Author Topic: Contemplating some upgrades to my CNC Router  (Read 2200 times)

Offline Tool-n-Around

  • Storm Trooper
  • ***
  • Posts: 211
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #30 on: March 14, 2023, 01:09:10 am »
PC690?  I did a lot of work with those before switching finally to the 3 phase liquid cooled spindles.  I burned up more than a couple of those PC690s running in aluminum.

Yes indeed. I have a few of 690s too. I also have quite a few of it's big brother. I doubt my 690s will see too difficult of duty with me. My big envelop needs are foam cutting and that's like running unloaded. After being stuck at 1/4", I do look forward to the 1/2" collet and the additonal power cant hurt. I did buy a Super PiD for speed control and CAM interface and my occassional harder materials......but I'm never more than a new router mount casting away from a spindle.

Best,
Kelly
« Last Edit: March 14, 2023, 01:15:07 am by Tool-n-Around »

Offline Bob La Londe

  • CNC Jedi
  • *****
  • Posts: 4358
  • ^ 8.5 pounds on my own hand poured bait.
    • View Profile
    • CNC Molds N Stuff
Re: Contemplating some upgrades to my CNC Router
« Reply #31 on: March 15, 2023, 00:39:25 am »
If I recall it comes with a 1/4 inch and a 1/2 inch collet/nut, but there is (or used to be) a 3/8 available special order.  I might have bought one, but I wouldn't have a clue where it is anymore.  What I did was turn brass bushings on the lathe for other size cutters. As long as the ID was on size, and it was concentric with the OD there was no problem.  A set screw held the cutter in place.  I made them blind hole so cutters couldn't get pushed in under hard cutting forces.  I also made them with a shoulder so the bushing couldn't get pushed into the router collet.  If I run across one I'll snap a picture.  It really opens up the tools that are available for use when you can make your own bushings or tool holders. 
Getting started on CNC?  In or passing through my area?
If I have the time I'll be glad to show you a little in my shop. 

Some Stuff I Make with CamBam
http://www.CNCMOLDS.com

Offline Tool-n-Around

  • Storm Trooper
  • ***
  • Posts: 211
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #32 on: March 15, 2023, 14:43:16 pm »
I bought an ER16 collet adapter and a cheap set of collets. Not ideal as far as runnout I'm sure but I only did so to accomodate the use of drill bits. I suppose in a pinch it could be used for a 3/8" shank cutter. The two main reasons I wanted 1/2" shank is of course sturdier bits but I also have a large collection of contour cutters, such as vee, coves, round overs, etc, which can be useful in my lost foam pattern cutting in fast 2.5D as opposed to having to cut them in a slower 3D routine with smaller ball nose bits.

My Z-axis new hardware is pretty close to done and I'm very happy with it. I'm a bit stuck on the overall project right now trying to decide if just sell my present machine and replace it with my own build. It appears I'll need to replace the controller to accomodate the the new Z ball screw since I dont have access to the OE controller settings. If I could convince myself, that I could get the desired X&Y speed with 5mm ball screws and Nema 23 steppers on X & Y, I'd do so. At that point there's not much left of the old machine and I may as well just replace the beams, build a new controller, sell my old machine and be way better off in both economics and machine quality.

I went back and re-read Dave Benson's earlier posts in this thread regarding the risk of overcapitalizing such a hobby machine, drivers, and steppers. The rub is, if stepper torque really starts to suffer above 300rpm, I'm 2x-3x short of myXY speed goal. However, 5mm BS should require 10x less torque, for a given inertial mass and cutting load. So seems like itshould work with good drivers.

I also noted Dragonfly's earlier post that he has used 5mm lead BSs with Nema 23s without issue albiet somewhat under my speed goals.

I'm just a bit out of my element on the subject machine controller and drive electronics, but am tryng not to let not knowing what I'm doing get in the way of making progress ;)

Best,
Kelly

Offline Bob La Londe

  • CNC Jedi
  • *****
  • Posts: 4358
  • ^ 8.5 pounds on my own hand poured bait.
    • View Profile
    • CNC Molds N Stuff
Re: Contemplating some upgrades to my CNC Router
« Reply #33 on: March 15, 2023, 18:27:30 pm »
I ran straight shank ER collet chucks too.  They are no worse than the shop made bushings for run out if you cut them short enough to be very close to the router collet.  Do not bottom them out (or any tool) into the router or spindle. 
Getting started on CNC?  In or passing through my area?
If I have the time I'll be glad to show you a little in my shop. 

