Answers Needed Should I Search Or Should I Post? How To Get The Best From The Cnczone

[NC Cams]

Adobe: Some people have a mind for figures, some have a head for cards, some know drill sizes and fractions by heart (some I do, most I don't), while some folks brains are filled with minutae - mine tends to be cluttered to extremes with technical minutae of ungodly proportions.

When that fails, I refer to a cheat sheet. In it, I copied and pasted links on the website to the commonly asked/replied to inquiries that I or someone else posted. Also pasted links to "important neat stuff" that needed to be remembered that unless written down, would soon be forgotten. Comes in real handy....

It is one thing to know the answer to stuff, but I learned from a neat old college prof that you don't need to know the answer. Rather, only the key "trick" on how to figure it out or locate the answer.

It also helps to know where to FIND answers to questions/problems. This was something that I learned during the creation of college research papers and in my formative years as an automotive engineer - namely, how to research technical stuff. The fact that I did this sort of research at my last employer in the auto industry should not be that much of a surprise...

It used to be hard to find information because you had to go to the library and use books - now it is a veritible cinch to do research due to the volumes of info available on the internet. PDF's take up so much less space than catalogs but I still keep my well worn catalogs.

Being single, I can pretty much focus on what I want to and let the other stuff slide - I can search for stuff on the net day or nite or whenever the urge suits me. Sort of became an information junkie.

I do recall certain VERY critical birthdays and which florist to call who can send flowers when I forget to send birthday cards.

The technical memory comes in handy as an engineer - it enables me to solve a lot of my own problems and get involved with a plethora of things. Besides it is fun to generate a "how did he know THAT????" reply.... It also enables you to prevent competitors from taking unfair advantage of you and/or your capabilities....

What was the question????? grin

[Adobe Machine]

NC Cams..Thanks, I also remember obscure facts, tolerances, numbers, torgue values, interferance fits, slide fits etc..but I do think that even with indexing, and the computor you have a real talent and ability to concentrate, KUDOS my friend, you have tained your mind well, a lot of us appreciate your knowledge.

Adobe (old as dirt )

[NC Cams]

In reference to post #12 concerning direction changing issues, I offer the following background information:

A couple pieces of dynamics information will be of assistance.

When something is moved, the displacement is measured in some scalar quantity (inches, feet, mm, etc).

The rate of change of displacement is considered velocity (ft/sec, in/ft, mm/sec, etc).

The rate of change of velocity is acceleration. THis gets a bit tougher as now you relate a change in velocity in a unit of time. The terms now are "ft/sec per sec" or ft/sec-sec. Acceleration is a byproduct of the external force that was applied to change the velocity of a body at rest or in motion.

There is also a rate of change of acceleration and this is referred to as "jerk" or the terms of "ft/sec-sec-sec". Jerk is also a quantificatio of the RATE AT WHICH FORCE IS APPLIED TO A BODY. Example: slow gentle poke versus sharp poke in the ribs - same force but the rate of application makes a big difference in the "feel".

These factors determine/quantify not only the force required to do stuff but also how fast the force is/has to be applied.

To move something a foot can be fairly easy if you take your time and don't work up a sweat. However, if you want/need to move something a foot REAL REAL FAST, you'd better input a lot more force. The faster you want to do it, you also need to apply more of it faster.

How does this pertain to steppers or servos?

Simple.

The faster you try to do something, the more force it will take to do it. Whether it be to accelerate it or decelerate it - faster activity (given the same mass), simply requires more force.

While experiementing with our servos on the Eztrak, we ran circular cutter traces - these force the motors to simultaneously accel/decel thru min and max velocities repeatedly thru the whole trace path.

Well tuned motors with minimal backlash, regardless of type, will cut round circles and have minimal "flats" at the 4 direction change points IF YOU RUN THEM SLOWLY.

As soon as you start to speed them up, we found more and more issues with "follower error" (electromechanical gremlins) and other issues more related to inertia and dynamics. THe faster we went, the circle got less round and more irregular and vice versa.

Essentially, whatever power you input to get something moving, you have to put an equal amount in to slow it back down IF YOU WANT IT TO FOLLOW THE SAME VELOCITY PROFILE in both directions.

IF you want it to stop FASTER than you sped it up, you'd better find a way to externally apply a braking force. Otherwise, the mass in motion (inertia) will overcome the braking force available from the motor/geartrain.

