machine control memory

[JG]

Hi All,

We have a mid 80's Hitachi Seiki HC 500 Horz, iniatially it had 32k of memory, we upped it to 512k (11M fanuc) the service dept say's that is the max we can go, I find it hard to believe, anyone have any experience with this machine/control combo and upgrading memory?

 

thanks for the help,

 

Jg

[Storkman]

Wouldn't surprise me. Most of the new Fanuc's max out at 2 mb nowadays. 512 in the 80's sounds about right. You need a dataserver if you want to go bigger.

[Guest]

That is entirely possible

[Fig sign]

You need a dataserver if you want to go bigger.

~~~~~~~~~~~

You can always drip-feed (DNC) the program

[JG]

thanks for the replies, drip feeding did not work out well, we keep emptying the buffer, the control's rs232 is very slow, even at 9600 baud, I'm told it is due to the overhead of the CRAPS ~ err~ C.A.P.S. system it has on board. I was thinking it could be expanded further because the manual for the machine has a option for 1.5 Mb memory, maybe something got mixed up in the translation.

 

thanks again,

 

Jg.

[Fig sign]

yOU CAN SUCCESFULLY MINIMIZE THE size of your programs using subroutings and FILTERING ,DO YOU KNOW ?

iF YOU NEED TO DEVIDE THE PROGRAM ONTO PARTS ,(means to devide long toolpathec )use c-hook section.dll

~~~~~~~~~~~~~~

 

Break an NCI file into a number of smaller files for machine tool

controls that have limited memory. The size of the output files

is specified in Kbytes. The output files are named nciname1.nci,

nciname2.nci, etc.

 

~~~~~~~~~~~~~~~~~`

[JG]

thanks Islander,

I am filtering, and I'm aware of the "C" hook, (as a matter of fact I've used that back in the 90's to section the length and z stroke, it worked great) the parts are airfoils 23" long x 2.5 high, and I have 335k of memory available, currently I'm using surface contour or water line to keep the Z axis "locked" and produce favorable conditions for filtering to produce arc's, I can produce the part, now I'm trying to balance "time to cut" (step down) vs "finish" (surface tolerence,filtering tolerence and step down) vs "fit" (335k of memory).

 

thanks again,

 

Jg