Some Stuff I Make with CamBam
http://www.CNCMOLDS.com

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1745
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #34 on: March 16, 2023, 02:12:53 am »
That’s a nice bit of Z axis bling.

Quote
 I'm having evil thoughts of just selling my machine and upgrading those too.
I'd have a much better more rigid machine and probably money in my pocket from the sale of old machine.

This is good idea, and lets you jump over the rabbit hole rather than end up with a hodge podge
of mechanical\electrical\frame\software assembles that while work, are not really optimised to
work together.

I’m thinking that the Z axis is heavier than the old assembly, and if I recall correctly
your motors were 370-400 oz/in, you can get Nema23 4 amp (560 oz/in) motors that would bolt on
I have one on my lathe X axis and a Nema 24 4 amp on the Z, they are just longer.
If you use the original motors then you may have to back off the GRBL Acceleration values.
You can change these values by using one of the many Gcode senders, if your not familiar with the
process (there are plenty of examples on youtube) then GRBLaser has a very easy to use GUI
to do this as well.

For a little more cash than you can get closed loop steppers with their drivers, or clear path servo’s
but where do you stop, as you would then want to upgrade the controller to (32 bit GRBL) “Fluid CNC”(free) to be compatible with GRBLMachine
or Mach3/4/Linux or the software that comes with one of the many USB/Ethernet boards.

With the extra height and added weight  there will be a greater turning moment on the Y axis assembly (and the beams)
so it wants to be linear rail and serviced (greased) regularly. (this is the most important thing to attend to and get right).
The linear rails will add some stiffness when bolted to the aluminum extrusions, I don’t know if this would be enough though.

Racks have their pros and cons, they are fast though, and if you are happy with the finish off the machine then I’d leave them until I had
measured the Y axis deflection, by running a test program which can be just a series of point to point G0 moves, I did this for my Z axis install
to check the Acceleration values Initially I set the values by running the axis and upping the values until the
axis started to miss steps then backed of the values 20%, I ran a few jobs and everything looked ok however I did a job which consisted of 500
drilled and tapped M3 holes, and after a couple  of hundred holes I noticed that the clearance plane seemed to have closed up some (it was set at 0.5mm)
I then ran a test file with 1000 point to point moves (10 mm) and found that the axis had drifted 0.6 mm so I had to back off the Acceleration some more.
Once you have the acceleration sorted mount a dial indicator not connected to the machine frame (perhaps mounted up off the floor on blocks) with the
indicator point in the middle of the Y axis beam and apply some force and see what the deflection is. ( manufacturers do have this spec) but
it depends on the fixing conditions on each end of the beam.

Your machine manufacturer would have taken this into account (among other things) when deciding to use a rack or a
ballscrew as a rack is much more forgiving and durable when the mounting conditions and\or rigidity of the frame are not optimum.
ballscrews offer no backlash but are speed limited by their slenderness ratio.
For example a Module 1 rack and a 20 tooth pinion will for 1 revolution of the motor move
63.82 mm, a common 5 mm pitch ballscrew would need to rotate 12.76 revolutions.
long unsupported sections of ballscrew tend to whip about at high rpm.
To some degree this can be mitigated by using a rotating ballnut and putting the ballscrew under
tension you need a heavier built frame to do this though.

ps. I just use the windows built in app snipping tool to take shots of parts of the screen.
Dave








Offline Tool-n-Around

  • Storm Trooper
  • ***
  • Posts: 211
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #35 on: March 17, 2023, 01:07:30 am »
…..I’m thinking that the Z axis is heavier than the old assembly, and if I recall correctly

It is heavier. So is the router. I’ll weigh it, but seems to me like the torque required to develop the same axial force on a 5mm ball screw, must be much less than the 38mm lead acme screw that is presently in the machine……..Like in the range of 38/5 = 7.6x less right? So even if the Z/router is twice as heavy, that leaves quite a bit of margin for reduction in stepper torque at higher speed, would it not?

……..your motors were 370-400 oz/in,

They list them as 270oz/in.

you can get Nema23 4 amp (560 oz/in) motors that would bolt on I have one on my lathe X axis and a Nema 24 4 amp on the Z, they are just longer.

