How To Select Motor Sizes

[EverydayDiesel]

Hello,

I need help selecting the right size motor for my 4x8 table I am going to build.

I will be using a rack and pinion type drive with a 3:1 reduction. The gantry will be all cold rolled steel (maybe 3x3 1/4 wall?) so it will be heavy. I would like to keep the speed up to ~1200 IPM for cutting thin metals for a target

2 Nema 34's for the Y axis
1 Nema 34 for the X axis
1 Nema 23 (or 34) for the Z axis

motor selection.png

How do I go about selecting the appropriate size motor for this setup?

Thanks in advance for any help you can give.

[robertspark]

what rack and pinion size?

why 3x3 (I am curious why you settled on that profile)?
With a 1/4" wall thickness? (again, just curious why)?

presuming 1' of carriage, your gantry will be 5' wide to accommodate a 4' sheet

It will weigh 8.81 lb/ft so 44 lb before you add any end plates to it, z axis, stepper motors (nema 34's are about another 8 lb each maybe a little more) so you are at 68 lb before you've added the end plates and any linear guides (if you are using them). say 100 lb as a reasonable estimate.

You mention 1200 ipm to cut thin metals (I've not seen a cut chart at 1200ipm, so what are you intending to cut?)

the table is 4x8, so 1200ipm will traverse the table in 41 seconds not allowing for acceleration and deceleration.

The biggest machine that I know of in the air plasma range is the hypertherm 105 (that publish detailed cut charts [keen to know of another as I know thermal dynamics do a 120A cutter, but the cut speeds look lower]) and the highest cut speed is 500ipm for production settings with 45A consumables on 26 guage mild steel

https://www.hypertherm.com/Download?fil ... &zip=False

at 1200ipm..... what is the smallest object you want to cut, what sort of accuracy are you looking for .... would you be happy with 1/4" rounded inside corners (1/8"; 1/16", 1/32"..... what were your thoughts?)

If maybe you come down on your cutting speed..... to say 500ipm..... again what rounded corners would you be happy with on an inside cut of a square say?

(remember you dont have to cut 26GA at 500 ipm..... you can use fine cut consumables and do it at say 325ipm ......... you could use fine cut consumables up to say 14GA.... that way you now flip to a 45A set of consumables and your cut speed is 320 ipm (on mild steel) [stainless is a little higher]..... so now you only need to design your cutting speed for accuracy at say 400ipm...... (your rapids can be higher than this as the accuracy is not an issue..... given you still consider what you will accept for an inside corner as being an acceptable rounding......

This rounding will tell you what acceleration you need, which in turn tells you what that gantry is REALLY going to weigh against the stepper motors..... being flung in the opposite direction....

simple analogy.... say you want 1G of acceleration..... that gantry at 100 lbs will have an effect of weighing 100 lbs on those stepper motors when then change direction in lateral force.... 1G of acceleration at 1200ipm will give you a minimum internal corner radius of 1 inch...... is that acceptable??

If not..... you need more acceleration..... twice the acceleration..... half the corner radius....

lets go the other way.... say you decide 400ipm is your design for accuracy.... and you still want 1G of acceleration....
your internal corner radius will be a minimum of 0.11" (just less than 1/8".....radius)..... 1G is quite high... not impossible but with say a 100lb gantry.... that could be a problem and you now need some really big motors.... which add more weight than you thought you needed and so it goes on....

Also remember that your profile is 3"x3".... this presumes that the force of both the weight of the Z axis carriage will be both the same when it is stationary and at 1G of acceleration (given that force would now be horizontal).

Maybe a better target would be 0.3G....... at 400ipm
Running the math.... the gantry would only need to be 1/3rd of the width as its height for the same deflection.... given a little less than 1/3rd would be transmitted horizontally under acceleration and deceleration ..... so your gantry is 3"x1" wide section and it now weighs 5.4 lb/foot... again you need 5 so 27 lb.... you just lost 17 lb (17% of the weight all else being the same) if you change this to aluminium.... the weight would reduce to 1/3rds (roughly) (0.1 lb/cu in for aluminium and 0.28 lb/cu in for steel density)

its all about understanding what is acceptable in terms of finish (would a 1/4"; 1/8"; 1/16" or 1/32" be acceptable)..... there is a bit of error here that someone may pick up on..... these acceleration calculations presume a centre line cut and do not offset the kerf to one side of the cut so in practice the finish would actually be higher.... as the kerf would cut into that 1/4"; 1/8"; 1/16" or 1/32" by 0.035 to 0.024 of an inch roughly (1/32").

