Re: motor sizing vertical slide EDM

Marcus,
Thanks for the info...

>Message 6
> From: "Marcus" implmex@...
> . Date: Thu May 11, 2006 8:09am(PDT)
>Subject: Re: motor sizing vertical slide EDM
>
>Hi Tom:
> Sounds like you've got a pretty comprehensive understanding of the whole
>process and are clear on your needs.
>One thing I'm having difficulty with though....how will you reconcile the
>mutually antagonistic goals of high speed versus fine resolution.
>I'm facing this same issue with a much simpler problem (CNC conversion of an
>engine lathe).

I design the system for hi speed, then dont use it :-)
Ususally just truncate the top velocity with some system parameter
Meanwhile I take advantage of the reaction capabilities to enhance
response at low speed.
I make sure it can respond to "1 vel"
which means, when at rest, give it the LSB of the controlling DAC and see it
move & viseversa(not move with 0 ).
This assures good low speed resolution but near
instantaneous respoonse to any other bit value handed to it.
This results in QUICK.

Top speed on todays' spreadsheets look like 610 rpm , direct coupled, 5mm
pitch, still means 120"/minute, a 3kW sureservo.com system, but limited
to maybe 1.5 or 2m/min (60 to 80"/min). Then insure that it can respond to
1 vel and it'll cut great. If it's overkill, well I got 9 more units to reduce
it on ;-).

>Finding a workable balance for my problem is a path well-trodden by many
>already, and I can follow in those footsteps, but you're accepting a far
>more expensive risk if you miscalculate.

Yep, expensive, but I'm gonna find out wether it works, then reduce.
Recent experiments say it the right direction. I'll overkill then backoff
later.

>Fine pitch leadscrew plus high rapid rates equals insane screw or ballnut
>rotation speeds (sounds like servo to me) together with the corresponding
>acceleration, screw whip, and thermal growth issues.
>Coarse pitch leadscrew to reduce motor speed equals poorer resolution and a
>far higher workload for the drivetrain components.
>Seems to me that the design choices of Hurco opted for resolution over
>speed, and Sodick had to break new ground to tilt the equation the other way
>successfully.

I still dont get linear motors for a gravity axis... my mind keeps thinking
sloppy, but its damn unlikely that anything Sodick Inova builds is anything
but very good. (I'm ex-AGie myself, also great builds )
Thermal mgmt? No, 1st I have to find out the characteristics of the given
frames. I'm rebuilding, not going from ground up. The reason for rebuild is
that very few shops can buy iron as stabile as postwar iron, even 60's iron
is way stronger than anything but the top of line stuff now.
More like Thermal notice, measure and workaround.

>Are you planning to do a sophisticated build with everything formally
>engineered, including thermal growth management?
>I don't see how you can really avoid it, given your stated goals.

I went to look at the roler screw, at that url,
it wanted me to sign in... I hate that,
why isnt there a damn catalog with prices and no salesmen?
(personal grip against all venditore... pirata! liars, cheats! )

>If this is going to be a high end project, have you considered using the new
>SKF roller screw instead of a conventional ballscrew?
>There are some significant advantages:

dang looks like I gotta talk to one of 'em.
I'm thinking of direct coupling, but was just going for 1.5" dia ballscrews
and single nut due to God & Gravity being on my side.

>First, you can get much finer pitches with hellaciously high load ratings,
>so you can often eliminate having to gear the screw down.
>Second, you can manage high cyclic loads far better, and for a lot longer
>before failure of the screw.(much larger contact area of nut to screw)
>Third, to the best of my knowledge, you can manage high speeds better
>(larger root diameter for a given screw OD therefore better screw stiffness
>at high speeds)

Is the root diameter differrence large (> 10%)?
I really dont know what these things look like.

>The only downside is that they carry a correspondingly audacious price tag.
>Check them out on the SKF website...they're worth a look at least.
>www.skf.com/portal/skf_ca/home/products?contentId=150201&lang=en
>Cheers
>
>Marcus

Thanks again
BTW so why do you know this stuff?
TomP