I’ve been assessing torque at 15rev/sec, but the only time I’d ever be at that speed is cutting foam and as you have noted, this may be more easily handled with acceleration setting which as you’ll see immediately below, may be target rich territory.

If you use the original motors then you may have to back off the GRBL Acceleration values. You can change these values by using one of the many Gcode senders, if your not familiar with the process (there are plenty of examples on youtube) then GRBLaser has a very easy to use GUI to do this as well.

I have a confession. I’ve been using GRBL Machine for over a year now. All of the dialogue boxes have been expanded and I forgot the settings/parameters were even down there (see atatched). Apparently the settings have been the condition for the last year+. What would be a reasonable acceleration setting and does it control spool up & down? Meaning will it control both the rate it decelerates and the rate it accelerates in a 180 degree return direction?

With the extra height and added weight  there will be a greater turning moment on the Y axis assembly (and the beams)

I would only be cutting foam with large Z extension so not much of a challenge there. I’d raise the bed or install a platform for hard materials so the extension would be minimal. I think it will be quite stiff in the first few inches of travel, at least compared to my present machine. The centerline distance from the router collet to the X-axis bearings is only ½” greater than my present machine. Considering the new router motor is .78” bigger in diameter and the ball screw ~.25” bigger, I thought I was doing pretty well with that.

The linear rails will add some stiffness when bolted to the aluminum extrusions, I don’t know if this would be enough though.

This may become a non-issue if I sell the old machine and replace the beams.

long unsupported sections of ballscrew tend to whip about at high rpm. To some degree this can be mitigated by using a rotating ballnut and putting the ballscrew under tension you need a heavier built frame to do this though.

The unsupported section can never be greater than 30”. I could be wrong but sure seems like a 20mm ball screw should be pretty stout and stable below <1000rpm.?

ps. I just use the windows built in app snipping tool to take shots of parts of the screen.

Thanks Dave. I thought there was way to capture the entire console dialogue. Really appreciate your input.

Best,
Kelly
« Last Edit: March 17, 2023, 01:09:42 am by Tool-n-Around »

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1745
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #36 on: March 17, 2023, 11:05:32 am »
Quote
It is heavier. So is the router. I’ll weigh it, but seems to me like the torque required to develop the same axial force on a 5mm ball screw, must be much less than the 38mm lead acme screw that is presently in the machine……..Like in the range of 38/5 = 7.6x less right? So even if the Z/router is twice as heavy, that leaves quite a bit of margin for reduction in stepper torque at higher speed, would it not?

It’s not a 38mm lead acme screw, it’s T8 2\4\8 mm pitch or a 3\8 12 something like that.
I had a look in the packing list but the pitch is not mentioned. Probably 2 mm.

Quote
I’ve been assessing torque at 15rev/sec, but the only time I’d ever be at that speed is cutting foam and as you have noted, this may be more easily handled with acceleration setting which as you’ll see immediately below, may be target rich territory.

15 revs a second is 900 rpm I doubt that you’ll have much torque at all.

Quote
I have a confession. I’ve been using GRBL Machine for over a year now. All of the dialogue boxes have been expanded and I forgot the settings/parameters were even down there (see atatched). Apparently the settings have been the condition for the last year+. What would be a reasonable acceleration setting and does it control spool up & down? Meaning will it control both the rate it decelerates and the rate it accelerates in a 180 degree return direction?

You have been using the millright settings with is good for the unmodified machine, I did the calcs for X and Y (couldn't find the Z axis pitch) and
at the quoted speeds the motors are running around 270 rpm at 300 rpm is full torque and it drops
fairly fast after that. The Accel and decel are governed by the jerk control setting, this should be ok though.

Quote
I would only be cutting foam with large Z extension so not much of a challenge there. I’d raise the bed or install a platform for hard materials so the extension would be minimal. I think it will be quite stiff in the first few inches of travel, at least compared to my present machine. The centerline distance from the router collet to the X-axis bearings is only ½” greater than my present machine. Considering the new router motor is .78” bigger in diameter and the ball screw ~.25” bigger, I thought I was doing pretty well with that.

This is not where you calculate the forces from, when I get time I’ll draw a diagram.

