The Circuit Farm

I just couldn’t face another marathon session today. I put some time in, and I got some good work done, but I quit before dinner and took the evening off.

I started the day at Home Depot. I knew that without a chuck key, I wasn’t going to be able to start really putting the parts together in a permanent way. While I was out, I also picked up some sandpaper and I was able to find a set of Allen drill bits. w00t!

The search for the actual chuck key took quite a bit of time, though. 2 hardware stores and a trip home to remove the chuck and arbor from the drill press were required before I found something that would work. The Ryobi chuck is irritatingly smaller than 1/4″ pilot hole, so it’s difficult to find a chuck key that will work. I finally figured out that 15/64 (that’s 1/64 less than 1/4″) works. grr.

With the old drill bit out of the chuck, I was able to pop in a Forstner bit and get all the pocketing done that I needed to.

I also figured out how to turn the head of the drill press sideways, so that I can drill into longer stock. That’s a big win.

Armed with a power Allen driver and a full set of gantry parts, I threw together the front of the gantry (complete with cross dowels) in … fairly short order.

The power went out for a couple of hours; there was a pretty big windstorm, but we really need to think about getting a generator. The stuff I was doing did not require power, but I took an hour off anyway, because I was starting to get tired and I was definitely hungry.

Back to work, I added one gantry side, but realized that the other gantry side is really not the same size and shape. I think maybe I did an “inside” profile on one of them by mistake. I had considered picking up a quarter sheet of ply while I was at the hardware store, but didn’t do it. I think that I’ll be re-cutting at least one of the two gantry sides (I can’t figure out which one is “correct” from the data I have).

I moved on to the Z axis. I really needed to watch and study the construction videos, because it was a case of painstakingly getting a bunch of stuff put together, only to realize that I needed to rip it all apart in order to get the next bit in. I probably built and tore down the axis 3 times today, each time a maddening succession of misdrilled holes and too-tight fits. I stopped once I got both router mounts in, with one side of the axis installed, and the other tacked in place with a couple of bolts.

I just didn’t have the energy to keep going.

I’ll get back to it tomorrow, and more progress will be made. If all the parts fit together perfectly, it would still take a little time to get everything running right. But just like the last CNC build, the edge drilling is really suspect, and it makes everything take forever to complete, because there’s a lot of drilling and re-drilling going on.

Todo:
– complete Z axis construction
* build the other side front
* drill the clamp holes for the router mounts
* attach the top and bottom assemblies
– more recuts
* gantry side
* vacuum shoe (drawing needs to be mirrored — I hope CamBam can do it)
– rails
* Need to buy 2x 59″ and 2×16″ (I suppose I could use the 52 and 8 to make 60″ rails and then just buy the 16″ ones)
* drill the rails (measurements? I imagine I will just lay the rails in situ and mark them as we go — drill slightly large for adjustment)

Without rails, there’s no need for bearing surfaces to get done.
Without bearing surfaces, there’s no real need to work on chain mounting.

I can’t start harvesting parts off the old CNC machine until I know it’s done cutting. That’s keeping the Z axis down for now.

I still need to stop by Tacoma Screw to pick up some ceramic bearings.

I’m suddenly realizing that I have a long way to go yet.

I got up fairly early this morning and started working on the CNC build. I think I was in the shop before 9am.

In a burst of inspiration, I decided that since the machine was already lined up and ready to go with the gantry top/bottom piece, I could just lay a second workpiece right on top of the first, rezero the Z axis, and spend a little less time worrying about registering the workpiece to the table. I also decided to swap out the previously-cleaned 4-flute bit for a brand new bit. All I have left are 2 flute bits, so that’s what I used. I am going through endmills
pretty fast. I better put in another order for some.

I can only say that this plan (new bit, layered workpiece) worked out perfectly. With very little muss and fuss, I had the first gantry top finished, and what with one thing and another, I now had outline marks on the bottom for all the drills, so I could see if the machine could hit them correctly, and it did. Awesome!

