Re: .ini documentation

[robin szemeti <robin-at-rszemeti.demon.co.uk>]

On Wednesday 18 September 2002 23:43, Paul wrote:
> Hi Robin
>
> 200mm per rev - That sounds like a *very* coarse feed screw. Are you sure
> it is a 7.87" pitch ? (or do you already have some gearing on it ?).

feedscrew? .. its a 40 tooth cog and some 5mm pitch linear toothed belt. 
Eventually I will gear down another 4:1 to get the motor into its ideal 
operating range ... the motor produces a nearly flat 4nm of torque up to 
around 2khz (20khz into the microstepping driver) ..  I've assumed I'll not 
get more than 10khz out of EMC which gives me about 300rpm at the motor ... 
thats 60m/min .. way too quick .. so gearing 1:4 or even 1:8 should get the 
8m/min I need.

Remember this is for a CNC plasma .. cutting speeds are high .. forces are 
small (except for accelertaion of the gantry) .. gear ratios of this order 
are about right .. 4nm of torque (with the 8:1 gearing) should still give 
(top of head) 1000 newtons of force, more than sufficient

> Basically, SCALE is equivilent to the steps per inch or millimetre, so
> using your "one step = 0.1mm" we get :-
>
> For steppermod with UNITS = 1
> INPUT_SCALE = 10   0
> OUTPUT_SCALE = 1   0
>
> And for UNITS = 0.03937007874016
> INPUT_SCALE = 254   0
> OUTPUT_SCALE = 1   0
>
> Running EMC with UNITS = 1 will probably produce following errors with the
> input scale set so low.

it does indeed  ....

> Increasing MIN_FERROR to perhaps 0.10 might help,
> but I'd be inclined to set the machine up with imperial units and set
> RS274NGC_STARTUP_CODE to G21 if you want metric.

Hmm .. possibly .. although imperial units aren't used much in England 
anymore (well .. I dont use them much anyway) .. it would be a bit of a bodge 
.. and I wonder what the rounding errors will do when to accuracy over a 
length of 9000mm?  Guess not .. as its got enough decimal places ... still 
seems a bit of an 'odd' thing to do though ...

OK .. and what does output scale do? ..  doesnt seem to have much effect ??

> The P I D parameters will have no effect with steppermod, so can be safely
> ignored.

Oh .. ok I misread something in the emcstepmot.c about P then .. fairy nuff 
.. must have been a different P :)

> Regards, Paul.
>
> On Wednesday 18 Sep 2002 10:51 pm, robin szemeti wrote:
> > What I have is a 9m (9000mm) axis with a 200 step motor and 10:1
> > microstepping (2000 pulses/rev) that provides 200mm of movement per rev.
> > (ie 0.1mm per pulse) ... yes I know its only a simple matter of division
> > or possibly multiplication .. but as yet I've found no clue as to divide
> > what by what ... and into which of the rather odd variables to poke the
> > value.

-- 
Robin Szemeti