Quote
The unsupported section can never be greater than 30”. I could be wrong but sure seems like a 20mm ball screw should be pretty stout and stable below <1000rpm.?

20 mm will be ok 16 would not, If I can find them (it’s been a while) I kept the balls out of a 600 mm 16 mm dia ballscrew that I was asked to look at, it had completely seized, the cause of which was that a few of the balls had split completely in half like split peas.

Dave

Offline Tool-n-Around

  • Storm Trooper
  • ***
  • Posts: 211
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #37 on: March 17, 2023, 16:17:30 pm »
…..It’s not a 38mm lead acme screw, it’s T8 2\4\8 mm pitch or a 3\8 12 something like that. I had a look in the packing list but the pitch is not mentioned. Probably 2 mm.

Wow, I’ve been mistaken about this all along and somehow had it incorrectly measured to be much higher lead. 1 revolution indeed produces 8mm of Z motion or lead. This is actually good news because it’s very near the lead of the 5mm or 10mm ball screws I have. The bad news is the new Z axis is quite a bit heavier but its probably no worse off than the X-Axis as far as required stepper torque.

15 revs a second is 900 rpm I doubt that you’ll have much torque at all.

I’ve been surfing around on a few linear motion manufacturer's sites that seem to have decent tech info available online for their products just to get a sense of the art of the possible.

https://www.anaheimautomation.com/marketing/stepper/high-torque.php

The products and info at this site seem to be representative. A plain Jane NEMA 23 is $69 but as you have mentioned, drops from 270oz-in full torque to 25oz-in at 15 revs/sec. In the attachment below I pasted the curve for the best higher torque NEM23 they offer in this regard. It drops to ~80oz-in @ 15revs/sec and costs $179. By comparison, their plain Jane NEMA 34 is still making 380 oz-in @15revs/sec and costs about the same. There is a lot of info left out about the driver and nature of the corresponding drive scheme, but if I’m going to build a new controller, this is why I was thinking I should quit noodling it, just buy 34s, and enjoy very high torque in the more usual speed regimes.

….You have been using the millright settings with is good for the unmodified machine, I did the calcs for X and Y (couldn't find the Z axis pitch) and at the quoted speeds the motors are running around 270 rpm at 300 rpm is full torque and it drops fairly fast after that.
…..The Accel and decel are governed by the jerk control setting, this should be ok though.

OK, jerk control. That’s yet another new one for me. Where is that at? or is it a non-grbl feature?

Just as fyi attached pdf are my present machine settings and what is reported in the console dialogue when I connect to the controller. I don’t understand the settings versus what I have measured for the R & P drive and Z acme screw.

Somewhere I got this information:

MegaV Stepper Motor Specs:
NEMA 23 frame stepper motor
Total Motor Torque: 270 ounce/inch
Digital Stepper Driver Max Amperage: 4 Amps/phase @ 48Vdc
Default settings for Mega V:
$100=57.288 X steps/mm   3200 steps/rev
$101=57.288 Y steps/mm   3200 steps/rev
$102=200.000 Z steps/mm  1600 steps/rev

Comparing what I have measured for XY&Z displacement per revolution, and looking at the GRBL GUI, I confirmed the stepper drivers must indeed be set at 16 microsteps on X&Y, and the Z driver at 8.

It also means the X&Y Steppers are running at approximately 90rpm for 5000mm.min target speed, and the Z would need to turn 625 rpm for same speed so required Z speed is evidently much less.

Where is the stepper driver microstepping setting and how is it changed? Is it software or DIP switch settings on the driver hardware?
So I connected to my controller just to see if I could alter the Z axis resolution. I typed $102=100 and tried to jog it .5”, expecting it to travel .25” and it did. So if I make the appropriate adjustment to $132 Zmax travel, it appears I should be able to accommodate my new Z axis with the existing machine. Same goes for installing ball screws on X&Y if the steppers can handle it. Light bulb is may be turning on.

Best,
Kelly
« Last Edit: March 17, 2023, 22:19:58 pm by Tool-n-Around »

Offline lloydsp

  • CNC Jedi
  • *****
  • Posts: 8820
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #38 on: March 17, 2023, 18:12:53 pm »
"The bad news is the new Z axis is quite a bit heavier but its probably no worse off than the X-Axis as far as required stepper torque."
-------
So, if it's too heavy, counterbalance it with a small air cylinder + regulator that has a stroke just longer than the Z travel.  Our older (of 2) ShopSabre routers has a heavy spindle, and uses a 1" cylinder to balance it, so the stepper is only working against inertia and cutting forces, not weight.