I would also look at the 3" section and consider what the Z axis carriage is likely to weigh..... and consider what the deflection is likely to be does it need to be 3" for a reason?

[EverydayDiesel]

Wow thank you for the response. You (and others on here) have probably forgot more about this stuff then I will ever know.

Here is the notes that I have updated

1. The table will be a 4x8
2. The gantry 'cross member' will be 3x1x.25 for rigidity (or should I go with something different?)
3. Target cut size is 20ga all the way to 1/2" but 95% of the time it will be 1/4" steel.
4. I have a hypertherm 45xp with cpc which I might upgrade at a later date.
5. I will be doing a similar build to precisions gantry. I really like the triangular sides as this design seems really stable to me because you can have two 'feet on the rails.

6. I got the 1200ipm from jd2 table that I really like. I see now that I dont need anywhere near that speed. https://www.jd2.com/mad-multiplatform-cnc-table
7. This is the rack and pinion I am looking athttps://www.finelineautomation.com/products/rac ... axon_id=35
with these racks https://www.finelineautomation.com/prod ... axon_id=35
8. These are the power supplies I plan on using (made in America) https://www.teknic.com/products/servo-m ... er-supply/


So if i move to a 3x1x.25 (or 1/8?) with one of the motors below (2 for Y axis, 1 for X axis) how would I achieeve say a 0.11 min corner radius?


Also when people talk about a 600 oz stepper they are referring to the peak torque right? So those motors are 1k+ oz?

Is it better to have higher RPM and more torque or less torque and higher RPM. After reading what you typed, I would assume more torque and less RPM for directional change.

[robertspark]

Wow thank you for the response. You (and others on here) have probably forgot more about this stuff then I will ever know.

No, I don't know that much, I have the limited knowledge of only my table and the other info I pick up along the way (job, life etc). I always start off why do you need that or why are you settling that specific parameter. If there is a reason for it that is fine. The point being if the reason cannot be justified then it may be changed to another specific justifiable parameter.

Point 1.... Plasma is all about acceleration
So if it can be made lighter then it will accelerate faster using less torque which means smaller motors possibly and less weight.

Point 2..... Jolt or jerk (in/s/s/s) is always the killer of acceleration
With stepper motors unfortunately they step and are not smooth accelerating devices so working out that first step and how much force you need to smack into getting the thing to move that is where your problem lies (in my opinion). Luckily with trapezoidal acceleration that makes life a little easier to work out as the time parameter is specific.

I'll look at the other stuff later, I'm sure someone else will chime in too.

Don't put too much faith in the advice you get.... It is free afterall.... Just consider it and move on as you have then justified your parameters in your own mind as you are the one who will be left with the boat anchor or pig later that does not do what you want because of someone else's advice or paid a premium for that is never or seldom going to be used.

It seems like your list of parameters are well thought out which should make the math on selection easier. Should being the operative word! You can and will always add margin onto it but again justify the margin because if it doesn't work.... You are the one left with the poor performance setup....or over engineered solution costing you a premium in money, weight or unutilised resource... Whatever.

[EverydayDiesel]

I was looking at the JD2 table and they use the same tecnik clear line servos (but i dont know which rpm they choose) I do know that they used a 10:1 reducer which I guess is for higher resolution. Maybe I should choose the 2370rpm servo and use the same reducer.

I am having a difficult time finding the formula that says X rpm == x inches.

