60 Degree tilt, vertical Lowrider? Also Yeti's and feeds and speeds questions

[jwoody18 (Jeff)]

Hi Everyone,

Hat tip to Ryan for all the amazing engineering he’s done, especially on the LR2. And Ryan, I love the fact you have the shop and offer the printed parts kit. I don’t have a 3D printer but the way you package and bundle things makes it an easy ordering decision once I learn a few more details.

I’ve been following the forum for a month or so and finally joined to ask some questions:

VERTICAL LOWRIDER?

(Note to Ryan: I suspect this was intended but I can not read Lowrider here without hearing “Low-ride-er” in that voice in my head. Nice one)

I purchased a MaslowCNC about a year ago, but life and a move intervened and I haven’t had a chance to build it. One of the big appeals of the Maslow is that it runs vertically (well at a 12 - 15 degree tilt off vertical) so it can operate in a fairly small footprint volumetrically. I think its about three feet of footprint. Alas, due to the chain driven nature, the current iteration has accuracy and repeatability issues in the edges and top center of the sheet. But it got me wondering, could a modified Lowrider run on an easel type “table”?

I found this the other easel/vertical CNC the other night when googling around: https://buildyourcnc.com/item/cnc-machine-greenLean-v1 and it looks like they have a V2 version of it, V1 and V2 photos attached.

If the LR2 was modified so the skate wheels were enclosed in a track of some sort, would it work vertically or would gravity likely throw things off? If it was vertical, I could probably leave it set up. What about running on more stainless pipe for the X axis with the “table” in a slanted vertical orientation?

YETI COMMERCIAL DESIGN

I’m also really intrigued by the new www.YetiTool.com Smartbench. Unclear if it can cut in a vertical position or if they are just putting it vertical for trade show display, but otherwise a similar but different approach to a walking CNC like the LR2: https://www.yetitool.com/products/product-details/smartbench#VideoMainContainer

No real reviews in the wild yet of the SmartBench yet as they are just shipping later this month. Intrigued to see where it goes. They just posted this background on how it works: https://www.youtube.com/watch?v=FIZJTIuOIrU&feature=youtu.be

That being said, at $6K it’s a very different beast and I can’t see how it’s materially more capable than the LR2 so it will be interesting to watch as more details emerge. It is the one device I’ve seen that is closet to the LR2.

FEEDS, SPEEDS, ACCURACY, PRECISION & REPEATABILITY

More generally I was also wondering about the specs of the LR2.

From what I have been able to glean from the forums, it sounds like a feed rate of up to 15 mm/s (~35 inches/minute) is reasonable at the high end. The Shopbot at my local Makerspace likes us to run plywood at ~70-80 inches/minute, so it’s pretty amazing the LR2 can match half of that for a fraction of the cost.

Is it viable to run 3/4" plywood in the LR2 at 15 mm/s with say a 1/3 (6.3mm) depth of cut using a compression bit?

How accurate and precise is the LR2? For example, would it be viable to make a Festool MFT style top on the LR2 where the goal is to have 20mm dog holes on a tight 96mm grid over a 3 x 5 foot surface? Is accuracy down to 1mm or .5mm or .1mm? https://www.festoolcanada.com/products/workplace-organisation/multifunction-table/495315---mft3

And how repeatable are the operations between runs? If I have to stop between toolpaths on a job and set the LR2 aside for a bit then put it back on the table and continue a project that is half cut, is that viable? How do your re-zero it accurately? Also, following the Festool MFT example above, if I ran the same file on the same kind of stock on two different days, how closely would they match each other?

MAKITA RT0701C PRINTED PARTS

I already have an RT0701C that I love, and I see someone has posted a spin of the router cradle and a few other necessary bits for adapting the Makita to the LR2. Is it possible to order a printed materials set with these subbed in (or added on) for us Makita fans?

ONE STOP SHOP (Almost?)

Is it correct that if I buy the LR2 CNC Bundle, the printed parts (maybe with Makita sub in?:)), and I CNC the flat parts at my local maker space - then all I need to do is source the SS tube and with my router in hand I would have everything to complete the build?

I see you offer an LCD controller for just $15 (Amazing that is possible at that price). Any other goodies you would suggest to add? Does the e-stop just wire right to any of the Rambo boards?

RAMBO MINI or RAMBO V1.4

I see the CNC bundle includes the Mini. Is it possible to pay an upcharge and get the RAMBO V1.4 board? And what benefit does the Rambo V1.4 bring to a new build?

