I am better do some real machining and research first. Good news is I am
basically ready to go. Yesterday graphics card freeze the screen and Eclipse
refused to work but now everything is good.
EDM machine is located in cold garage and temperature is to low this weekend so
I postpone test run but next weekend look more promising.
As use my own hardware run on Micro controllers I have been thinking a bit
suitable protocol or more exact message format for protocol since it would be
good if messages could be sent over several different medias: Ethernet, maybe
RS-485, UART, SPI. CANopen might be a choice since it have both PDO and a
dictionary for parameters.
>
>
> On 01/06/17 15:55, Nicklas Karlsson wrote:
> >> Nicklashello
> >> the time from voltage turn-on until the voltage dropis the ionization time
> >> this can be used as an indication of the gap 'clean'.
> >>
> >> as the cut progresses and more debris is generated,
> >> this time measure will change and decrease
> >> because the debris will reducethedielectric quality (the anti-conducting
> >> quality).
> > Time for a peck cycle?
> no the ionization time decreases as the conductivity increases ( due to
> debris in the oil, makes it more conductive )
> >
> >> IF the time is too small,
> >> then the gap is dimensionally too small ( too close )
> >> or too dirty ( needs 'peck' cycle )
> >> or more fluid flow ( sweep up the dirt )
> >> or more offtime or a combination of these
> >>
> >> this ionization monitor cause & effect loop should be faster than 100mS
> >> you cant change flow rates that quickly but you can add offtime to
> >> subsequent pulses
> >> other strategies are in scientific papers
> >>
> >> re: The sparc counter
> >> not sure what the count deal with yet
> >> it sounds like you could get a number related to (sparcs_attempted /
> >> sparcs_actual) = sparc_efficiency
> >> and then see if adjusting edm servo target voltage or offtime would
> >> increae that ratio
> > Yes "ionizations attempted"/"sparcs", then to high a peck cycle is needed?
> well, the ionization time is a better indication that peck cycle is needed
> and
> ionizations attempted / sparcs ( a low percentage ) can be used to
> indicate
>offtime needs to be increased
> and/or
>gap distance needs to be increased
>
> if changing offtime or gap distance has no appreciable effect, its
> usually best to return to previous values
> no effect indicates the system is tuned well enough already
>
>
>
> >
> >> adaptive control in edm is a great idea
> >> BUT
> >> make sure the adaptive control can be turned off!
> >> many times in AGie and Elox and Mitsubishi cutting, we found that
> >> turning the autopilots off was neccesary
> >> ( let ME drive the damn thing!)
> > I wait with adaptive control until I figured out how it works but to
> > present some kind of average seems like a good idea.
> >
> > There is:
> >Number of ionizations.
> >Number of sparc.
> >Some theorethical numbers.
> >It might also be possible to measure some times like time to ionization
> > or other if this provide some useful information.
> i thought you were measuring time to ionize already, so i am not clear
> on what the 'number of ionizations' vs 'number of sparcs' means
> one ionization to me is [ a rise of voltage towards Uopen, then a drop
> towards Udischarge ]
> i think you must have a different event for one count, like just the rise
> a rise ( without a fall ) will occur when tool is too far away from work
> to sparc
>
> yes, get some experience before tackling adaptive control system
> watch a lot of sparcs! they are quite interesting visually, on voltage
> scope, on current scope, on frequency domain scope
>
> just for interest
> the auto jump is very important for thin deep slots ( called 'ribs' )
> and heres a paper with a pulse discriminator, and auto jump circuit
> description ( no schematic or CPLD code :-(((
> https://www.researchgate.net/publication/251896377_Control_Strategy_for_High_Speed_Electrical_Discharge_Machining_Die-sinking_EDM_Equipped_with_Linear_Motors
>
> HTH
> tomp tjtr33
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
