supports, 1 extruder support and 2 bearing arm mounts: 2(6&3/8")+2(9&1/16")+2(5&15/16")+1(2.5")+2(9") == 53&1/4". not including the kerf from cutting. The 43" is correct, if you're only cutting the six stage-supports - it omits the extruder and bearing arm mounts.

End result, I need another 13" of aluminum channel because of my own shortsightedness, and another 18" of angle aluminum because I didn't crosscheck the second parts list for accuracy against the inline parts list. Ah well, if it was straightforward, everybody'd have one of these already, right? Think I'll hit Metal Supermarket to get these bits, though - can't spring for another couple four-foot lengths right now.

The driver boards for the steppers are en route now - probably be another week or two before they arrive. I have lots to keep me occupied before then.

Construction begins.

2009-04-08T20:52:00.000-04:00

This, in a nutshell, is the base of the McWire: a few pounds of black pipe, wrenched together into something relatively orthagonal to itself. It's snugly threaded together enough to stand up to a fair amount of abuse now: Once I've marked the starting and finishing points on all joints, I'll put the stickers to guide drilling the pipes in place, dismantle, drill, apply threadlock to help hold it together, and then reassemble.

I'm vaguely considering painting the base. There's no need, really - corrosion isn't likely to be a concern in the basement (famous last words, I'm sure), and prepping the pipe will be a pain in the rear - black pipe is pretty grimy, even brand new. And yet, the aesthetics of a yellow-and-black industrial bot appeal to me. We'll see. Patience is one of the virtues I'm trying to exercise.

Nearly the end of the beginning.

2009-04-08T15:31:00.001-04:00

I've got an order in at Fastenal for the bolts, nuts and whatnot: I still need to trim down the UHMW plastic, and pick up three mounting brackets from Generic Hardware Store, and then I'll finally have all the parts for the mechanical structure. Then the fun really starts. (No, really!) And I have skate bearings! Ludicrously expensive. I could have probably picked them up for half the price or less on Ebay, but they were there, so was I, the shipping's already done.

And they're green. Shiny. A potent motivator. n.n

Think I'll start photodocumenting this as I go along the actual assembly - the Zork approach is a bit dry.

More parts.

2009-04-07T09:43:00.000-04:00

Found another length of 3/4" aluminum channel at Rona, bringing me up to 8' total - for the linear glides, that's all I should need. (I hope.) Also found a box of 10-24 machine screws, pan head. No luck with the 5/16 machine screws - just a bit large for the three big boxes I've visited, Rona, Home Depot and Canadian Tire; I'm going to have to hit Brafasco or Fastenal for them, and the rest of the mounting hardware. Fair enough; it's just a little bit much to expect one-stop shopping for this project.

I've also apparently lost one of the bearing arms I cut out - fortunately the patterns include a few spares, so I still have enough to make the bot - for that matter, I have enough leftover plastic to redo any individual parts, if needed.

Back again.

2009-04-07T09:42:00.000-04:00

Seems Google thought I was a spambot for a bit there. Amusing: I'm writing a blog about building a 'bot, and apparently that fails the reverse-Turing. Or, as a friend pointed out, "I passed the Robot Test!" Now I can walk around with deely-boppers on my head making 'meep-meep' noises. n.n

PTFE vs UHMW

2009-04-06T10:57:00.000-04:00

I was at Lee Valley Tools to pick up an 82-degree countersink - they had the best price for a bit I'm only going to use for plastic/wood work, with occasional metalworking. Half of the machine screws are flat (countersunk) 5/16" bolts, so countersinking the plastic they'll thread into is important to keep surfaces from getting scratched and caught mechanically on each other. It also looks more finished.

While I was there, I grabbed a couple strips of ultra-high molecular-weight plastic. The specs for the McWire bot list strips of 1/8" thick, 1/2" wide Teflon for linear bearings, to help keep things moving smoothly (or not, depending on the direction of movement). Lee Valley has a generic UHMW, which should have a coefficient of friction very similar to PTFE, and even better wear resistance - it was also in stock, which is handy. I'm trying to source as many parts for this project locally, trying to reduce the global footprint associated with it. Of course, local parts all have to come from somewhere originally.. (shrugs)

Anyhow. The PTFE strips spec'ed are 1/8" thick, 1/2" wide, 2" long; there are 16 of them. I picked up two strips of UHMW plastic, 3/8" thick, 3/4" wide, and 2 feet in length. The thickness will need to be corrected - that will affect the alignment of the feed nuts. So I'll be passing it through a table saw - if I use a narrow-kerf blade, I might be able to squeak two 1/8" strips out of each 3/8" thick piece. More likely, I'll wind up sacrificing half of the thickness to get it down to size. The width doesn't seem to be a problem, but since I'll have the saw out anyhow, I might as well pass it through a second time to shave 1/8" off the side and bring it to a finished dimension of 1/8" by 1/2", and then I can just cut the 16 2" lengths with a mitresaw.

I should start carrying around the parts list in pocket, for chance opportunities to pick up bits and bobs for this project...

The Beginning.

2009-04-05T21:17:00.000-04:00

So, I'm building a bot. This one, in particular: the McWire 3-axis Cartesian bot, revision 1.2. It's an intermediate step before tackling a RepRap - they use the same control electronics, and parts for the latter can be manufactured on the former. So far I've begun manufacturing the acrylic stages of the bot, and I've acquired the plumbing components and most of the aluminum. I'm also encountering some issues that might be addressable for a future revision of the plans: this blog will be a combination of build record and notes along the road, as it were.

I'm guessing that the majority of the McWires in existence were built using lasercut parts: I don't have a lasercutter, and I like doing things the hard way. ;) So, I've printed the part layouts onto adhesive 8x10 labels, and have been cutting bits out using a jigsaw. The printing of labels only encountered one failing, namely my printer cropped the borders on the largest stages by a milimeter or two - still, good enough to work from. (Remember to turn off 'Fit to page' - you want to do this at 1:1.)

I don't have the teflon strips for glide bearings yet - I'm considering using UHMW plastic instead, depending on what Acrylic Depot has handy.

The plastic templates fit with room to spare on a 8.5" wide, 5' long sheet of 1/4" acrylic. (The thickness of the plastic isn't specced elsewhere, so far as I've noticed..) When you're sticking the templates down, start with a corner and smooth them down in one slow sweep with a rigid ruler to the other corner, peeling off the backing as you go - don't take the backing off and try to flop it down in one go, you'll ruin it.

None of the template holes have center marks, making hand-drilling them a little harder than necessary; none of them are marked on the template as to hole size, or if they'll be tapped later. (For the record, most of them are either 5/16" diameter (of which some are then countersunk), or #25 drillholes, tapped to 10-24 - they advocate using self-tapping machine screws, but I've got the taps handy, so I might tap them by hand instead. Most of the holes are marked if they're to be countersunk, except for four which are countersunk from the other side. (A dotted line might work for indicating that?)

I need to find a good inline skate shop in Mississauga to pick up some sealed skate bearings with a 5/16" internal diameter.

I used a circular saw to cut the stage platforms - nice straight edges - and a jigsaw to cut the smaller pieces, like bearing arms. Still have to cut the endpoints and vertical arms - for rounding the corners, I'm going to use a sanding wheel on a dremeltool, rather than risk the edges with a saw. Accuracy of the edges isn't supercritical, but it's nice to have a good-looking piece of equipment when you're done, instead of something jagged and sloppy.

Another day, another blog.

2009-04-05T21:09:00.001-04:00

(waves to internet) Interesting content will follow. None of it'll make much sense for a while, though.