Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....

Babar wrote:

> now while pc still turned on and machine off, i
> detached P1 cable from STG-II P1 connector, Detached
> LPT cable from lpt connector present on pc and last I
> detached the P3 ribbon cable from the breakout board.
> Even this time I saw the same errornous counts
> although
> the machine was all off, all the DAC, encoder
> connections were open
> nothing was turned on except computer but these
> errornous count still
> generate. This seems in my case this problem is
> explicitly with
> the STG-II board itself. since there was no connection
> of STG-II board
> with external world this time, but still i saw the
> errornous counts generated in stg2diag
>
> Do you think that this may be due to not very good
> powersuply of
> my computer itself? do you think that changing of
> computer may result
> in removal of this errornous count problem? anyways i
> plan to do that
> as a last effort.

I think you should try a different PC, with a different motherboard, at the very
least, to see if that will run better. If it also has the same problem, I would send the
STG board back to the factory for repair. If the other PC works better (hopefully
100% normal, no jumping counts) then I think it must be a compatibility problem
between the ISA bus of the motherboard and the STG board.

>
> cant the module that reads the 24 bit counters of
> LS7166 coded
> again to fix this bug? Isnt there possibility to
> ignore these errornous
> counts in software without depreciating the accuracy
> of EMC system ?

EMC already does this! But, if more than 50% of the position reports are bad, how on
earth can software determine which one is the good one. Anyway, on most PCs, this
problem is VERY sporadic, about 1 - 2 events per second when reading the position
1000 times a second. EMC can supress this error rate without much effect on the performance
of the servo system. But, if long strings of position reports are bad, it means the servo
system goes open-loop until the encoder counts come back within normal range. This will
cause massive instability in the system.

Jon