I am doing a related project that requires some extensive professional CNC work and I discovered, to my pleasant surprise, a large scale aluminum boat builder in my area that does CNC work. They have both a 8’x30’ (!) Multicam router and a 6’x25’ Flow waterjet router, and they sell machine time for smaller jobs on the side when they are not cutting boat hulls etc. Haven’t run a project yet but talked to one of the senior folks there and they seem quite reasonable.
Since I am going to be cutting things anyway, I was thinking that maybe I should get a nice sled cut for my Maslow build at the same time. I was wondering what the current recommendations are for idea shape and material and details. They principally work in aluminum but will cut MDF and wood and other materials. Is aluminum a viable choice for a sled?
A theory here (backed a little with observation):
As the sled base I went from a heavy 21mm c-plex to a light 8mm acrylic sled.
My effort of bringing the centre of gravity in Z-direction down with a 150mm c-beam and mounting the z-motor at the bottom was compromised by this decision. Sled #4 would be steel with a PTFE (or similar) film, if i would build it.
The theory is to bring CG-Z as low as possible to allow the chain-mount on the sled to be low. As a result the distance to move the motors out becomes less. The closer the motors get to the top beam, the less flex is to be expected was the unproven thought behind this.
I would be careful about saying that a steel sled needs no additional weight. The design of the sled to allow this must be such that the bulk of the weight is below the router otherwise it will tend to spin and can cause issues. I would say that using a steel sled reduces the amount of extra weight needed, but unless you design your sled like this you are likely to need some weight to keep it vertical. Note that the steel used in that design is 10mm thick.
The metal maslow also uses a c-channel z-axis which does contribute weight to the bottom of the sled, but I found that this was not enough to keep the sled oriented properly and to counteract the force of the bit driving into the work piece.
Edit I did look back a the metalmaslow design and it is bottom weighted. I also saw this in their initial descriptions, though I don’t know if they’ve designed away from the need:
Are you saying the metalmaslow sled will not work…not stay oriented properly? I think (and hope) not but the way you worded that could be misinterpreted.
I am not saying that. The metal maslow sled is not a circle centered on the bit, if I recall properly, so their weight distribution may allow it to orient. It may still require some additional weight to counteract drilling forces, but that can also be solved by going slower.
Just wanted to make sure that folks are taking weight distribution into consideration. I don’t remember who it was now, but someone looked at radially symmetrical weighting and found that it had a lot of problems. If one were to simply cut the standard sled from steel, they would encounter these problems as well (my sled is essentially the standard sled, just from steel with uhmw on the bottom) even though the weight might be as heavy as others’ sleds.
Thank you for clarifying that. I’m trying to be more conscious of how others might interpret the statements I make here.
I agree. Keep up the good work.
I think I know what this means/looks like but not sure I fully understand. I had a question about the placement and rotational orientation of the bricks while I was building my sled and I don’t think I found an answer so I just eyeballed it. I have recently wondered what an additional weight, or maybe some percentage of added weight, at the bottom center of the sled would do.
EDIT: I thought it was good to have the bricks at the same general angle as the chains, while sled is at 0,0, and that is how I eyeballed it…looks like this and works well enough, I think.
Perhaps @dlang or @Gero can illustrate the purpose of the brick location and rotation?
My understanding, and subsequent experience, is that the bricks/weight need to bias the sled vertically as much as possible, this means that the weight needs to be below the bit. The lower you get that weight on the sled, the more effective it will be in countering rotation. In other words, if you have a circular sled and you put the weights at 100 degrees and 260 degrees (assuming straight up is 0 degrees), the weights would not be as effective as if you put them both at 180 degrees. This can be understood by tilting the sled, the additional height of the left weight is somewhat counteracted by the weight on the right having it’s own moment arm around the center. If both weights are at 180 degrees, any tilt will cause the weight to want to right the sled.
What you did will work perfectly well.you could also put the bricks so that they are vertical and it would work very slightly better since the center of gravity of the bricks would be closer to the vertical center-line of the sled and would thus have a smaller moment arm. But, honestly, I doubt the gain is worth the effort compared to how you have it set up now.
Disclaimer (TL;DR accepted):
All of this is observation and crazy made up theory with no claim on truth!
I did a lot of (extensive) ‘weight and balance’ tests with my new sled. I’m holding back, because i don’t trust my observations yet, due to the weightless 8mm acrylic base. As far as i can remember, it was @blurfl that tested distributed weight with bad results, but wasn’t that with brackets? With quadrilateral the twist angle around the Z of the sled is part of the math (we measured the distance from the bit to the centre of gravity in Y direction in an attempt to get that right).
With triangular, I can tell that a long ‘pendulum’ weight down the middle does not work well (tested).
