Squaring the Gantry on a CNCRP Pro

[GME]

I am looking at purchasing a 4848 Pro, but I have concerns about squaring the gantry. The CNCRP website speaks to squaring the gantry, and suggests adjusting the homing sensors until the gantry is square. I'm not a fan of this technique. If you uses homing sensors, you are essentially forcing the gantry to give square readings when measuring diagonals, but something is necessarily out of wack somewhere. As they say, 'somethings gotta give.' If it help in figuring out where I'm coming from, I had the gantry on my present machine squared to less than .001", before I recently took it apart (and that's a whole other story).

In looking through the assembly instructions, it does not appear that there is much, if any, adjustability. On the other hand, I don't find any discussions about gantry squaring, when searching the web, including here. To me, it looks like (1) folks are just using the homing switches and calling it good, or (2) there is enough play to get the gantry squared mechanically, so no one feels the need to complain or comment on it.

What's the real scoop? Is there mechanical adjustability? If not, has anyone found a workaround, like modifying the interface plate (e.g., slotting the mounting holes) and using shims on the joining plate?

Thank you in advance for your help!

Gary




Similar Threads:

• Problem- Squaring the x axis gantry
• Gantry squaring questions?
• Squaring the Gantry to the X-axis
• Squaring a gantry - without a square?
• Squaring the gantry

[GigaWatt]

When I received the extrusions from 80-20 for my Pro4824, I spent time checking each piece with a machinist square and feeler gages. Each part was cut very accurately as to length and squareness. I spent a lot of time using squares, clamps and measuring to insure the base was as square as possible during assembly. When finished the gantry was less than 1/64" out of square. I made a very slight adjustment during the squaring procedure. Since I do mostly woodworking that's as far as I went. That was 2 years ago and the cnc still cuts square enough for me. I don't think you will have any problems although I'm sure there are members here that do far more accurate work than I who will give their opinion.

[dgage]

I use my 60120 CNCRP Pro machine to cut home theater subwoofers. I cut 2 sheets of 60x60 Baltic Birch plywood at the same time. My subs need large circles and squares cut so having the machine square is critical. I squared the machine as they said with 46” drilled holes and measured the diagonals and adjusted the sensors as needed. Once dialed in, my machine cuts perfect. So maybe this is one of those concerns that doesn’t translate to the real world with the CNCRP machine. But if you did have issues, I’m confident CNCRP will take care of any issues as they have with me and others.

[GME]

When I received the extrusions from 80-20 for my Pro4824, I spent time checking each piece with a machinist square and feeler gages. Each part was cut very accurately as to length and squareness. I spent a lot of time using squares, clamps and measuring to insure the base was as square as possible during assembly. When finished the gantry was less than 1/64" out of square. I made a very slight adjustment during the squaring procedure. Since I do mostly woodworking that's as far as I went. That was 2 years ago and the cnc still cuts square enough for me. I don't think you will have any problems although I'm sure there are members here that do far more accurate work than I who will give their opinion.

I use my 60120 CNCRP Pro machine to cut home theater subwoofers. I cut 2 sheets of 60x60 Baltic Birch plywood at the same time. My subs need large circles and squares cut so having the machine square is critical. I squared the machine as they said with 46” drilled holes and measured the diagonals and adjusted the sensors as needed. Once dialed in, my machine cuts perfect. So maybe this is one of those concerns that doesn’t translate to the real world with the CNCRP machine. But if you did have issues, I’m confident CNCRP will take care of any issues as they have with me and others.

Thanks for your responses. Both of you tend to confirm what I suspected.

I have two reasons for not wanting to square with the homing sensors. 1. I want to do inlay work, and 2. I will be doing dovetails off the front of the table using JointCAM. Both call for a high degree of accuracy. I'm not confident that using homing/limit sensors will give me sufficient accuracy.

I have engineers squares and a 36" straight edge good to .0002' in 12". I also have a Woodpeckers 45/45/90 18" triangle with .001" tolerance to help with an accurate set up. On my present machine, I used shims as small as .001" to ensure crossmembers were square. All this helps to minimize an out-of-square condition, but having to make some adjustment is probably inevitable. I still had to make adjustments on my present machine.

Would appreciate if someone could tell me about mechanical adjustability.

