If the mounting point for the chain is in the middle of the cross bar, such as in the 45 degree version, aren’t the side arms always parallel to the chain? And isn’t the cross bar always perpendicular to the chain?
If I am right about that, then why does the 45 degree version need any special chain mounting?
I get what you are saying but I’m actually saying that only the Top Mounted design needs care for the chain. I’m not talking about modifying the bars to accommodate the chain. I’m suggesting using simple shackles to connect the chain to the bar:
My point is that for the 45˚ version there is no interference with such a shackle (you can look at them in use in the pictures I posted the other day). With the top mounted version they will have the propensity to hit the bar if the bar is of sufficient width and nothing else is done about it.
Shackle + 45˚ design = no problem.
Shackle + Top Mount design = problem.
That’s all I’m saying.
Obviously we could design more complicated bars with cutouts and offsets to account for the chain (as you have shown in your drawings) but I’m trying to keep it as simple as possible (i.e. straight, smooth bars).
With the shackles I used in my mockups I can easily achieve a 27˚ internal angle without interference. On the 45˚ linkage design the minimum chain angle from the bar is 62.07˚ if the sled is at top center and 66˚ if the sled is in the lowest corner.
The internal chain angle for the Top Mounted design is 10˚ when the sled is in the lower corner, which would cause interference.
Except, we never measure from the bit to the sprocket under tension.
We position the sprocket at a known position and have it play out X amount of
chain with no tension, then we attach that to the sled and enter a number for
how far this point is from the center of the bit.
so if it’s not under tension when it’s measured, we have problems
and you can’t measure to 1/64" manually in any case.
the side bars are parallel to the chain, but the center bar is always at a fixed
45 degree angle, so the angle between it and the chain varies.
Good point on thickness and chain offset (offset from the workplace)…
I think half of the issues/confusion is that there aren’t good ways (or at least common across people) of describing some of this 
For my bearing point, I was trying to address what was discussed here:
http://maslowcnc.trydiscourse.com/t/throwing-my-hat-in-the-sled-modification-ring/193/234?u=bdillahu
The slop or side load in the pivot points due to only having one vs. two points of support/contact.
But either way, the @krkeegan demo has that covered, and you’re right on offset vs. thickness…
Yes!
No!
The angle between the chain and the crossbar changes depending on where the sled is, however the side arms will always be parallel to the chain.
The measurement will need to be automated for any of these designs, and as the slop will be under the normal working load when the automated measurement is made, it will be part of the measurement, yes?
Shoot, you are right.
In the 3-bar topmount design, one of the arms on each side is a tension member. These could be made adjustable while under working load to correct the slop and make the chain arms truly parallel. A turnbuckle or eccentric might provide enough adjustment. Care would be needed to make the adjustments equal on both sides, but this would make the mechanism more forgiving to build.
To add to what @blurfl is saying…
Slop in the 45˚ version should always be radial to the router bit, meaning that the slop only affects the length of that side of the triangle (the chain triangle). While this is important for good calibration it is consistent and can be easily accounted for.
Slop in the Top Mounted version actually moves the tip of our triangle off the router bit! This can probably be accounted for but I think it’s much more complicated then telling the software how long the chain is.
Technically you should be able to have 4" of slop in the 45˚ version and still get accurate cuts (after calibration).
So, that seems to really lean toward the 45deg version… I had been leaning toward the top mount version, both because of router clearance and because it just seemed to neat
With a 20 pound sled in the top-center position there will be 33.26 pounds of tension on each chain. I believe that is the most tension the system should see (not accounting for drag). That puts ~16.6 lbs of force on each pivot, max. That’s really not that bad…
Yeah, I keep going back and forth too! I like the clearance idea but I don’t know exactly why you need clearance (I don’t have a Maslow so there could be something I’m missing) They are both pretty cool in my opinion!
Now there’s dedication - you’re doing all this and don’t even have one yet!
I got my frame pretty much together last week, and most of the unit assembled last night. It will be a day or two before I can complete things.
I am missing a router that I can use for this (what I have won’t work with the z-axis I believe. I’ll probably get a R2200, but not having that nailed down makes the clearance issue more important to me (mentally). It’s not like I couldn’t change, but you know how it is
Totally. I had considered making one with a piece of all-thread as the bottom member so I could adjust the slop away… There should be zero torsion on that member so it could be made of anything that won’t stretch (chain, cable, wire, rebar, a frozen fish, etc).
I modeled a possible top-mount sled attachment. This has the benefit of being removable for experimentation and easier to align (and it looks like a ship from Star Trek).
Here’s a link if you want to play with the model. The sled is 18" (mm due to scaling issues, but same diff) and the “router” in the middle is 7".
None of the bar (or triangle) lengths matter much, as long as all holes are the same distance as their same-colored peers.
Adjustable linkages were mentioned in an above post, and if the 3-part blue arms were bolted together with the internal holes elongated, we could possibly have the adjustment that @blurfl and @pillageTHENburn are talking about.
Here’s the adjustable 3-part bar modeled. Tight bolts and friction would (hopefully) keep them from slipping. A removable tacky adhesive could be added between layers to prevent slip, if needed. Maple syrup would work in a pinch 
I am in the process of building a Maslow (currently gathering all components), and was also thinking about the math involved with the changing center point with the current design. Has anyone tried mounting the chains to a single point on the sled? Say 4" to 6" directly above the router bit (on the Y axis). This would make the center point easy calculable triangles, and you only have to offset your workpiece the same distance that’s between the mounting point of the chains and the router bit. Only concern I have is how stable the router bit will be during movement with only one mount point and the two bricks. I guess best way to find out is build it and try it. I will be adding a thin sheet HDPE (1/4") to the bottom of the sled to reduce drag. Any thoughts? Thanks
If that worked, you could make that a software configuration and not have to offset the workpiece…
The flip side is I think you may get some drag from the bit cutting itself (hits a hard spot, whatever) that would/could cause issues.
Note I’m not speaking from experience yet, so…
Not just drag from the bit, but also from the sled itself. Horizontal accuracy would take a huge hit with a single mounting point for the chains. That’s why we’re focusing on methods that will hopefully give us the benefits of a single mounting point (easy math) without the loss of accuracy.
Definitely going to be on my “things to test” list. I’m also planning on using stepper motors and counter balances (so the steppers don’t have to lift all the weight). I’m going with stepper motors because I already own half a dozen Vexta PK296-03AAs that should be more than enough for the job. Coupled with 15 tooth .25" pitch sprockets and a 16 micostep stepper motor driver I could theoretically get a resolution of 0.001171875" per microstep. Even if I go with full steps I’ll get a resolution of 0.0175", which is more than enough for any projects I have planned.