Thanks in advance for the details!

-Jeff

 

[jeffeb3]

@jwoody18

If the LR2 was modified so the skate wheels were enclosed in a track of some sort, would it work vertically or would gravity likely throw things off? If it was vertical, I could probably leave it set up. What about running on more stainless pipe for the X axis with the “table” in a slanted vertical orientation?

This would be pretty awesome, but it’s uncharted territory. The Low Rider is based off of the crawlrbot (sp?) and to make it run vertical, I would guess the biggest problems are 1) The router will always be pulling toward the ground in the Y direction, so you might need more torque in the Y and 2) You need to put something on the top side to hold the weight for the X axis so it can still move slowly and won’t translate in the Y. Something as simple as another set of wheels against the side might do it.

The Z is lifted by the same motors, but they are lifting using a leadscrew which will give it much more turns/mm (about 4x) which is enough to compensate gravity. The belt and pulley on the Y might have to be replaced with something with more turns per mm. A pulley maybe? Backlash shouldn’t be a problem since it will be pulled to one side of the backlash by gravity.

That’s my arm chair engineering, at least. I’m a software guy.

@jwoody18

Is it viable to run 3/4″ plywood in the LR2 at 15 mm/s with say a 1/3 (6.3mm) depth of cut using a compression bit?

I would say that’s reasonable, but there are no guarantees. The material you’re cutting in, the bits, and how you’ve set up the machine can make a big difference in those statistics, which makes them really hard to specify. I haven’t even used a compression bit either, but those numbers seem reasonable for a single flute upcut bit. As you get more experience, you can push it a little farther, but it’s a compromise between risking your job and just taking your time. As you get more experience, you’ll be able to tell easier if you’re just cutting too hard, or something else is wrong with the machine, which will give you confidence to try faster/deeper cuts.

@jwoody18

How accurate and precise is the LR2? For example, would it be viable to make a Festool MFT style top on the LR2 where the goal is to have 20mm dog holes on a tight 96mm grid over a 3 x 5 foot surface? Is accuracy down to 1mm or .5mm or .1mm? https://www.festoolcanada.com/products/workplace-organisation/multifunction-table/495315—mft3

I’m doing estimation here, and the quality of your CAM skills and the quality of your bits will make a big difference. But, here’s my guess:
Getting the holes 96mm apart, along one axis, is no trouble at all. The belts are going to determine that precision, and they are made very precise. Getting exactly 20mm holes is a little harder, but if you do a finishing pass (which basically cuts a 19.9mm hole and then travels around the last 0.1mm in one smooth kiss with the bit), it should be very good. The common errors are more in squareness. You’re going to want your holes to be circles, not elipses, and you will want the grid to be square, not parallelograms. IMO, having a 5 foot long gantry makes that pretty easy. If you build a nice square table, you’ll have a great reference to start the gantry against one end, and since both sides are moving in lock step, it will stay square. How close can you get the gantry if you’re physically pulling it against a square hard stop, 0.1mm? And you’re talking 0.1mm over 5’… Even with a tape measure, you can get within a few mms, and over 5’, that’s still pretty darn square.

@jwoody18

And how repeatable are the operations between runs? If I have to stop between toolpaths on a job and set the LR2 aside for a bit then put it back on the table and continue a project that is half cut, is that viable? How do your re-zero it accurately?

I set mine up with hard stops. I clamp a scrap piece of wood on the end of the table and pull the gantry against it when I energize the motors. The Y is all the way against one side, and the Z is at the same height because it’s dropped by gravity. I think it’s pretty repeatable.

@jwoody18

Is it correct that if I buy the LR2 CNC Bundle, the printed parts (maybe with Makita sub in?:)), and I CNC the flat parts at my local maker space – then all I need to do is source the SS tube and with my router in hand I would have everything to complete the build?

How are you going to get the makita parts? Maybe your local maker space also has a 3D printer you could use?

But I think that’s right. You also need a table.

@jwoody18

I see the CNC bundle includes the Mini. Is it possible to pay an upcharge and get the RAMBO V1.4 board? And what benefit does the Rambo V1.4 bring to a new build?

The mini rambo has 4 drivers. The rambo has 5. The benefit of the larger controller is you can drive each motor on it’s own driver. The benefit of that is that you can put an endstop on each one, and when you home, if one motor has to travel a little farther to get to the switch, it can. This is the “Dual endstop” setup, which is really “Dual motors” and “Dual endstops”. Doing Dual endstops on the LR isn’t as turnkey as doing it on the MPCNC, so you’ll have to be comfortable with configuring Marlin and/or electronics to do this (or be willing to get comfortable).