I really dont know but I know there is alot smarter people on this forum then myself. I am just here doing research and trying to find out what works (and what doesnt)

[robertspark]

Servos... Different animal, constant torque device
https://www.motioncontroltips.com/stepp ... ervomotor/

You need to know what pinion you are going to use with the rack

[robertspark]

Servo I would select on the rated rpm as being equal to or a little higher than your intended design velocity.

say you want 1200 ipm as your rapid speed, and your servo motor is rated at say 4000rpm, given its a constant torque device what you do (in my opinion) is design for your servo motor to operate at 4000 rpm at 1200 ipm given you are then using all of the machines torque and you are leveraging that torque through gearing

a 20 pitch rack is normally coupled to a 20tooth pinion so it will travel approximatly 3.142 (pi) inches in 1 revolution

so 1200 ipm / pi (3.142) gives you approx 382 rpm.

you would like your machine to run at the end of its torque rating (before it drops off) so the pinion is travelling 382 rpm and the motor is traveling at 4000rpm (in this example)..... 4000 / 382 = 10.47:1

so a pulley ratio of 10:1 will provide your optimum 1200ipm with a 20 pitch rack and 20 tooth pinion.

10:1 means that torque will be increased by 10 times from the motor

so a servo motor that puts out its rated torque (which is constant from start to its rated rpm) of say 50 oz in (round number) will become 500 oz in at the pinion.... the pinion to rack ratio is 1:1 (20 tooth / 20 pitch) so 500 oz/in (of TORQUE) to the rack

torque does not mean a lot as 500 oz/in applied at say 2 foot (24") is only a force of 500 / 24 = 20 oz of force applied at the end of that 2 ft wrench.

or if you have only a 1/2" long wrench.... 500oz/in now becomes 1000 ounces (62.5 lb) of force you have to apply.

going back to our rack and pinion 20Tooth pinion.... 20 pitch rack..... means 1 revolution will be 3.142" (pi) of travel..... therefore the (effective mating) circumference of the pinion will be 3.142 inches.... circumference of a circle is pi * D.... hence circumference / pi = diameter (pi/pi = 1inch)..... the force is applied from the centre of the pinion (i.e. the radius of the pinion) so 500oz in / 0.5 inch means it will provide a force for movement of 1000 oz (62.5 lb) from one motor..... if you have 2 on both sizes of the gantry..... the total force you can move is twice this

In this example I've used the
https://www.teknic.com/model-info/CPM-M ... tage=75VDC
or
https://www.teknic.com/model-info/CPM-S ... tage=75VDC
which produce 49 oz/in of torque all the way to 4000 rpm.

now the 62.5 lb of force from 1 motor needs to relate to something..... if the z axis weighs say 62.5lb and you accelerate it at 1 G then the force to move it will be 62.5lb..... EXCLUDING FRICTION of both the slides and the drive train (drag will be negligible as it's only moving at 1200ipm (1.1 mile per hour)

friction is a tricky one... as you can guess it or measure it..... an even if you measure it ..... are you sure you are accurate about it..... this is where the fudge factors come in to compensate for guess work....

There is a phenomenon known as slip-stick.... basically when something is stationary and just as it begins to move it can be stuck to the surface because all the lubricants used on the mating surfaces are stagnant..... (I do not profess to be an expert at ANY of this mind you so please anyone feel free to correct me as I'm always willing to learn and correct what I've picked up along the way)
https://www.machinedesign.com/mechanica ... phenomenon.

This is a very common problem with lathes and milling machines.... because the mating surfaces are large between the carriage and the bed and they are lubricated with a specific oil...... sometimes (ask me how I know) .... people clean up their lathes and milling machines and polish all the mating surfaces using a very fine rubbing compound to painstakingly hone the mating surfaces..... wash and clean everything grease it all and put it back together (without the leadscrews yet) and give the carriages a slide only to realise that "great" general purpose, bearing, high load whatever grease sticks like "£$%% to a blanket..... and until you overcome that initial jerk it then slides fine.....

The way to correct for this is to use a mill / lathe bed lubricant which is formulated to avoid / minimise slip-stick .... clean off all the stuff you added to your finely honed surfaces add the correct stuff and it slides with a finger push.... back and forth as smooth as silk.

It looks like you intend to use v-groove bearings such as RM2-2RS or something similar on the Y axis carriages.... these will be low friction in my opinion lower than using linear slides (but I've not used / held a genuine quality THK linear slide so I'm not so sure).

On the X axis.... they also look very similar... again (in my opinion) a good low friction solution.

I am REALLY guessing here... but I'd say at +20% to the weight to account for the friction..... so your motors can now only drive a 50 lb X axis carriage at 1G given 20% is the allowance for friction.

