Re: Ferror & minferror

Steve,
Thanks for your reply... I think you have rattled some grey matter...
In fact The machine is running in "inches" and yes 0.1 does seem
huge...but it works...tighter settings ie ferror = 0.005 minferror =
0.001 have cause much pulling of hair in the past.
What you said about when the proportional ferror can't be used it
reverts to the fixed minferror... makes a lot of sense. I was under
the impression that they were both proportional values and... that
was giving me the "headache".
The beast should be capable of tight tollerances. 2.5 hp ac servos
running a machine about the size of a Bridgeport Boss.
I will try to quiz the guys at EMC users... I am a member of that
august group...but I have lurked so long I have no idea what my
password is...but I can fix that.
thatnks Mike


--- In CAD_CAM_EDM_DRO@yahoogroups.com, Stephen Wille Padnos
<spadnos@s...> wrote:

>
> zbushz wrote:
>
> >Well we finally got " the Beast" up and running under EMC2. The
> >"Beast" is a Ramill VI, Big Iron, VIA mini computer, running Linux
> >(BDI rc-46) Motenc-100 motion board driving brushless AC servos

that

> >has been in the works for sometime.
> >It was running under EMC about a year ago... but we were plagued

by

> >following errors. I'm wondering if anyone has a good starting

point

> >for setting "ferror and minferror" in the *.ini file right now I
> >have it set at
> >ferror =0.100
> >minferror = 0.01
> >I just looked at the EMC Handbook and I don't really understand

it...

> >
> >Any words of wisdom out there?
> >thanks
> >Mike Bush
> >
> I can tell you how the two parameters work, but not too much about

how

> to tune them for your machine. You should probably post this

question

> on the emc users list:
> emc-users@l...
> I can say that they (the real experts) will probably need

information

> about encoder resolution and motor power to be able to help much.

Also,

> it looks like you're using mm units (0.1" would be a huge following
> error) - that kind of info is useful as well.
>
> Here's how they work:
>
> Emc calculates the allowable error as a fraction of FERROR,

proportional

> to the present feedrate as a fraction of the max feedrate.

basically

> the calculation is allowed_ferror = FERROR * (current_velocity /
> MAX_VELOCITY). If the result of the calculation is lower than
> MIN_FERROR, then emc uses min_ferror instead. The reason is that
> without the MIN_FERROR clamp, you'd effectively have 0 allowable

error

> at low feedrates, and the machine would stop all the time with

following

> errors.
>
> Assuming that you have a MAX_VELOCITY of 5000 mm/M, and you're

running

> at 100 mm/M, this is the calculation emc would use:
> allowed_ferror = 0.100 * (100 / 5000) = 0.002
> since 0.002 is < 0.01, allowed_ferror is set to 0.01
>
> Hope this helps
> - Steve
>