[jwoody18 (Jeff)]

@jeffeb3

This would be pretty awesome, but it’s uncharted territory. The Low Rider is based off of the crawlrbot (sp?) and to make it run vertical, I would guess the biggest problems are 1) The router will always be pulling toward the ground in the Y direction, so you might need more torque in the Y and 2) You need to put something on the top side to hold the weight for the X axis so it can still move slowly and won’t translate in the Y. Something as simple as another set of wheels against the side might do it.

Jeff - first thank you for the detailed response, I really appreciate it.

I see in another thread about fold-up tables that there is a feeling it’s probably not viable but I really appreciate your walking through some of the details. I was curious about the effect of the gravity pull in the Y axis and whether or not there was enough juice in the motors to overcome that. Your later point about using mechanical advantage in the drive system was intriguing too, to get more oomph.

I think I’ll stick with a standard design and get all that sorted out before I ask much more about going vertical.

@jeffeb3

I would say that’s reasonable, but there are no guarantees.

Ok, thank you. It does give me a general sense of 3/4" plywood speed which would be my typical use. Maybe 1/2" too. I just wanted to make sure my expectations were appropriate as it is very material dependent and I haven’t seen too may people talking about 3/4. Sounds like 15mm/s (~35 ipm) well tuned (as in tool path, bit, chip load etc) is a good benchmark for me to expect.

@jeffeb3

I’m doing estimation here, and the quality of your CAM skills and the quality of your bits will make a big difference.

Modeling skills are pretty modest but have been steadily getting better. CAM is a new area but just did a fairly ambitious project and forced myself to learn Fusion 360 enough to get started. Still very much learning the basics but had some great feedback from the owner of the Makerspace on how to improve toolpaths during a discussion yesterday so my knowledge is growing a little each day. I’d just like to have it in my own shop with the LR2 rather than have to drive an hour or more each way (more with traffic) to use a machine…

@jeffeb3

IMO, having a 5 foot long gantry makes that pretty easy. If you build a nice square table, you’ll have a great reference to start the gantry against one end, and since both sides are moving in lock step, it will stay square. How close can you get the gantry if you’re physically pulling it against a square hard stop, 0.1mm? And you’re talking 0.1mm over 5’… Even with a tape measure, you can get within a few mms, and over 5′, that’s still pretty darn square.

Good feedback on the hole setup and the finishing pass. On the ShopBot I had trouble initially getting the diameter spot on, it came in under but a second pass did get me there. Your suggestion is basically what worked on the second piece I did, two passes and finished it accurately on the final way through.

I’m not quite sure I get what you are saying with the gantry, but I think it’s just a perspective thing. I’m so used to 4’ being the width discussed, I wasn’t thinking creatively enough.

I believe you’re suggesting building the gantry wide enough to enable a 5’ Y axis since the X Axis with the LR2 is limited only by table length and the Y-axis accuracy is dictated by the accuracy of the parallel long table edges which should be more easily achieved? So that would give a nice sharp repeatable Y for the grid.

It sounds like the X would already be ok, but I’m infering that. Is this due to the dual motor design so it’s easier to hold square?

@jeffeb3

I set mine up with hard stops. I clamp a scrap piece of wood on the end of the table and pull the gantry against it when I energize the motors. The Y is all the way against one side, and the Z is at the same height because it’s dropped by gravity. I think it’s pretty repeatable.

This is wonderful news. Thank you for explaining how you do it.

@jeffeb3

How are you going to get the makita parts? Maybe your local maker space also has a 3D printer you could use?

I’m afraid I don’t have any 3D printing experience yet so if Ryan could offer an upcharge to print them an include them with the kit I’d like him to benefit. If he’s not interested I’d probably try and find a service bureau to run it for me. That’s a solvable one.

@jeffeb3

But I think that’s right. You also need a table.

I was pondering this one. I use some Ikea kitchen islands now for work benches. It’s not the right dimension but out of curiosity I was wondering if that would be fine? It’s got a nice square CNC factory edge. Picture attached.

I’d love your feedback on what I’m thinking of doing for a table. I thought I could cut a spoil board and put it over the dog hole table I just made (wider and longer than the table itself which is ~3’x5’). It would save making a base.

