[project5k]

I'm sitting here working through a couple options and i think i know my math is right, but if someone would double check me please...

the facts:
5tpi ballscrew
10ustep driver
200kHz max step rate
direct drive from stepper to ball screw
200 step/rev motor.

my math sais that the max IPM = 20.. is this right? no wait, i think i just found my error.. i forgot to multiply the 200kHz by 60seconds to get IPM..
so then its 1200ipm.. nahhhh that cant be right...

here's how i got there...
200kHz is 200,000 cycles per second
so 200,000*60 seconds is 12,000,000 steps per min
5tpi*200step-rev*10ustep= 2000 steps-rev*5rev-inch= 10,000 steps per inch
12,000,000 step-min/10,000step-inch = 1200ipm

first off i understand that nothing i build will move like that, and secondly none of this is allowing for max motor rpm, max screw speed, friction or any thing like this... but anyway, if my math is right then i get 10,000 steps/inch, that would mean .0001 per step, not allowing for backlash, machine flex and so forth...

the more i do this, the more i'm second guessing myself...

where am i going wrong?

[hoss2006]

I don't see anything wrong with the 1200 IPM calculation.
Seemed ungodly fast for a stepper but I found this post discussing getting
6000 rpm from a stepper so who knows.
It might work on paper but in the real world the torque would be almost nothing
at those feed rates.
Interesting though.
Hoss

[project5k]

oh yea, i understand that its not likely to happen, spinning a stepper that fast is probably far beond what i can do at this point and time.. but i was just making sure that i was doing the math right.. cause somehow i missed the seconds to min conversion first time around.. no, in all reality if i can get something in the 60 ipm rapid neighborhood or 10K steps/second, then i would be jumping outta my skin happy... thats only 300 rpm at the stepper, very do-able from my experience..

ok, thanks Hoss for answering....

[Stepper Monkey]

I think a big part of the discrepancy was just using the max speed at which the controller can send step signals to the motor to determine actual stepping speed. If it were actually possible for the motor to respond that quickly, the stepper would be getting driven at 6000 RPM! Yikes. If only, right?

[dertsap]

I think a big part of the discrepancy was just using the max speed at which the controller can send step signals to the motor to determine actual stepping speed. If it were actually possible for the motor to respond that quickly, the stepper would be getting driven at 6000 RPM! Yikes. If only, right?

i agree
you need to look at the motors specs and calculate according to the torque curves

[project5k]

yea, i get all of that, i was just making sure that i was doing the math part correctly, cause at one point my math said that i could only get a max of 20ipm and thats just too slow.. i know to check the torque curves, and raising the Vdrive helps and so on, i also know that there is a max you can spin a certian diameter screw mounted in certain ways certain lengths before whip becomes an issue.. bla bla bla... all i cared about was that the formule that i was using was working...

[dertsap]

just trying to help man

[project5k]

i appreceiate it... even if i can't spell it.. hehehehe

[Robin Hewitt]

thats only 300 rpm at the stepper, very do-able from my experience..

Your experience agrees with mine, I find little useable power above 5 rps on a stepper.

In your experience, do you think the motors will actually resolve 10 microsteps when you start driving a slide? Between step positioning is springy and even the best slide oils don't completely abolish the dreaded stiction.

OTOH, I am no expert and with about 1.5 thou of backlash my .25 thou step resolution is about as useful as a chocolate teapot

Perhaps the maths could be more understandable if you converted to revs/second at the offset.

200,000 steps/second divided by 2000 steps per rev = 100 revs/second.

100 rps on a 5 tpi screw is 20 inches/second, done.

[project5k]

i see your point... well in the real world, i would hope for something in the 60+ IPM range for rapids, but knowing that my cut speeds will be much slower than that, and considering that when cutting is when you need more power, then it works out in my mind to be about as good as i can get on my budjet... I'm a little concerned how much backlash i will have once i get my ballscrews installed, but as for right now i really have no idea.. just have to see what it is and then figure out how to reduce it...

as for resolving 10usteps, its not really about getting it, cause nothing i'll ever make will need more than oh say .001 accuracy, maby .0005 at the absolutely tighest... for me the ustepping is more about resonance damping, and smoothness of rotation.. by subdeviding the full stepps and running multiple coils at different % of power, the transition between this full step, and the next will be considerably smoother, and thus less likely to resonate... or at least thats what i have in my head right now.. application may very well teach me otherwise...

[Robin Hewitt]

The backlash doesn't exactly bother me either, I just insert a boring head when I want truly round holes to seat bearings in.

I am redesigning my X screw mounts to have an axial ball race at either end and a hefty die spring trying to stretch the screw, (bought some green springs on ebay that don't move very far when I stand on them).

With the Y axis I can only hold one end which is tricky fitting motor and handle. I have decided to fit the spring, two axial bearings, a gear and then the motor. The handle connects to the gear so it turns in the old familiar direction without needing a left hand screw.

The Z axis can't fit a motor directly on the end so it has to come in from the side.

I couldn't understand why my z axis assembly pulled apart. Turned out that locking the quill was not enough when I hammered out that tight fitting taper on the collet chuck, I now release the screw before hammering

Only a few months experience but enough to appreciate some of the problems, so here goes for what it's worth...

It does a lot of up and down movement so that wants to be quick.

Beware the sudden reversal at rapid speed, you need to apply accelerations but not all the time.

When your tool discovers a piece of hard steel buried in that aluminium block, you discover how flexible your 'rigid' set up actually is.

Carbide tooling allows faster quill speeds than HSS and sounds good but my mill simply doesn't spin quick enough to warrant it. I find HSS cuts better in aluminum and I quickly discarded the listed feed rates.

I'm finding that with a bendy bench mill feeds around 1.5mm/second are about right. OTOH a fast transit between cuts saves a lot of time because there is a lot of that.

My big bugbear is chatter on the tool tip which leaves horizontal lines on my lovely vertical surfaces, especially when I want to go deep and have to use long series cutters. They seem to be dependant on cut direction and echo on the deeper finishing cuts so they never go away. Still working on that one.

An OMIGOD button that stops all X, Y, Z movement instantly, preferably without losing position, is a seriously good idea

[project5k]

yea my estop button will probably have "OH FECEES" on it.. hehhehehe
i still havent gotten to work on mine, still fighting with getting the last couple driver boards just the way i want them, the first one came out so nice first try, but i'm having hell replicating it.. for somereason, anyway.. between that and traveling all over the place, i just havent gotten anything done.. been to portland, been to new orleans, went to the coast fishing, bla bla bla...

[Robin Hewitt]

yea my estop button will probably have "OH FECEES" on it.. hehhehehe

You know you're in trouble when you are stalling the x axis with your right hand, the Y with your left, kicking your shoe off hoping to reach the keyboard before it starts moving in the Z