I’m ‘eyeballing it’ when I speculate somewhere between >10’ and <14’ is both reasonable (realistic for many to build/store) and is predictable to be (knowing wider/higher adds accuracy, but recognizing width adds chain sag and eventually limits height, as well as stock chain length limitations) better than a ‘stock’ frame width. (just under 10’)
I’d recommend the same as @dlang, starting at 10’, simply to have a common baseline for the unistrut frames (the 3rd known build?) and then think about modding from there once you’re dialed in.
Remember that alignment, while there’s room for error, is still pretty important, re: wingnuts, you’re not going to be changing the machine’s dimensions w/o leveling/straight-edge elements afterward. At that point, you’ll probably want to tighten things a little beyond hand-tight, consider that you’re talking about loading full sheets onto the back and on to and off of the machine. imho, a bump with a full sheet is enough to tweak ‘hand tight’ bolts.
I propose that we by pass the entire ‘miter saw sled’ concept and put those efforts into establishing maslow procedures that usurp the need for such an addition by:
making the frame rapidly adaptive to multiple thickness of stock
faster, more accurate.
extending ground control to produce simple/common cuts, change stock, iterate to ensure only successful active cuts.
Not that there isn’t a lot of work towards all this already, but following the rapid progress of the frame designs, accuracy, software, etc from the early summer beta units to today’s ‘all initial units shipped’.
It is a quite dramatic and rapid change, and it doesn’t have to slow down yet, I think that there’s enough to still figure out as far as ‘best practices’…
put a hook on the unistrut at the center and a hole/eyelet at the top of the
sled so that when you need it out of the way, instead of disconnecting the
chains, just lift the sled up and hook it to the top unistrut
making the frame rapidly adaptive to multiple thickness of stock
unistrut makes this very easy, you just slide the top strut in or out as needed.
put a hook on the unistrut at the center and a hole/eyelet at the top of the
sled so that when you need it out of the way, instead of disconnecting the
chains, just lift the sled up and hook it to the top unistrut
good call, I’ll work something up along these lines.
unistrut makes this very easy, you just slide the top strut in or out as needed.
not that easy, yes it’s movable, but it would still be a 5min operation and without a guide rail on the frame’s “Z-Axis T-Joints” could require a re-calibration. (even though the motor distance might not change much, it still could be askew to the workpiece +/- ~x mm.)
I’m just saying that there’s more work to do all around, and that if we focus on speed/agility of the machine itself, and developing some good solutions for common operations we’ll probably get further and faster than a custom solution to do a single operation (in this case, Vertical Full Length Cuts for breaking down full size stock)
If we could make the top beam itself move in and out based on material thickness it would alleviate this problem without needing to re calibrate. In fact, your setup is already most of the way there. If you flipped the Unistrut that the beam mounts to, it would give you that level of adjustment.
you are talking about two different things.
mrfugu is talking about the distance from one motor to another
you are talking about the distance from the beam holding the motors to the back
of the frame.
changing the front-back location of the beam doesn’t require any calibration.
changing the distance between the motors, even a tiny bit, requires calibration.
First, some shots of the perfectly good chain guides, using the included Delrin roller, trimmed ~1/4" to fit 1.5" screws. The screw is kept loose enough for the roller to spin freely.
Note: This position (mounted directly to the motor face) places enough pressure to keep chain on at least 180deg of the sprocket, and keeps the exit chain aligned entering the sprocket.
This is important because as you’ll see I have the loose chain twisted about 45deg, due to not wanting to add much to the frame in order to tension the chain.
Everything worked great during calibration, I still need to test running it out to the far corner, but I don’t anticipate many problems, if anything I’ll need to add another foot of chain to each side.
Add ons:
I took @dlang’s advice and added a hanging mount for the sled in the center of the main cross bar:
Tomorrow I’ll have 3 more modifications and intense calibration:
Some spoilboard stiffeners in the form of more L-Brackets on the back side of the bottom and middle horizontal bars.
z-axis= 0 connection via foil backed insulation foam (1/2" or 1" ) stacked on top of a 1/4" spoilboard. This should leave enough space left on the bottom rail to mount 9-18mm ply w/o much overhang of the bottom strut.
Looking awesome! And thanks for the closeups of your motor mounts. I like how you put the motor mounts on the opposite side of the brackets. That configuration should solve an issue that I have been wrestling. I am planning to use a single piece of unistrut bolted to a 2x4 frame on which to mount the motors. It seems like folks have established that rigidity between the motors is the main area that is important.
Out of curiosity, are you getting flex in your spoilboard? And are you using plywood as your spoilboard (judging from the pics you posted). I am planning to use 1/2" MDF with a 1/2" foam insulation panel over top of it. Mounted to a 2x4 frame I expect that the MDF should provide plenty of rigidity, but perhaps I am missing a force?
Also looking forward to hearing how this works out for you!
Resistance to warping is a factor as well, having the workarea flat is important to avoid chain jumping the sprockets. Not sure how MDF behaves with respect to sagging between supports, but I found that even 3/4" ply would bow unless I took steps to prevent it.
What distance between supports were you using when you saw the 3/4 ply bowing? (and was the bowing just happening just due to gravity, or was there a force applied? This is why I am asking about a missing force in my post)
Generally MDF is more dimensionally stable than ply (assuming it doesn’t get wet), so I figuring that I won’t have to contend with warping due to seasonal changes.
My underlying frame supports the MDF all around the perimeter and through the center with a maximum span of 3’-8" which is pretty long for 1/2" MDF, but I am hoping that that mainly vertical orientation of the frame will mitigate any sagging. If not, I can add supports, but I would be surprised if I needed to.
It was the normal frame arrangement with the two angled supports, and I think the force was a combination of gravity and the leverage of the motor arms, as well as my relatively humid climate. Sheet stock at the big box stores around here is often ‘pre-bowed’, though I had high-graded to pick a flat sheet to use. I replaced it once after seeing the problem cause chain jumps, but the new sheet bowed within two weeks.
It’s just something to watch for and think about. It doesn’t seem to have affected others - Bar has done all his work on a stock frame, in much the same climate 175 miles south of me. Go figure, I’m ‘just lucky’ I guess.
Out of curiosity, are you getting flex in your spoilboard? And are you using plywood as your spoilboard (judging from the pics you posted). I am planning to use 1/2" MDF with a 1/2" foam insulation panel over top of it. Mounted to a 2x4 frame I expect that the MDF should provide plenty of rigidity, but perhaps I am missing a force?
My thoughts are to install a 1/4" plywood with 1/2" foamboard as the spoilboard.
there is about a 3/4" sag in the middle of the spoilboard when sitting in my rack currently, I think i can fix this just by putting some opposing L-Brackets along the back of the horizontal support rails spaced in 3rds across the supports.
ascii diagram of frame, as seen below the x-bar:
= : L-bracket back edge of supports.
-|----=------=----|-
| |
| |
-|----=------=----|-
If this doesnt do it, I can always add vertical members and physically secure the spoilboard, but I think that with the L-brackets, after adding a workpiece, it will lie flat after clamping for actual milling.
Ah, ok. Yeah, the stock frame does seem like it is relying on the plywood’s strength alone. As a suggestion, a 2x4 frame (cut an edge to be flat) attached to the back might be an option instead of trying to find a flat piece of ply.