The table has a precise grid of 3/4" and 20mm holes so I was pondering cutting 1/2" MDF on the shopbot to my desired dimensions (so it will be super square edges for the LR2 to ride along) and putting some matched holes in it to register it to the far edges of my dog hole top using bench dogs. The top is sitting on two saw horses with 2x4s along the X axis to make the large frame (photos attached).

The upside being I can take it apart easily and put it away, take some or all of it to help a friend with a project etc. Of course the bench dogs would have to be outside of the cutting area to avoid any contact with the router bit but I can likely achieve that by using a 3/4" spoilboard and just insetting the dog hole pockets on the underside of the spoil board 3/8".

Ryan said something awhile ago that struck me. He said he doesn’t need a super big machine because he’s only willing to babysit a job for so long. I hear that. I also am really interested in jobs at or below the size I can cram a cut partial sheet in to my old Volvo, which is 42" x 78" or so. I don’t want to rent a truck or buy a trailer. But your point about a 5’ Y intrigues me. I guess I’d need about 7" more so 67" in that direction? and 49" on the X? (just going back to your earlier point about improving squareness by putting the longer axis on the Y).

I guess it would be ideal to stop at 49" or just under so I can get it out of an MDF dimensioned board without having a seam. That would give me a 5’x4’ area which is quite flexible I think. Big enough to do some larger things but small enough to be manageable from a space and materials handling perspective.

I know people talk a lot about tables but I haven’t quite gotten my head around it all. Is it enough to have square edges or do I need to frame those edges so they are quite a bit taller in the Z dimension? Does my “temporary” sawhorse base cause any foreseeable issues?

@jeffeb3

The mini rambo has 4 drivers. The rambo has 5. The benefit of the larger controller is you can drive each motor on it’s own driver. The benefit of that is that you can put an endstop on each one, and when you home, if one motor has to travel a little farther to get to the switch, it can. This is the “Dual endstop” setup, which is really “Dual motors” and “Dual endstops”. Doing Dual endstops on the LR isn’t as turnkey as doing it on the MPCNC, so you’ll have to be comfortable with configuring Marlin and/or electronics to do this (or be willing to get comfortable).

Thanks again, those details really help. So this is the setup that is detailed in that awesome video series recently posted by the Australian Teching Tech guy. I believe he went through some of those changes in detail, I’ll take a closer look. I’d be willing to give it a shot as the capability it adds is very appealing - especially if I’m going to regularly be removing the LR2 to do other things with the table.

Jeff, you’ve been really really helpful and you’re a great writer. Thanks for explaining things in detail, it’s really got the wheels turning in my head and my excitement up for the LR2. Based on this I think I’m going to pull the trigger on the kit, I just need to figure out the additional 3D printing for the Makita router.

-Jeff

 

 

 

[jeffeb3]

@jwoody18

Ok, thank you. It does give me a general sense of 3/4″ plywood speed which would be my typical use. Maybe 1/2″ too. I just wanted to make sure my expectations were appropriate as it is very material dependent and I haven’t seen too may people talking about 3/4. Sounds like 15mm/s (~35 ipm) well tuned (as in tool path, bit, chip load etc) is a good benchmark for me to expect.

You can’t talk about speeds without DOC. So 15mm/s and 6mm DOC is doable (with the stipulations I had before, maybe not the first settings to start with). That’s pretty close to 8mm/s and 12mm DOC or 4mm/s and 20mm DOC (although there can be other troubles going this deep in one pass). Ryan’s advice, that I remember is that about 8mm/s is a sweetish spot, and if you want to push harder, go deeper first, until you’re hitting the depth of your material, or you are running into chip evacuation problems or something. 15mm/s is doable though, but you’ll need to go shallower than if you’re going at 8mm/s.

When I say skills with CAM, that’s a bit of a mistake. It’s not the familiarity with the software as much as seeing where the cuts are going to fail in the software, adjusting settings, including DOC and speeds, but also pocket strategies, or adaptive milling, or lead ins, or whatever. Then it’s also watching it run, listening to it run and knowing what to change to stop wasting time, or cut a little more aggressively. After using it a bit, you’ll get your speed up, but it’s much safer if you’re a novice to just start slow.

@jwoody18

I’m not quite sure I get what you are saying with the gantry, but I think it’s just a perspective thing. I’m so used to 4′ being the width discussed, I wasn’t thinking creatively enough.

A 4’ working width means something like 4’8" long pipes, and a little more stuff on each side. So I call this a 5’ gantry.