[ger21]

I have two reasons for not wanting to square with the homing sensors. 1. I want to do inlay work, and 2. I will be doing dovetails off the front of the table using JointCAM. Both call for a high degree of accuracy. I'm not confident that using homing/limit sensors will give me sufficient accuracy.

I would expect that squaring by using the sensors would be easier to setup than mechanically squaring, provided that it's pretty close mechanically.
I don't really know how much flex a CNCRP gantry has in regards to racking. But again, I think that it should be pretty easy to assemble it very close to square without a lot of measuring.
I think that you are underestimating how well homing to two separate switches works. We're talking a small amount, hopefully less then .03" or so worst case.

If it's pretty close, a little out of square won't affect dovetails. My machine is out of square a bit. (Enough that I know it's out of square by comparing parts...) and it has no noticable effect on the joints I've cut. I've cut every type of joint that JointCAM can make while testing during development.

[dgage]

[QUOTE=GME;2235478]I have two reasons for not wanting to square with the homing sensors. 1. I want to do inlay work, and 2. I will be doing dovetails off the front of the table using JointCAM. Both call for a high degree of accuracy. I'm not confident that using homing/limit sensors will give me sufficient accuracy.[/QUOTE]

I will be doing dovetails too, I just haven’t had time to use the jig I’ve built. Should do it within the next month and I have no concern on accuracy. And even for inlays, you square it up once at the beginning of the day and shouldn’t need to square it again until you turn the machine off or have a mishap such as material not held down well. Usually I square it up once and then cut on it until the end of the day as the motors will be energized the whole day and won’t lose position normally. So I don’t see an issue. If you’re that concerned about accuracy then you need to be looking at a servo controlled machine with constant movement feedback.

[he1957]

Are you referring to squareness of the gantry (X) to the rails (Y) (ie: X perpendicular to Y) or squareness of the table/slots to X and or Y linear movement? The table itself may not be square/perpendicular to X and Y travel but as long as the X and Y are perpendicular to each other it should cut square with the spindle trammed to the surface.

Or am I missing something?

[islaww]

Build the gantry as square as you can. Use the homing switches to get it perfect. 1/1000th at a time, if needed. That's the way that all but the most rigid gantry machines are squared, an extrusion machine is surely not among them.

[GME]

Are you referring to squareness of the gantry (X) to the rails (Y) (ie: X perpendicular to Y) or squareness of the table/slots to X and or Y linear movement? The table itself many not be square/perpendicular to X and Y travel but as long as the X and Y are perpendicular to each other it should cut square with the spindle trammed to the surface.

Or am I missing something?

No, you are not missing anything. I'm referring to the relationship of the X axis (gantry) and the Y axis, i.e., whether they are perpendicular to one another. If I want my table to be square to the axes, all I need to do is run a cutter along 2 edges. It is sometimes useful lining up stock for cutting, like when cutting dovetails. Also, having an edge parallel to the Y axis can be useful when running tiled toolpaths.

I am familiar with tramming to a properly surfaced spoilboard. I use an Edge Technology Pro Tram and a thick piece of plate glass to get the Z axis perpendicular to the spoilboard surface.

Gary

[GME]

[SIZE=4][FONT=tahoma]


I would expect that squaring by using the sensors would be easier to setup than mechanically squaring, provided that it's pretty close mechanically.
I don't really know how much flex a CNCRP gantry has in regards to racking. But again, I think that it should be pretty easy to assemble it very close to square without a lot of measuring.
I think that you are underestimating how well homing to two separate switches works. We're talking a small amount, hopefully less then .03" or so worst case.

If it's pretty close, a little out of square won't affect dovetails. My machine is out of square a bit. (Enough that I know it's out of square by comparing parts...) and it has no noticable effect on the joints I've cut. I've cut every type of joint that JointCAM can make while testing during development.

Thanks, Gerry. Actually, I'm not underestimating how well homing to two sensors works. I understand it, and have done it. I just prefer to do it mechanically, if possible. I also have the CNCRP NEMA 34s, so I have enough muscle to rack the gantry, as least a small amount, if necessary. What a smallish "auto square" would do to linear rail wear is an open question.

Good to hear that I needn't be concerned about dovetails. I know that Z cut depth is critical to a dovetail joint fitting (or not), just like it is when using a handheld router and dovetail jig. I have firsthand experience from cutting dovetails on my Leigh D4 jig (the older model of the D4R Pro).


