Re: motor sizing vertical slide EDM

TomP,

Not quite sure how to reconcile "15Krad/sec^2" with "40mS response
time" and "8X acceleration that I stated".:-)

I use 15Krad/sec^2 as very respectable acceleration goal on a
servomotor. This reflects a very good motor torque to inertia ratio.
This value is motor-bound in that all mechanism acceleration rates
will be limited to or below this value.

You said the goal is 2"/sec on a 5TPI screw (600 RPM) reduced from
the motor's 3,000 RPM via 5:1 reduction.

Using that accel rate, the motor will take 21mS to reach 3,000 RPM.
The linear velocity will go from 0 to 2"/sec in the same period of
time if I understand your mechanism correctly.

Solving for "a" (a = v / t) gives 95.5"/sec^2 or 0.249 G (G =
384"/sec^2).

The power needed to accelerate a 700lb mass at 0.25G to velocity of
2"/sec is only 39.5 Watts, the energy invested is only 0.83 Joules.

Mariss

--- In CAD_CAM_EDM_DRO@yahoogroups.com, tomp-tag <tomp-tag@...> wrote:
>
> Hello All:
>
> Graham:
> thanks, To get fine reaction ( measurable motion when the
smallest change in
> control voltage is detected), Hurco used large gear reduction (
meaning the
> motor kept moving more ) and fine pitch lead screw ( smaller motion
but
> easily discernable by the control system). Good idea, my AGie AB
uses 5000
> rpm dc motor with 5000:1 double worm reduction and 2mm pitch lead
screws.
> This also allow the motor to stay out of the 'dead' area , respond
to very
> small changes in control voltage, and move very small amounts.(max
velocity
> is 2mm/minute! but can cut with .01"x.01" tools )
>
> Mariss:
> thanks,
> 1) yes the reversals create near 3500 lbs force, didnt know that
was 5G. That
> general idea was why I was opting for 1.5" dia ballscrew, maybe it
oughta be
> bigger!
> 1a) reasonable systems react at 40ms ( acceleration times that are
8x of what
> you stated, so it's true that better reaction will be had with the
more
> powerful motors). The reversal is misleading tho, at the beginning
of the
> reversal, the velocity will be near 0 but 'forward' ( EDM is slow
but quick,
> sort of 'nervous').
> 1b) I'm sure I'll be buried inside the accelration ramp time
during my
> reaction, and wont arrive at the full velocity. The reason for the
outlandish
> desired velocity is to decrease this time. Rather than the
slow 'rapids' of
> most machines ( a result of good control at low velocity
sacrificing top
> velocity ) I opt for higher rapid rates only to increase the
reaction speed
> at low velocity ( this gets me the 'quick' ).
>
> re: question
> Does the use of the high velocity to get 'quick' make sense? is it
valid?
> Recent experiment showed systems with nasty old generators ran
sweet when the
> mechanics and motors were changed ( 150W dc brushed to 400 watt
Yaskawa ).
> Just this change contributed to the control loop enough so the
system
> suddenly competed with new equipment.
> 2) that's a lot of calcs, I gotta study ;-) thanks
>
> Marcus:
> Anytime the EDM control loop handles the gap position better, it
runs better.
> It isnt very noticeable at 'rough' but is always noticable
at 'finish' ( what
> rough and finish are to different disciplines varies a lot ). But
systems
> always respond better when quicker and able to resolve finer
position.
> I know about crawling: imagine systems with 3 meters of X travel
and rack
> type tool changers! The operators avoid changing tools!
> I plan on servo but see a great advantage in using steppers on the
contol
> side. It's just that steppers of that power are more expensive and
hot. I
> could use the amp styles that take servos with rotary encoders
and 'step'
> them.
> re: linear motors: yes the street buzz is that Sodick Inova got
near 5G on
> those 'jumps'. I've gotten near 2.5 with the Yaskawa setup
mentioned above,
> and suddenly could cut 'no flush' with simple generators. Quicker
is better!
> re: mother of all rams: Nope, several machines have much larger. A
SIP in
> Elmhurst Illinois was made for steel on steel, and the ram face was
1.5 meter
> square, It was hydraulic, didnt move as fast as we said earlier,
but way way
> over 700lbs.Also many of the older Nassovia's ( bought by
Charmille, bought
> by Geo Fischer) were much bigger. I've got 350 lbs of platen based
fixturing
> to begin with, and also have a 300 lbs slide ( the motor sees near
700
> static ).I can actually ignore the weight of the graphite
electrode! It
> doesnt change the equations enuf!
> The ability to burn deep skinny ribs can be viewed as an overall
performance
> test of your system. The systems that cant cut .020" X x 1.0"Y x
2.0"Z will
> be having trouble with simpler forms, and the better system wont.
With the
> new Yaskawas, I cut 10mm deep with a 250um wire, no spin, no flush,
just
> jump! Quick is good, stiff is neccesary.
> In this control type, the ability to cut small deep and fine
are 'poles' in
> the equation. Quicker expands the distance between the poles, and
makes the
> system more 'robust'.
>
> Thanks guys!
> I think that I've gone overkill, but not by much. I doubt I'll
reduce
> anything till I test it on a real system. Get pragmatic, get dirty,
then
> reduce.
>
> regards
> TomP
>