If you get the math wrong.... well the worst that it can do is the machine won't perform as you thought it would.... your accelerartion will be lower than you thought it would be.....

I know I've not answered your EXACT problem (choose some motors) but really you need to work out weights and exact parameters in a clear sheet of paper and then work back from those.

Start with the Z axis and the weight it will carriage, the acceleration and feedrate you want it to run at.... choose its motor..... you now can work out the weight of the Z axis and in turn the X axis carriage and then the gantry.....

The table is really to support the weight of you material + the carriage and designed as such...... note that moving at 1G( for simplicity) you will be flinging that carriage around and its lateral force will be as much as it weighs....

[robertspark]

its also probably worth mentioning that the earlier example ended up with a 50lb x-axis carriage and a 100 lb y axis gantry being able to accelerate at 1G.... if say you get the math wrong and your x-axis weight 100lb... well your acceleration will be 1/2 what you wanted or thought it would be with that example.

I've seen 3:1 gear drive reduction mentioned as "optimum" for steppers (not servos!) ..... I've not looked at this too much but it works as a 3:1 torque multiplier so I'm sure it will improve the low end torque to get things moving.... but steppers run out of torque as they are constant current devices

Power = 2 x Pi X RPM x Torque / 60 ...... a stepper being a constant current device and although the power increases (up to the point where the voltage available runs out) really all they are doing is trying to add more power (voltage) to maintain the current through the windings to compensate for the back EMF.

A servo is more superior that a stepper (whether open or closed loop (hybrid) stepper) as they run out of puff (torque) trying to keep up with the losses and maintain current flow. (imo)..... if you can afford them that is.... (clearpath are nice servos, just dont use them for sensorless homing with a cnc machine as I don't think the cnc machine will get the feedback to say that the servo has actually reached its home position (they use a stall method to determine when its homed (the TMC2209 tiny stepper drives for 3d printers also provide sensorless homing but for stepper drives).


Again I am absolutely no expert and am always willing to hear / read and learn others viewpoints and references (if we never take them on board we will never learn anything and progress). I don't work in the cnc industry and have zero affiliation to any manufacturer for any cnc product. It is always good to hear other manufacturers viewpoints like Tom Caudle and Jim Colt who come on here from time to time and learn their viewpoints and perspectives.

[EverydayDiesel]

I was quite suprized that the example used a 50oz (continously) motor. Especially when i was looking at 150+ continous (1000+ max)

The 3:1 gear reduction rack and pinion assembly is the only one I have found like it. Do you know why it would not be recommended.



Double bearing drive spindle. Features a 20 tooth pinion gear and 64 tooth timing pulley (5mm pitch). Comes pre-installed on the plate with a precision ground stainless steel shaft.

There has to be an optimal rpm range for these tables and i think that would really clear things up.

Ultimately I am looking for (mimicking in the high quality tables)
1. Target rpm range (using 20 tooth)
2. Target torque range.

I realize it's not that simple but there has to be a general range when selecting motors.



EDIT: i found these...... https://www.rovercnc.com/collections/ra ... 0275210309

[robertspark]

I was quite suprized that the example used a 50oz (continously) motor. Especially when i was looking at 150+ continous (1000+ max)

The 3:1 gear reduction rack and pinion assembly is the only one I have found like it. Do you know why it would not be recommended.

Double bearing drive spindle. Features a 20 tooth pinion gear and 64 tooth timing pulley (5mm pitch). Comes pre-installed on the plate with a precision ground stainless steel shaft.

There has to be an optimal rpm range for these tables and i think that would really clear things up.

Ultimately I am looking for (mimicking in the high quality tables)
1. Target rpm range (using 20 tooth)
2. Target torque range.

I realize it's not that simple but there has to be a general range when selecting motors.

It is not 3:1 is not recommended.... 3:1 will work fine with the right motor, but 3:1 says that the torque is multiplied by 3 times and the RPM of the motor is 3x that of the pinion.... so it's probably not operating in the optimum range for the servo motor and you are leaving money on the table by buying something that you don't use the full range of it.