I spent a little time getting the final board registered; this is a fiddly process, because I don’t really have anything on the table that tells me where to put stuff. So I end up laying it down “close”, then running the bit back and forth until reality matches what the computer thinks is going on. Then I find a convenient place to throw in some hold down screws, and I”m golden.

I’ve started putting hold down screws in each part if I can; it seems to help with that final cut out. If the part has a 1/4″ hole in it, that makes it easier. If the part has only 7/16″ holes, I have to add a nut (in use as a washer) to the hold down. It’s a bit fiddly, but it works out ok.

That said, there were a couple parts that I simply couldn’t screw down, so I had to hold them down with my fingers. This is a process that takes a lot of concentration. Stay away from the router bit, but hold the part steady and firmly. I did ok but it’s one of those heart in the throat moments each time I do it.

With the final board all cut out, I was able to shut down the CNC PC and move to the drill press to see if that jig was going to work.

There is a lot of edge drilling in this build. I assume I’m going to have to use about 100 cross dowels, and each one will require an edge to be drilled for it.

The jig did… alright… There were a few glitches. I learned late in the edge drilling that the model I’d used was not at a standard distance from the edge; I had blamed my CAD skills and the CNC machine. I feel a bit justified now.

All of the edge bores that I did with the dirll press seem to be OK. But the ones that I had to do by hand are another matter. One of the gantry sides is in really bad shape, and I’m generally unhappy with my skills with a hand drill. I was also getting pretty tired by this point, as it was after 4pm, and I hadn’t had anything to eat all day, and I’d been on my feet doing CNC stuff for most of that time. I’m not complaining; I signed up for it. But I really should have saved the hand edge drilling for a time when I wasn’t in such a “just get it done” frame of mind.

I still have one piece left that I need to counterbore; it’s the piece that ended up being mirror-imaged. I would have done it, but, of all things, I can’t find my chuck key for the drill press. I’m hoping that it’s still in a box somewhere. But, for now, I am stuck doing only 1/4″ holes with the drill press. sigh.

I started to fit the parts together. Things are still a little tricky and tight-fitting. But I got all the pieces test-fit on the gantry (except for the non-counterbored piece) and the Z-axis.

That is a LOT of bolts. I need to make sure everything is ready to be put together permanently, then cross-dowel it all together. With those pieces built up, everything left is either a motor mount or a chain mount. Well, the Z table needs to be bolted on, too. But that’s mostly putting on bearings.

The TODO list got a little shorter:

– buy a chuck key that fits my drill press
– sand all the new parts (maybe pick up some sandpaper while I’m at Home Depot)
– decide whether I want one more quarter-sheet of ply, so that if there’s a part that’s egregiously messed up, I can recut it (only one gantry side, and one piece of the vacuum shoe, fit this criteria)
– drill out any troublesome cross dowel holes by hand
– cross dowel the test-fitted parts together
– once I have access to my forstner bits, there are a few pockets that were missed (one gantry side, Z axis bearing seat, and the re-counterbores)

With those tasks accomplished, I will be ready to drill the rails. I haven’t quite decided what to do about the fact that the rails are all too short. Probably head over to online metals and buy some more parts. Road trip! I just have to figure out what I need.

With rails in place, I’m essentially done with the gantry. Then I’m off in tablesaw land so I can build a torsion box.

Having a whole day to myself really helped to move the project way down the road. Thank you to my loving wife who had a pre-planned engagement for the kids that took most of the afternoon. Here’s another look at that gorgeous gantry.

I really like the 4-flute endmills. They leave a very clean edge on the cuts, and I feel like they eat wood better than the 2-flute bits.

But they seem to foul pretty quickly.

I started through the cut-out pass on the first (auxiliary) board, and about 2/3 of the way through it, the sound of the cutter changed a bit, and “dark” sawdust started coming out. I knew this was a bad sign, but I was almost done. By the time it got to the last part (a very complicated cut, that one), the smell of the sawdust suddenly changed, and I realized that something was smouldering. I got the ember put out (I ended up sucking it into the dust collector, not a great idea, but I got lucky this time), and opened up the shop to let the smoke out. There was a lot of it.