Gary

[GME]

Build the gantry as square as you can. Use the homing switches to get it perfect. 1/1000th at a time, if needed. That's the way that all but the most rigid gantry machines are squared, an extrusion machine is surely not among them.

If a machine is designed to allow sufficient adjustment, it's not that hard to dial one in to within one or two thousands, or better. It can be a bit fussy, and takes some patience, but it's not that hard to do. I've done it. So, if getting the gantry "as square as [I] can" means that it is possible to mechanically square to perfection, then I can accomplish what I want. However, so far, no one is saying that. That's why I asked how much adjustability the CNCRP has. What I seem to be hearing is that mechanically squaring the gantry on a CNCRP pro is, at best, a crap shoot. At worst, it isn't going to happen.

[ger21]

I'm not implying that you shouldn't assemble it as square as possible. This should always be the main priority. But imo auto squaring should also always be used as a backup, especially on a machine where the rigidity (or lack of) allows for some independent movement between the two sides.
And I don't think the small amount of squaring will have any effect on bearing wear. You'd have to approach the point of binding to see any wear, imo.

Yes, with dovetails, the depth is the most critical (along with tool size). The depth is used to adjust the fit.

What I seem to be hearing is that mechanically squaring the gantry on a CNCRP pro is, at best, a crap shoot.

Funny how people can read the same things and see something different. Again, I don't have one, but I haven't read anything to lead me to believe that you can't assemble it square. I've spoken to many CNCRP owners, and have never heard of any issues.

[wmgeorge]

Why don't you try one of the commercially made and marketed machines, as Camaster or other US made and ask the same question? I would be curious to hear what answers you would get about .001 inch over 48 inches accuracy.
https://www.camheads.org/showthread.php?t=3876

[ahren]

Hi All,

I'll jump in here. First, it is rarely actually necessary to mechanically square the system unless you have a significant deviation between the two sides of the machine. We typically recommend relying on the sensors to precisely square the system, as they are repeatable to less than 0.001", even under significant temperature changes. Please refer to our leveling, squaring, and tramming video for instructions on adjusting the sensors for machine squareness:

Leveling, Squaring, and Tramming Your CNC Machine | CNCRouterParts

While most users will find no practical advantage in doing so, if you do want to pursue a mechanical approach, it is possible to change the squaring of the gantry by shimming behind the red joining plates that connect the gantry to the risers -- see the picture below. Please keep in mind that as stiff as the machine is, over the span of the gantry it's always possible to rack the system, so you would need to "pin" your system against a hard stop every time you engage or disengage the motors to insure squareness. As Ger mentioned, you should feel free to adjust the system as square as possible mechanically, but the sensors offer a practical way to refine your machine squareness on a regular basis.

Ahren
CNCRouterParts

[GME]

[QUOTE=ahren;2235792]Hi All,

I'll jump in here. First, it is rarely actually necessary to mechanically square the system unless you have a significant deviation between the two sides of the machine. We typically recommend relying on the sensors to precisely square the system, as they are repeatable to less than 0.001", even under significant temperature changes. Please refer to our leveling, squaring, and tramming video for instructions on adjusting the sensors for machine squareness:

Leveling, Squaring, and Tramming Your CNC Machine | CNCRouterParts

While most users will find no practical advantage in doing so, if you do want to pursue a mechanical approach, it is possible to change the squaring of the gantry by shimming behind the red joining plates that connect the gantry to the risers -- see the picture below. Please keep in mind that as stiff as the machine is, over the span of the gantry it's always possible to rack the system, so you would need to "pin" your system against a hard stop every time you engage or disengage the motors to insure squareness. As Ger mentioned, you should feel free to adjust the system as square as possible mechanically, but the sensors offer a practical way to refine your machine squareness on a regular basis.


Thank you, Ahren. Knowing I should be able to mechanically square the gantry is good news. I appreciate you advice about using the sensors. My intent was to achieve mechanical square, set my sensors to such that they home at that same relative position, and use reference all home in the usual way. I am aware that steppers may not lock in at startup in the exact relative position, which I assume is why you commented on pinning against a hard stop when engaging or disengaging the motors. so the sensors serve to correct for any difference, while still preserving mechanical square. That was my intent all along.

By the way, I looked at a number of your videos, including leveling, squaring, etc. You folks produce some nice videos. It's a great value added service.

i should be pulling the trigger on a purchase in the next several days.

Thanks again,

Gary