Gear reduction understanding

Personally, I think the "1 inch rule" came from Tom Caudle at CandCNC.
Jim Colt's rule of thumb was 0.3G for acceleration. Any more than that and the inertia really shakes things around so the table and gantry needs to be very stiff and probably bolted down.
I never shared my dial indicator tests for a reason

The other factor to consider in the design is are you outside of the resonance zone at cut speeds? Gearing helps here.

Hypertherm talk about centripetal radius limits. Your acceleration determines the minimum arc radius you can negotiate without slowing down. Faster acceleration will give crisper corners.

Use a TB6560 drive with microstepping, to much vibrations for direct drive.
Use a TB67S109 (ebay TB6600) drive with microstepping, it is much smoother.
Use a LAM Technologies drive with microstepping, very smooth as it can do 256000 or more micro steps.
-
Servos are a different beast-
-
The 1 inch/turn rule is old, it should be used as a rough estimate, anything within reason will work with minor differences.
-
Homework for those all:
-do a move back and forth on one of the axis while lowering the velocity till it hits resonant frequency and the machine starts vibrating,
-do the same with the other axis,
-try to avoid those speeds while cutting
-problem solved.

tommylight wrote:

-
The 1 inch/turn rule is old, it should be used as a rough estimate, anything within reason will work with minor differences.
-

 

What is the "1 inch /turn rule"?

Is that a rule of thumb that the designer should target 1" of linear motion per motor revolution? 

Well, at the rate of which this conversation has escalated I no longer have anything to offer. All of this is over my head. I need to do more research and learn better about motors, voltage, torque, and all things involved with that.

Here is where I stand on plasma CNCs: I have built two, the first one with the idea of starting a business and not having enough money to buy a 15k plasma table. It worked great and I learned so much. After making it though I knew I could do so much better. So I sold it. Well upon trying to sell it I was commissioned to build another table. So I did, it was much better then my first but I still know it could improve. So before I offer to make any other table I would like to perfectly fine tune and research all the little things that would improve my table design. I love building plasma tables. Im a welder by trade.

just figured I’d give some background

spumco wrote:

Is that a rule of thumb that the designer should target 1" of linear motion per motor revolution? 

[/quote]
It somehow became a rule at the time when microstepping was very expensive and there was no "resonant frequency compensation" of any kind in the drives, so with reduction steppers would turn faster hence lower vibrations.
Try to run a direct driven machine with no microstepping (drives set to 1/1), it is a terrible noise !

Aldenflorio wrote:

Well, at the rate of which this conversation has escalated I no longer have anything to offer. All of this is over my head. I need to do more research and learn better about motors, voltage, torque, and all things involved with that.


 

Everyone has something to offer - even it's just anecdotal evidence of something not working as desired.

In your case, you had/have a perfectly acceptable plasma table without a reduction system.  If you know, more or less, what the smallest corner radius you could cut (at high speed) with acceptable quality is, then let everyone here know and we'll all be better informed.

For example:
Gantry estimated weight - XYZkg
Motors & driver voltage - Nema34, 1200oz-in, 80Vdc
Pinion diameter - XYZmm
Smallest acceptable corner radius:  XYZmm @ (insert metric nonsense) per minute on 1mm (steel/aluminum)

You can certainly contribute something like that.  Remember, most of the 'above your head' stuff is theoretical.