A drag (bottom friction) is introduced that can be compensated with raising the chain mounts, but who wants that
1 of the simpler tests i did was:
Dry run horizontal along the top of the sheet with the top of the sled at same height as the sheet edge (full sled on the sheet). The same mirrored at the bottom. With a sled balanced in the middle of the sheet (i mean putting a rod horizontal under the sled in line with the raised bit and it does not tilt in any direction), moving from left to right shows a slight drag at the top (top falls behind/top-friction), while left to right at the bottom shows a slight drag at the bottom (bottom falls behind/bottom friction). If your sled shows top-friction at the top and the bottom, your chain mount is to high. Opposite, if it shows bottom friction on both runs, your chain mount is to low.
I used almost a complete roll of double sided tape and tested a series of ‘flat steel arrangements’ in various combinations. What i ended up doing was to put the sled on the table and put a rod under it in X direction in line with the bit and shifting the 2 outer weights i have upwards until the sled balanced on the rod. I do have weight down in the centre, what i believe still keeps the sled level. So does the upside down ~V i did with the outer weights i guess.
Compared to my overweight ~16kg Bosh GOF2000PE setup, i was able to cut the chain mount height more then half with this ~8.5kg (might not be enough) sled. With a steel base i could have done better.
With 16kg i could go down to a frame angle of ~9° before seeing the effect that there was not enough force in Z direction for the bit to go in. With 8.5kg i’m back to 15°. Tweaking all this variables is quite a task.
Lighter weight is better for future speed.
How can we motivate a larger group to upload their Maslow-specs and cut the same file so the collected data would help big time? Does it already fail because similar calibration results would be a must? Several attempts from various members failed till date, as you would need lager numbers than a handful to get usable results and conclusions.
Respect if you made it until here
Kind regards from the turning to dried meat (55°C/131°F expected next week), Gero
How many manage to get this with still an error +/- 0.5 mm right, with a temporary sled, cutting the ‘final’ one, with a questionable calibrated machine? We might loose allot here already, giving up because the results don’t match the expectation? Non of my sleds are cut on the Maslow, for a reason.
If i could vote, i would vote the above as #1 reason.
@gero, if the bit is centered in the triangulation kit, the sled rotating should not matter [1]
one test that makermade does is to run the bit into the wood, then manually grab the head of the router and rotate the sled side to side. Then withdraw the bit and look at the hole. If it’s enlarged, then the bit isn’t centered.
[1] the sled rotating matters when you hit the limits of your triangulation kit and the chain needs to flex rather than remaining pointed directly at the bit. At one time I didn’t understand this and advocated a balanced sled, someone built one and showed the problem of the motor torque spinning it against the limits of the triangulation kit and the distortion this caused.
Out kit comes with two 4x12" by 1/4" thick steel bars. One can use them on the sled, but I suspect most people will use them as counter weights for the chains since the sled alone seems to work well.
If they do this with the router hanging on the chains, a person can inadvertently pull on the router/sled and obscure the results. I did this test as well, at the advice of Jay_Settle. I may have influenced the cutter; hard to tell. As you can see from the irregular hole, installing the router and linkage is no joke. If you read through my posts you can see how carefully I tried to do it the best way possible (using a drill press) and still ended up with a distorted hole. This is another reason why I ordered metalmaslow’s kit; thinking that, being laser cut, some of the critical holes and dimensions are done for you).
This all ties back in to what I was saying yesterday about kit makers doing a bit more to prepare, or even pre-assemble (for a fee), critical components like this and also what gero was saying here:
looking over your post (with the updated images), you did pretty much how I intended the kit to be mounted. while your resulting hole is slightly enlarged, it’s pretty close (I’ve seen some posts where the hole becomes a 3/4" wide smile)
I agree that a laser cut sled is going to be more accurate than anything done by hand, however careful you are.
while making a temporary sled and then a permanent one is a nice bootstrap, I think it’s more trouble than it’s worth.
My suggestion would be to
mount the router to approximately the center of a hunk of plywood, screwing it down.
Drill a center hole with the router
remove the router
mount the router to some scrap of wood or cardboard (or even stiff paper if you are very gentle with it)
put a hole in your scrap around 9" from the bit.
use this scrap as a circle cutting jig to cut the sled
put another hole in your scrap about 6" from the bit
use this to cut mounting holes for your bricks at a consistent distance (or even a shallow groove and then drill the final holes manually)
repeat steps 7 and 8 with the right distances for the ring mounting holes (twice, once for each distance, I don’t know what distances these should be) with a smaller bit
enlarge the center hole (does not need to be precise, freehand with the router works, as does a hole saw)
remount the router to the holes made earlier, and you have an accurate sled, no CNC work needed.
other than the ring mounting, none of the distances are critical, you just want to have everything centered on the bit.
Is aluminum a viable choice for a sled?
You could, but why? aluminum is about 3 more expensive than steel and about 1/3 as dense. Unless you absolutely cannot find 1/4 or 3/8 steel plate nearby, check local fabrication or welding shops or local ebay/craigs list sales, I would say steel is the way to go. Thinner sled and lower center of gravity.