I changed over to a cleaned 4-flute endmill (I’ve been soaking them in Simple Green, which seems … not that effective — I will try Acetone next, then go search out some saw blade cleaner), and started on board #2. I decided to do the board that has the gantry top and bottom next, since it had the doweling jig on it. I outlined the parts and then did the cutout of the jig only. I was still a little wigged out about the near disaster, and it was starting to get late.

I got a small pile of parts for the Z axis, and the doweling jig cut out. I also got one half of the vacuum shoe, which is (I’ll say it again) a very cool looking part.

I have the parts for one of the gantry top/bottom outlined; just have to cut those in the morning.

By the time I was done, though, the “cleaned” endmill was starting to sound rough again. I left it attached, but I’m going to have to swap it before the next run. And put in an order for some more endmills.

The doweling jig looks really cool all put together. I hope it will work as good as it looks.

I have a full weekend of CNC building ahead of me. I hope to make some big progress.

The last box of parts arrived today, so I started construction on the CNC machine.

I also headed over to Home Depot to pick up the lumber for the torsion box; I ended up with 3 sheets of 3/4″ MDF (which is about $33/sheet, a lot more than I was expecting) and 2 sheets of lauan ply (1/4″ stuff). I bought a new sawblade too, for doing all the MDF cuts.

Anyway, to the build.

The small inaccuracies in bit position and angle added up to some of the holes not… quite… lining up. I mean I think everything is going to come out OK in the end, but a lot of the pieces fit tight.

I ended up with 2 x-axis mounts, for some reason.
I cut one of the Y supports backwards (it’s a mirror image of what it should be).
But I got the Y gantry together, mostly. And I got all the rest of the parts sorted into piles with their various assemblies.

Here’s what’s left to do.

To buy:
– new Y axis rails (I bought 2×52″, they need to be 2×59″)
– new Z axis rails (I bought 2×8″, they need to be 2x~16″ — measure against the appropriate piece)
– washers 3/16″ thin and thick (need ~16 of each)

To be designed:
– router mount (probably make one for the DeWalt and one for the Craftsman)
– gantry top and bottom (these were too long to cut on the book machine, and too wide to cut with the scraps; I need to cut shorter lengths and make do. Either 2×24″ and 2×18″ OR 6×12″ will do the trick. Need to line up the holes, though. Cut from the 24″ stock that’s left and design from the DXF)
– cross doweling jig (is this part hiding out on the DXF somewhere?)
– Figure out placement for all the re-cuts.

Re-cuts (I should have plenty of stock for all this; about 12″x24″ on each of 4 boards, plus some small scrap):
– 2 router mounts
– vacuum shoe (2 parts)
– 2 small router mounts (square, with 5 cross dowel holes)
– 2 large router mounts (can just route them flat, maybe)
– 2 Z rail supports (can just route them flat, maybe)
– Z motor mount (it’s really bad)
– a couple of the chain clamp parts (they’re pretty quick and are in tough shape)
– gantry top and bottom pieces

To drill:
– 7/16″ holes in the Y gantry backer board.
– all the cross dowel holes
– a few places where pockets were missed (bearing seat on one gantry side, bearing seat on Z axis bearing mount)
– holes in the rails (measure against the as-built)

To grok:
– figure out the new Z-table (compare the 4.1 and 4.0 ones and study the changes; in particular, did the bearing mount points change?)
– X motor mount (3rd cross dowel?)
– Z screw mounts (how do they work?)

It’s going to be awhile before I’m finished. And I am more certain than ever that I am going to cut a new Blackfoot as my first project on the new Blackfoot. For $50 of plywood, there’s no reason to hesitate. I am unimpressed with the accuracy of my first couple of boards, but I am not certain that running 2 more boards through the Book machine is worth it. I can hide most of the egregious errors, but I’m really fighting the little “off by a touch” places whenever I try to put a screw in.

… in the GW Bush sense of the term. 🙂

All kidding aside, I made it through board #4 today, so all the major cuts are done at this point. A lot of the pieces need cleanup (maybe even recut). And of course I need to build the table, more on that in the next post.