I don't know what those gantries weigh, I am sure if you are designing yours you can work out what all of the axis and carriages will weigh without the motors using the components you've selected which will give you a number you are happy with.

if you want to rework the example at 3:1 .... the table moves 1200ipm as a rapid say, 20tooth pinion 20 pitch, so it moves 3.142" per revolution of the pinion..... 1200/ Pi = 381.9 rpm of the pinion, with a 3:1 reduction (from the motor), this would then be 381.9 x 3.... so at 1200ipm the servo motor will be doing 1,146 rpm approximately.

This NEMA 34 clearpath servo is rated at 1410 rpm (max) but the RPM drops off after about 1200 rpm and it costs $299. It has a rated torque of 389 oz in, so at 3:1 the torque at the pinion will be 1,167 oz/in at the pinion and 1,167 oz
2334 oz (149 lb) of force on the rack (the driving force) so it will move a carriage at 1G weighing about 125 lb up to 1200 ipm (and a bit more), and a gantry weighing 250 lbs at 1G.

https://www.teknic.com/model-info/CPM-M ... tage=75VDC

The thing is.... it weighs 4.4lb and requires 247Watts (and up to 352W peak) at 75V but its cost comparable with a lighter and smaller Nema23 ..... so it will require a larger power supply and use more power when its working than a lighter machine ..... costing you more than it would need to if it were to be run as a production machine, given it will be burning through more electric than you need to, increasing the costs of producing the parts or lowering your profit margin if you have to bid against other machines that may be better designed.

But hey it will probably move any weight of gantry, X-axis carriage or anything you want to put on the Z axis you can imagine.

As a design engineer.... its over designed! .... but hey it'll work with ANYTHING you can throw at it including if you wanted to put a driving seat on the gantry.....

You shouldn't design on peak torque (in my opinion) but this will help (increase) your acceleration and overcome slap-stick and high jerk/jolt at 1G acceleration. Also some servos have S-curve acceleration built in and because of the following error allowance they are able to smooth high jolt/jerk and lower the acceleration to compensate for these instantaneous acceleration moments.

[EverydayDiesel]

I cant thank you enough for your replies. You really understand this stuff while I am over here struggling. I really appreciate you helping me

I like the drivers seat on the gantry idea

This NEMA 34 clearpath servo is rated at 1410 rpm (max) but the RPM drops off after about 1200 rpm and it costs $299. It has a rated torque of 389 oz in, so at 3:1 the torque at the pinion will be 1,167 oz/in at the pinion and 1,167 oz
2334 oz (149 lb) of force on the rack (the driving force) so it will move a carriage at 1G weighing about 125 lb up to 1200 ipm (and a bit more), and a gantry weighing 250 lbs at 1G.

https://www.teknic.com/model-info/CPM-M ... tage=75VDC

Am I reading it wrong? The page says 289oz for that motor



In this video at 0:42 seconds
you can see that the precision plasma cnc uses a 3x3 AND a 1x3 to span across the gantry. I find it funny that none of the pictures on their website show this.
The precision plasma cnc weighs about 150lbs according to them... "The weight of a 6’ gantry with motors and torch can be as high as 150 lbs. "
http://www.chacich.com/INFO/hdg/prints/ ... enough.pdf
When I contacted PP they said the aluminum and the steel weighs the same due to the thicker aluminum required to maintain the same rigidity.

I think its safe to say that my gantry will be around 150lbs as well, realistically these things add up and you know how it goes....starts out being 80...then 100... then 120lbs etc


So with your help I have determined that I am looking for these specific things
1. Gantry weight 150lbs
2. 600ipm target cut speed (max)
3. internal corner radius will be a minimum of 0.11"
4. 3:1 gear reduction
5. 20 tooth rack/pinion

You suggested this motor which is great but.....
https://www.teknic.com/model-info/CPM-M ... tage=75VDC

What does a higher rpm do to this equation?
say with this motor https://www.teknic.com/model-info/CPM-M ... tage=75VDC


https://www.teknic.com/products/clearpa ... 8]=200|500

[Rodw]

I'd challenge them on their decision to use steel over aluminium. Our deflection analysis did not support that it was required for rigidity. You should go close to being able to halve that gantry weight. I think mine weighed in at about 35 kg (77 lb). 2.5: x 2.5" with NEMA 34 motors.

I have not looked at the Clearpaths but we found that a 30mm diameter pinion (which your 20 tooth will be close to) and a 5:1 pinion hits the sweetspot across a range of stepper motors. I'd also look at a few more servos before settling on Clearpath. I would have expected a greater range of usable torque at higher RPM from a servo. I have heard some people mention some servos are steppers in costumes...

[EverydayDiesel]

What gantry do you have? Diy? Do you have any pictures?

Do you have another brand suggestion of servos I should be looking at?


I also found this http://www.cncrouterparts.com/pro-rack- ... p-226.html which claims to have a 3:2:1 reduction which i dont understand what it means. What is the difference between 3:1 and 3:2:1

[Rodw]

Its 3.2 is to 1.

If you can build accurately, there is no need for a spring loaded pinion drive. I asked the question here right at the beginning of my build
https://forum.linuxcnc.org/30-cnc-machi ... ve?start=0
Oops, My X axis has a 3:1 reduction becasue thats what I had! It works Ok.

I did do a build thread here on this forum
https://plasmaspider.com/viewtopic.php?f=103&t=22826
There is a lot more detail on my Linuxcnc build thread (55 pages) so the 6 pages here is more of a summary.

My gantry is 40mm x 80mm x 3mm hollow rectangular section Aluminium with 2 x HGR 15 linear rails and rack and pinion. From memory the gantry assembly weighed 20 kg (44lb) and the gantry ends with motors mounted were 7kg (15 lb) each, so say 35kg (75 lb) all up. I just cold not see the added weight and expense of profiles was worth it to take the easy way. They certainly don't improve on the stiffness of the gantry.

This stepper powered machine runs 21 m/min (825 in/min) rapids and accelerates to 0 to 10,240 mm/min (400 in/min) in 115 ms.

I've not really looked at servos but If I did use them, I'd probably use the cheap Chinese 80ST 750 W servos (about USD $250 each) and you can see they scorch the Clearpaths but are probably more difficult to tune from what I've heard. You can run these with a Mesa 7i77 or their 7i95 ethernet version

Screenshot 2020-06-07 at 5.50.06 PM - Edited.png

AND the Plasmac THC is not too bad after I improved the probing speed from this video.

[Rodw]

I forgot to mention, I'm only using a DC 48 volt 10 amp power supply and am working on an upgrade to more commercial motors so time will tell.

[robertspark]

I also found this http://www.cncrouterparts.com/pro-rack- ... p-226.html which claims to have a 3:2:1 reduction which i dont understand what it means. What is the difference between 3:1 and 3:2:1

It is 3.2 : 1 ratio

Not 3:2:1

Aluminium is the way to go but 150 lb for a gantry with 4' of travel is really poor design unless you are milling the steel out with it and need the mass.

You can use thinner wall steel too... 1/4" is unnecessary in my opinion. It's all about the moments of inertia (bigger profile in the direction of the forces you anticipate as the centre of the profile is called the neutral axis and does not deflect (on an equal profile (getting complicated now)) .... For a downwards force the top of the profile is placed in compression and the bottom in tension.

At 1g design acceleration this effect would be the same on the sides of a profile ...... So using an I beam for example would not work very well as it would suffer from greater deflection under side to side acceleration and deceleration but is good for buildings and bridges....

Hence at 1g design acceleration a square profile would be best (in theory).... At 0.3g design acceleration a rectangular profile of 1/3rd width to height would be do the same defection in both the vertical and horizontal axis. At a 2g design acceleration it would be a 2:1 profile with the width being wider than the height for the same deflection.

I say in theory..... A square or rectangular profile.... Now you have to consider what the effect of deflection is.... Vertical deflection is it an issue?? If you have torch height control and you don't use linear guides .... Vertical deflection can be compensated for and provided your gantry is no too high the effect on the cut bevel should be minimal. Defection on steel and aluminium profiles over 5ft span should be very very small too given the use of standard chs and rhs profiles in this application even with a heavy z axis carriage.

I use a 2"X2" with a 1/8" wall section with one of these kits. At the time it was the most cost effective solution as I did not have access to a mill, router or CNC plasma cutter at that time (I've since found a low cost laser cutting service that I have used when I want something more precise than plasma) and have a mill and lathe too

https://i.pinimg.com/originals/07/a7/dc ... 02e145.jpg

https://i.pinimg.com/236x/64/0d/d8/640d ... -table.jpg

Quite a few manufacturers have used this design in the past and some still do such as one of the sponsors on here
https://xtremeplasma.ecwid.com/4x4-CNC- ... p153634730
Plus plasmacam (do/did)
And I think precision plasma started out selling kits like these a few years ago


It's light and meets my needs and can be extended and also designed to be taken apart quickly as I use it in a 5mx7m (16x23ft) garage that is shared with my other hobbies

It's all a steel frame, I have considered changing the gantry to aluminium to improve its performance but it's quick enough as it is for everything I throw at it hobby wise 2mm to 6mm (16ga - 1/4") steel, not tried aluminium yet (only been about 5 years of use ). The z axis has always needed improvement as the kit came with a piece of threaded rod and a long studding nut which had a lot of play and was replaced with a delrin nut and tr4-10 I think it was trapezoidal leadscrew (4mm pitch) but that was only ever a temporary solution as the delrin does not like high z axis acceleration being a polymer.... I have toyed with various improved z-axis designs over the 5 years I think I've had it now including using open builds linear rail / open rail, belt drives, and 10mm pitch leadscrews.... And I'm now playing with a ball screw and very compact and lightweight all in one design idea which incorporates home, floating head and breakaway microswitches and laser pointer and will see how that pans out as I have an air scribe too to mount on the shared carriage ..... I have also used scanything in the past to trace parts with it so the new breakaway idea will allow for changing heads to allow for a pen marker, and camera tracer too with led ring light.

I do not use the table for production it is just a hobby (at present .... While the day job still pays me more than I could ever make steel cutting as a one man band ... Who knows with covid19 and the UK ... 3million + job losses anticipated (10-15% of the workforce))

My stepper motors are well oversized at 4.2A nema 23's but at the time I built it I did not know what I now know... (story of life) .. And there is no point in replacing them due to sunk cost effect. If I did need to change my hobby to a living then the only changes I would make is swapping out the belt drive for rack and pinion and making it a longer table for full sheets. And I would improve the downdraft performance of the table and metal particulate capture. The bearings, 2x2 and stepper motors are all cheap and quick to replace and I have Soares available on the shelf

Rod however uses his to earn his bread making 4x4 accessories I believe it was and has put (100's? of) tonnes of steel through his and has a water table too.

[Rodw]

Rod however uses his to earn his bread making 4x4 accessories I believe it was and has put (100's? of) tonnes of steel through his and has a water table too.

Actually Rod built his table as a hobby really but he does cut a few parts on it that he used to get laser cut and is getting more adventurous with it. I cut down $200 of steel the other day which are now at powdercoating which will become $2000 worth of parts. I've also got another part to go to powdercoating this week where I made a part to use up some faulty laser cut tube pieces that were cut wrong so they cost me nothing.. But really I try to avoid making things myself as I have a business to run!

The original table design was to allow either a water pan or downdraft unit to drop in. When I started it, I was working where a plasma table could have been handy so I moved away from a ghetto build to a full on industrial quality build which by and large I think I achieved. I ended up getting a sheet metal business to make the water pan from stainless steel sheeting as it was the cheapest solution and I'm amazed how well it works. When I get around to it I am going to put a plastic RV tank under it on a trolley with a pump attached so I can empty and fill it easily and roll it out to be cleaned out and serviced.

The drivers and motors are very basic Chinese ones. So I have a project to upgrade them under way. Not that there is anything wrong with them.

The crazy thing is that I was approached by somebody in the US to help them bring a table to market and the plan is to offer them for sale here in Australia and the US. All the time consuming hard bits have been done. Engineering, motor selection, Z axis, control box design, so it will come together pretty quickly from here. So 4WD parts might play second fiddle moving forward. Thats why I want to update the motors so it can be sold as a commercial prototype rather than a DIY table so I can upgrade to the new design..

Whilst I'd love you to use our yet to be built Z axis stage, I saw this by accident today and I thought the single rail and pinion drive was a pretty neat way of getting the weight down.

https://www.ebay.com/itm/Z-axis-and-X-a ... fresh=true

[robertspark]

... a ghetto build ...