@jwoody18

I believe you’re suggesting building the gantry wide enough to enable a 5′ Y axis since the X Axis with the LR2 is limited only by table length and the Y-axis accuracy is dictated by the accuracy of the parallel long table edges which should be more easily achieved? So that would give a nice sharp repeatable Y for the grid.

It sounds like the X would already be ok, but I’m infering that. Is this due to the dual motor design so it’s easier to hold square?

X is the direction on the roller wheels, right? Y is the single motor moving the router across the gantry.

Either of them are very accurate along their dimension, because you’re stepping a motor on a timing belt, and it’s accurate enough to not be the bottleneck on accuracy.

But to make squares and not parallelograms, you need those two axes to be square. So you need the gantry to be square to the belts driving X. The two motors will move in lockstep, so you need to start them perpendicular. Once you do that, they will stay square.

The X axis runs right along the sides of my table. My table is very square (as square as I can measure). So to start my gantry square, I just need to start it parallel to the end of my table. To do that, I clamp some boards on the end of my table, pull the gantry into that board, and when that’s where I start.

There are lots of ways to make physical hard stops. Some people just pull it to the end of the belts, or there are jigs for this, but they all amount to starting the motors square, and they will track square (as long as nothing goes wrong).

@jwoody18

I’d love your feedback on what I’m thinking of doing for a table. I thought I could cut a spoil board and put it over the dog hole table I just made (wider and longer than the table itself which is ~3’x5′). It would save making a base.

IMO, make a simple table first, 2x4s and sheet goods. You can always make it again, and now you’ll have the LR and a specimen to experiment with. I think there are probably as many tables as there are LRs.

Make the smaller dimension Y (the pipes) because that will increase rigidity exponentially. The smaller you can make Y the better.

[jwoody18 (Jeff)]

@jeffeb3

When I say skills with CAM, that’s a bit of a mistake. It’s not the familiarity with the software as much as seeing where the cuts are going to fail in the software, adjusting settings, including DOC and speeds, but also pocket strategies, or adaptive milling, or lead ins, or whatever. Then it’s also watching it run, listening to it run and knowing what to change to stop wasting time, or cut a little more aggressively. After using it a bit, you’ll get your speed up, but it’s much safer if you’re a novice to just start slow.

Understood. But always a welcome reminder that there are a lot of variables at play here and a whole world of knowledge to amass.

As someone said to me a long time ago, first you get good then you get fast. I wasn’t looking to leap right to that, only to understand what the machine could do once I get that knowledge, so I can get a sense of how it could fit in the shop.

@jeffeb3

A 4′ working width means something like 4’8″ long pipes, and a little more stuff on each side. So I call this a 5′ gantry.

Ah ok, so there is no need to make the Y wider. Since I am ok 42" usable area, I would be around 48" total length with the long pipes, correct? [edit: meant to say 50" total, but to keep it in sheet size would probably drop the short axis to 40" + 8" for traveling components on both sides, keeping it on the sheet]

@jeffeb3

There are lots of ways to make physical hard stops. Some people just pull it to the end of the belts, or there are jigs for this, but they all amount to starting the motors square, and they will track square (as long as nothing goes wrong).

Ok I’m, seeing what you’re saying now. Hold it square and energize it and it will continue to hold square from that point using motor power. I am guessing this is something you need to do any time it loses power? But not between tool changes for example because it won’t lose power, only if you power it down and remove mains power, correct?

I’m guessing some tall dogs in holes in the spoil board can give me two or more points of contact to manually square it. Or if I do the double end stop option that will do this automagically instead?

@jeffeb3

Make the smaller dimension Y (the pipes) because that will increase rigidity exponentially. The smaller you can make Y the better.

Good to know. Makes sense, will do it that way.

And how much extra distance do you need in the long axis past the target working area so the machine can work right to the edge of the target working length?

-Jeff

[jwoody18 (Jeff)]

Just found the calculator, off to explore the dimensions in more detail. Sorry for asking questions that are already answered on that front!

-Jeff

[jeffeb3]

Yep, yep, yep. You’ve got it now. There is a ton of info if you read through all of Ryan’s instructions. I just like looking smart in the forums and I hat saying RTFM.

[barry99705 (Barry)]

The build your cnc lean machine uses some beefy springs to hold the weight. I forget what they’re called, but they’re the same kind as in a tape measure, just bigger. It’s also a much larger and beefier, and expensive machine.