I prefer to leave just a little weight in the downward position to help with cutting forces, also.

Lloyd
« Last Edit: March 17, 2023, 18:15:13 pm by lloydsp »
"Pyro for Fun and Profit for More Than Fifty Years"

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1745
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #39 on: March 18, 2023, 04:08:59 am »
Kelly

Before you pull the trigger, there are some things to know about the motion control options
open to you.

For standard hobby machines that most home enthusiasts either buy or put  together as a project
A standard bipolar stepper motor and stepper driver plus power supply sourced from ebay or aliexpress seems a pretty good deal, but
this can often cost more and offer less
performance. I’ve snapped some pics off the net, showing some examples of other offerings.

There are  bipolar two phase stepper motor drivers, like an MAH860 that requires a separate power supply.

There are  bipolar two phase stepper motors, similar to  MAH860 that do not require a separate power supply.
They plug into single phase so no extra power supply taking up space in the control cabinet and the associated extra wiring, this can
if not eliminate reduce potential  noise issues.

There are 3PH stepper motor drivers and motors The three phase refers to the motor coil construction not the incoming power required,
These run off single phase too, and are industrial rated.

There are AC and DC servo motors an example of a DC servo would be Clearpath.
There are literally hundreds of AC servo manufacturers.

I’ve snapped three random examples, The stepper motor drivers will give between 4 and 5 NM
of torque at 1000 rpm, The AC servo is way to big it’s just example.

You have to shop around to find a good price, I’ve seen a couple of hundred dollars difference in
the same units for a three pack.

Often you can get NOS units from industrial wholesalers at very good prices once you are aware
that these units exist.

Will an MAH860 + motor + power supply be ok for what you are doing yes, I have 9 in service
and two as a backup (never needed them though) I also have the 240 volt units and a gecko and Leadshine servos and some
SEW Eurodrive and chinese inverters all have been good.

Dave

Offline Tool-n-Around

  • Storm Trooper
  • ***
  • Posts: 211
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #40 on: March 18, 2023, 12:56:49 pm »
…….So, if it's too heavy, counterbalance it with a small air cylinder + regulator that has a stroke just longer than the Z travel.  Our older (of 2) ShopSabre routers has a heavy spindle, and uses a 1" cylinder to balance it, so the stepper is only working against inertia and cutting forces, not weight.

Just thinking about this Lloyd, how are you maintaining the cylinder pressure/counterbalance force throughout the stroke? -In my case 12” of stroke. I can certainly see how the regulator does so initially and as the cylinder volume expands, but what eliminates the charge from being compressed and pressure/force increasing on the return stroke?

.......Kelly, Before you pull the trigger, there are some things to know about the motion control options open to you.............

Not going to buy anything just yet. There are a mind-numbing number of options as far as motors and drives and I've been reading up and learning (a lot). The R&P drive does keep economical steppers in the sweet spot of their performance, so I understand why the manufacturer took the approach they did for the niche their product offering serves.

I have an extra stepper from the 4th axis. It’s the same as the other three. I’m going to finish up the Z with that stepepr and experiment with speed and acceleration settings to see how fast I can actuate it without lost steps, maybe in both horizontal and vertical positions so I can see if my inertial calcs are correct. From my reading, manufacturers recommend keeping the inertial load <10x the rotor inertia, and <4x for high speed/performance systems.

I have a suspicion that the X-Axis could become the limiting factor on speed. So, to test that premise, I think I’ll just replace the router in my machine as it presently sits with a mass that brings it to approximately the same entire mass as my new Z assembly and do the same test with speed and accelerations settings and see what performance can be achieved with the existing hardware.
Might be good enough. We shall see.

I still need to raise the gantry and my may go ahead and take your advice on measuring beam deflection, but I can tell you with relative certainty, the two biggest culprits in the (poor) rigidity and deflection of my present machine are the four corner posts (which are terrible), and the track wheels. I can literally feel the posts deflect (a lot) in the Y direction under inertially induced loading as the machine runs so I’m sure they also deflect under load and were major contributors to the waviness you saw when I machined the decorative ribs in polycarbonate air cleaner lid. Those flimsy strap type posts are also mounted to a 16 gauge steel skirt on the plasma table which also deflects.