It's not about the cooker it's about the taste of the cake and finished parts for the application .... Function over form always with a design engineer... It's marketing and architects who love form over function

[robertspark]

I always remember a program I saw 10 or so years ago about the race for land speed records and there was a statement about how Art Alfons built the green monster with nothing more than an arc welder and a pedestal drill

https://www.google.com/search?gs_ssp=eJ ... e&ie=UTF-8

And his reuse of a classified jet engine was ingenious but simple to rebalance it

[Rodw]

I must say I am a function over form guy but the guy I'm playing with is a form over function guy so it might work well! When you hear stories of guys filling their factory with the inferior red machine over the superior brand of white routers, I do see that form now matters...

[robertspark]

I must say I am a function over form guy but the guy I'm playing with is a form over function guy so it might work well!

It's a good mix

I am carp at business and I know it (tried it a long time ago) as I'm an analyst type of person... For business you need an assertive type of person... Unfortunately those two types just rub each other up the wrong way as one wants to do it right and the other just wants to get it done, get the money in and move on. If you can get both to work together or be a mixture of the two you have a winning combination.

https://www.google.com/amp/s/blog.black ... s_amp=true

I've got to ask red and white router?
I got a big Triton (you'll know it... orange) but the CNC has a Makita (green)
I know about the blind love of red and blue welders (I got a yellow one .. ESAB.) kind of like blue ovals and bow ties... But did not know there was a router affinity.

[EverydayDiesel]

Is there a type of aluminum I should be looking at?

[robertspark]

What do you mean type?

Profile or grade?

[EverydayDiesel]

Both really. I found some T6 that I think would be good for this application

4 x 4 x 1/4 wall 6061-T6 Aluminum Square Tube 3.5lbs /ft
3 x 3 x 1/4 wall 6061-T6 Aluminum Square Tube 4.5lbs /ft
https://www.metalsdepot.com/aluminum-pr ... quare-tube

I also found these from the JD2 machine, looks like aluminum to me

[robertspark]

T6 is the grade (hardness) , 6061 is the alloy
http://www.aalco.co.uk/datasheets/Alumi ... s_145.ashx

6061-T6 is a general grade which is fine, common and low cost.
+ you can weld it with 4043 and 5356 too
https://www.thefabricator.com/thewelder ... 6-or-4043r


4x4 is a bit overkill in my opinion

3x3 would be better (but everyone has an opinion and they are all different.

3x2 would be fine too in my opinion as you've got enough space to mount the rack and also allow for stability of the carriage... you can always make a wall / axis stiffer by slapping a plate on it and bolting it or spot welding it on to the section you want to stengthen.

think about your carriage sizes and the amount of additional rail you need in length to accommodate the carriages. A smart design will offset the gantry back to accommodate the torch position and travel (otherwise you are paying for bits of rail you will never use (A 4x8 table can be done at 5x9 rail allowing for 1' of carriage dead zone.... because you cannot use the 6" to the left and right of the torch (my Z & Y + Y' axis carriages are 7" wide as I'm stood next to them changing wiring).

likewise you can offset your rack to allow for the actual travel length you are going to use and not just span it across the whole of the gantry and Y axis.

(I keep looking for an interesting article written by an old timer I read about the cost of square inches on a milling machine that never get used but can't find it).

Have you done any drawings or sketches of what you are thinking about?
I suggest drawing it out in fusion 360 or one of the other 3d packages and it will give you a chance to learn how to draw too as once you've got the thing unless you are going to just cut what everyone else has drawn you are just copying work.
fusion360 is available free if you are a hobbyist (and if you aint selling anything yet... you are a hobbyist!)

I've tried learning freecad in the last couple of weeks and could not get my head around it so am using fusion 360 (but I've come from 20+ years of autoCAD, Revit and Inventor.... so fusion sort of makes sense a little (but different [easier than inventor])

I would really be reluctant to put orders in before getting drawings and cut sheets to hand and a say 80% complete bill of materials (there will always be nuts, bolts and washers etc missing.... just buy a bulk bag of bits and you don't need to count them out.... no need to draw your mounting holes out if you know they are 2" centres etc etc starting at a specific point....)