This cut went very cleanly. I swapped endmills before the final cutout, since the first endmill (which had done all of board 3 and all the drills and pockets of board 4) was starting to worry me.

This board was the one that went all the way to 38″; the rightmost edges were actually hanging off the end of the table. The pieces were also all very close to each other. Unlike the other boards, this one just has a big hole in the middle of it.

With a complete pile of parts (at least at some level), I spent some time sanding everything. I need to start putting everything together to see how much stuff needs to be recut. I also need to do all of the edge drills, but my understanding is that an edge drilling jig is provided, so I need to figure out what that’s about and then get started on that.

Then I need to build the table. But I’ll explain that in its own post.

I wish I’d had a project to cut before the Blackfoot, to get myself back into the swing of things. I had a pretty rough time with the first two boards of this project, but once I started really tuning up the rig, board number 3 went off without a hitch.

Now that it’s raining, I decided that I had to move the CNC inside to keep it from getting rained on. I did not want to have the same dust problem as before, so I knew that I needed some way to contain the rig. Kristi offered the family tent, as she is apparently in the market for a new one, so I set it up inside the shop and put the CNC table inside.

Let me take this moment to say that it’s ludicrous that I can set up an 8-person, two room tent (approximately 7’x12′) inside the shop. Back to the show.

I got the vacuum set up on a separate circuit so that I could gather dust while it was being made.
Knowing that I was going to be in an enclosed area, I grabbed a dust mask (!), safety glasses, and earmuffs, and got ready to cut.

I also set up a MOP to throw in a 0.5″ deep drill hole at 0,0, as suggested by a previous commenter. This, above all else, was probably the major reason that today went so smoothly. I re-homed the axes after each MOP (outlines, small drills, large drills, pockets, and cut outs), and the axes were off noticeably each time; just a couple hundredths of an inch, but I am betting it would have added up by the end of the day. Re-homing saved my sanity; thanks again for the tip!

I have also been enjoying the “outline” MOP that I’ve been throwing in. Being able to see where the part boundaries are really helps me to visualize what’s going on with the rest of the cut.

I decided to watch the whole cut, like a hawk. I also vacuumed up a lot of the dust right away, and was pulling parts off the machine as they finished. Watching the router to see where the cut would finish, then holding down the part with my hand once the router got to a safe place… it was something that required concentration.

Finally, I decided to re-grease the X and Y axes after each MOP. It’s probably overkill, but the axes were running really smoothly, and I began to look at my screws as “machine parts” instead of “fasteners” for the first time. I understand why having an oilcan around moving-parts machinery is a good idea. Thanks, Tin Man.

These careful considerations made for a pretty easy cut.

I have just one gripe about the way the cut went today. I have been having the same old problem where my < 3/4" workpiece (it's nominally 18mm, .70") is not getting cut all the way through with my -0.75 Z depth. This is probably because the Z axis gets out of synch sometimes. I had to re-home it twice during the day, and I noticed that a lot of pieces had holes that were not punched all the way through. Oh, bother. I'll do them on the drill press, or even with a hand drill, since they are drilled almost all the way through. Setting the Z depth to -0.85 was not the right answer. That messed me up on board #1. The edges on the parts are really nice today; there is a noticeable difference in the tearout between the earlier cuts and this one. I think I might be sold on the yellowish TiN coating. Or maybe it's the 4-bladed endmill. I dunno. I like it. I threw in some screws to hold down small pieces of scrap in between the parts. I think it helped. One more board to go. The final one will test my X axis travel, as the parts go out to 38" instead of 36". I checked carefully, and the board is set up with a hair over 38" of cuttable length. It extends past the X table by a couple inches. It's all fully supported by spoilboard, of course, and the axis could physically travel at least another inch. But it's new territory, certainly. All the parts need sanding. I need to head out to the lumber yard and get 4 sheets of MDF and a bunch of 2x4s and 1x4s to make the torsion box, too. And I need my nuts and bolts to show up on Monday. But I am getting awful close to having all the parts for a new CNC machine. Awful close, indeed.