I’ll make a tubular steel subframe weldment and stout corner posts to remedy this and am already committed to replacing the wheels with linear rail.

Best,
Kelly
« Last Edit: March 18, 2023, 12:58:54 pm by Tool-n-Around »

Offline lloydsp

  • CNC Jedi
  • *****
  • Posts: 8820
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #41 on: March 18, 2023, 13:33:46 pm »
"I can certainly see how the regulator does so initially and as the cylinder volume expands, but what eliminates the charge from being compressed and pressure/force increasing on the return stroke?"
-------------------------
Some (many) regulators 'relieve' excess pressure.  If it's a good one, it'll keep the pressure constant within a couple of PSI regardless of whether it's being extended or compressed.  In this particular case, a very small regulator seems to be more 'sensitive' than a big one would be.

Another solution to help mitigate that issue would be to use a cylinder much longer than the required stroke.  I didn't find that to be necessary.

Lloyd
"Pyro for Fun and Profit for More Than Fifty Years"

Offline Tool-n-Around

  • Storm Trooper
  • ***
  • Posts: 211
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #42 on: March 18, 2023, 17:11:47 pm »
........Some (many) regulators 'relieve' excess pressure.  If it's a good one, it'll keep the pressure constant within a couple of PSI regardless of whether it's being extended or compressed.  In this particular case, a very small regulator seems to be more 'sensitive' than a big one would be.

Another solution to help mitigate that issue would be to use a cylinder much longer than the required stroke.  I didn't find that to be necessary. Lloyd

Since I'm already at 12" a much longer cylinder wouldn't be very practical, but simply running a small line to a remote rervoir that was large compared to the swept volume of the cylinder would make any pressure change quite small and make it zero air consumption. Since I only need to counterbalance maybe 20lb, that could easily be done with a 3/4" bore x 12" stroke cylinder. (double acting would make it package better). That would be inexpensive to the tune of like $30-$40. I'll keep that one in my back pocket if needed. Thanks for the suggestion.

Best,
Kelly

Offline lloydsp

  • CNC Jedi
  • *****
  • Posts: 8820
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #43 on: March 18, 2023, 17:21:30 pm »
"...but simply running a small line to a remote rervoir that was large compared to the swept volume of the cylinder would make any pressure change quite small and make it zero air consumption."
---------
Yep, that's an excellent idea!  Good thought.

Lloyd
"Pyro for Fun and Profit for More Than Fifty Years"

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1745
    • View Profile
Re: Contemplating some upgrades to my CNC Router
« Reply #44 on: March 20, 2023, 00:14:00 am »
HI Kelly

Quote
I have a suspicion that the X-Axis could become the limiting factor on speed. So, to test that premise, I think I’ll just replace the router in my machine as it presently sits with a mass that brings it to approximately the same entire mass as my new Z assembly and do the same test with speed and accelerations settings and see what performance can be achieved with the existing hardware.
Might be good enough. We shall see.

This is a good idea.

I’ve snapped some accelerator apps, with these you can take a baseline measurement and as you
make adjustments see what effect they have.
The idea here is to load the app and fix the phone to safe place on the Y axis post and measure the
ringing, before and after you have modified the frame for example.

Quote
I’ll make a tubular steel subframe weldment and stout corner posts to remedy this and am already committed to replacing the wheels with linear rail.

This is the rabbit hole.
Linear rails require a pretty good surface, for example they require very little surface deviation
in flatness and rotation (twist) not to bind, and a standard extrusion has some of each.

When you have some spare time pop over to mycnuk.com and have a look at the built logs
It’s very interesting to see what solutions people have come up, one guy even sent his newly cut work piece
away to be measured with a CMM, He was a little underwhelmed with the result
although he had taken a really good methodical build. I have to say that the results were pretty
spectacular for an aluminum framed router I guess it just depends on your expectations.

I followed (at that site) a fellow who wrote some code and used a webcam with the ir filter off in a machined al case and a builders
 laser level to to achieve a few microns of accuracy to level his rails.
See pic.

You have a good attitude, and I’ looking forward to you progressing to a satisfactory result.

Dave