It all comes down to F=MA where

F= the force applied to speed up or slow it down
M= the masss of the system in motion or at rest.
A= the ACCELERATION force input by the motor or the DECELEREATION force applied by the motor or brake to slow it down/stop it.

The velocity, acceleration and jerk are purely byproducts of what the motor is capable of doing as it acts or is acted upon by the mass being moved or trying to be stopped.

If guys would look at their CNC'd whatevers as cars that someone is trying to drive around a course, they'd SEE that you need varying amounts of control forces to affect the motion path that they are trying to generate. The faster you want to do it, the more horsepower (wattage) you need - period. BTW, the braking system in most cars generates WAY more horsepower than the engines - the cars stop WAY faster than they can accelerate.

Why would servo or stepper motors act/react any different? They have to follow the same laws of physics, no?

Hope this helps...

[Adobe Machine]

Fricton ? Gravity ? formulas ?

Adobe (old as dirt)

[NC Cams]

Adobe: as you well know, the devil is in the details.

Basically, the novice comes up with a "spec" whereing they want to cut whatzits at 100"/min and ask what size motor. To REALLY answer that question, you need to know the mass of the devices being moved (so as to find the equivalent inertia), the coefficient of friction between the moving member and the supporting member (to find out frictional resistance) as well as gear ratios, motor inertias and a bunch of other minutae that add up in the equations.

Having worked as a bearing applications engineer, I can tell you that professional machine tool designers can't answer all the questions needed to "do the math" - it stands to reason that "I dunno" will be uttered more often than not by the DIY'er.

Unfortunately, "I dunno" is an input value that results in too many unknowns in an equation thus making it unsoleable.

Ok, make some assumptions. By, as an experienced business many, you already know that when you assume something, you make an "a$$ out of u and me".

THere is a lot of charm in actually measuring the force needed to turn a mill or lathe handle - this will give you a NET force that takes all the friction and inertia into relative consideration. You can then turn that into a motor requirement spec and proceed with a form of calculation.

HOWEVER, when you add heretofore unspecified acceleration factors into the mix, the calculations become MUCH more complex. Bigger motors WILL accelerate things faster BUT the larger motor will also have more inertia thus they will be harder to slow down - see prior post about power versus rate of velocity change.

Formulas????: Mariss gave a neat disseratation regarding the "basics" on that very subject here:

http://www.cnczone.com/forums/showth...078#post214078, see post #15

[Geof]

Everything NC says is right on the money but my experience from the first (and only) CNC machine I built over 20 years ago combined with more recent reading is that you can shortcut a lot of stuff.
Jump straight to this step: "when you add heretofore unspecified acceleration factors into the mix, the calculations become MUCH more complex. Bigger motors WILL accelerate things faster BUT the larger motor will also have more inertia thus they will be harder to slow down..."
If your machine is stiff enough and your motors big enough to handle the acceleration loads and give acceptable rapids it is almost certainly stiff enough and powerful enough to handle any cutting load.

[NC Cams]

Geof: the issue I see is the guys come up with some "vaporware" speed at which they want to move the axis, irrespective of whether or not you can cut the material at that rate.

Perhaps with routers or the like you can go zooming right along but when you cut metal you can more oftern than not easily exceed the ability of the cutter with regard to the axis speed that you can jog at.

From what we learned via experiementation, if you want to reduce follower error, SLOW THE FEED RATE DOWN. Both cut accuracy and finish (when cutting metal) are usually improved.

[Geof]

Originally Posted by NC Cams Geof: the issue I see is the guys come up with some "vaporware" speed at which they want to move the axis, irrespective of whether or not you can cut the material at that rate. Perhaps with routers or the like you can go zooming right along but when you cut metal you can more oftern than not easily exceed the ability of the cutter with regard to the axis speed that you can jog at. From what we learned via experiementation, if you want to reduce follower error, SLOW THE FEED RATE DOWN. Both cut accuracy and finish (when cutting metal) are usually improved.

You are correct slow is more precise but with modern machines it is sometimes surprising how fast things can be pushed and still keep with better than a thou tolerance and when this is acceptable for the part and time is money that is the way to go.

However, really what I was getting at is that the biggest load your structure and servos will ever see is during acceleration so design for that and then you are sure you can handle the cutting. Also you can always back down the oomph on an overdesigned and built system but the reverse is not possible.

EDIT: I forgot this: http://www.cnczone.com/forums/showthread.php?t=22589

I had a good example of the follower error a while back.