Post by thread:[EE]: Multiple values accelerating at different ra

I've got this application were multiple RAM values in an array are given a target value which can be above or below their current value and must reach this target within a given time. For example : ORIGINAL VALUES A=010 B=020 C=055 D=255 E= 000 etc ... TARGET VALUES A'=020 B'=010 C'=055 D'=000 E'=255 etc ... time required = 1,000,000 processor cycles (50 Hz refresh rate with a 20MHz processor) Given these conditions : A must increase by 10 in 1,000,000 cycles B must decrease by 10 in 1,000,000 cycles C must not change in 1,000,000 cycles D must decrease by 255 in 1,000,000 cycles E must increase by 255 in 1,000,000 cycles etc ... The first obvious way to do this is : (A'-A)/1,000,000 = RateA = 0.00001 per cycle (B'-B)/1,000,000 = RateB = -0.00001 per cycle (C'-C)/1,000,000 = RateC = 0 per cycle (D'-D)/1,000,000 = RateD = -0.000255 per cycle (E'-E)/1,000,000 = RateE = 0.000255 per cycle and continuously multiply the values in memory by these constants until we're done counting 1,000,000 cycles, phew! Since I don't need such a high resolution (1 cycle) and couldn't possibly calculate all these values fast enough (with a PIC)I could simplfy these operations by updating the values only every 20,000 cycles (50 Hz refresh rate with a 20MHz clock) his still gives some rather hefty fractional values to handle with a PIC. Besides, this method requires twice the amount of memory to store the Rate constants ... Any thoughts ? regards, Tobie Horswill Le Projet Ex Machina, QC, CAN. spam_OUTthorswilTakeThisOuTexmachina.qc.ca -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam@spam@mitvma.mit.edu

> Since I don't need such a high resolution (1 cycle) and couldn't possibly > calculate all these values fast enough (with a PIC)I could simplfy these > operations by updating the values only every 20,000 cycles (50 Hz refresh > rate with a 20MHz clock) his still gives some rather hefty fractional values > to handle with a PIC. Besides, this method requires twice the amount of > memory to store the Rate constants ... > > Any thoughts ? How about a Bresenham algorithm? ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, olinKILLspamembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam.....mitvma.mit.edu

Olin, I did a quick search on the net for "Brensenhams Algorithm" and quickly found several explanations of this Computer Graphics method for selecting which pixel should be illuminated when drawing a line on-screen. It fits nicely with my objective, it's simply a matter of replacing the X displacement by the number of processor cycles and the Y displacement by the difference between the target and the original values =) I also found an Integer version of the algorithm which I think can be adapted quite easily to PIC assembly. =)) For those interested, here is one of the best documents I've found : graphics.cs.ucdavis.edu/GraphicsNotes/Bresenhams-Algorithm/Bresenhams -Algorithm.html Have you worked in computer graphics before or is this algorithm considered common knowledge ? Anyway, many thanks! Tobie Horswill Le Projet Ex Machina, QC. CAN. EraseMEthorswilspam_OUTTakeThisOuTexmachina.qc.ca {Original Message removed}

> Have you worked in computer graphics before or is this algorithm considered > common knowledge ? I used to design display controllers and first ran into Bresenham's algorithm there, however it's widely used in other applications. Jack Bresenham's original problem was to make a pen plotter driven by stepper motors follow an arbitrary straight line. I wrote a paper once about how Bresenham's algorithm can be extended to start and stop at fractional end points (subpixel addressing) and still maintain the same integer algorithm after setup. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, olinspam_OUTembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics