Code check and refinement suggestions?

[HenrikOlsson]

It depends on where in the code you clear the screen.
If you only clear the screen at the start of the program then yes, you need "erase" whatever garbage that might or might not be there - which it looks like you ARE doing with the code you just posted.

If you clear the screen at each update (ie you refresh the whole screen) then you obviously don't need to "erase" any "garbage" that might be on there - which is the method in the snippet I posted earlier.

Up to you to decide which is the most appropriate. (Thinking about it I'd probably go with your method of "erasing" whatever might be there, it'll be quicker but MIGHT take slightly more code space.)

/Henrik.

PS. Are you actually running this code on anything to see if it works or are you just churning out code "blind" so to speak?

[kevlar129bp]

Hello Henrik.

If you don't see a large problem with me wiping the individual characters, than I'll give that a run. As far as "live" testing, not as of yet. I designed and etched a board some time ago that I just plop different code on when it hits me. As far as the purpose of this code, I have dreams of hooking it to a dc brushed gearhead motor that I fitted with an encoder (not a true servo motor ) that will be mounted to a rock tumbler. The board will monitor, as I'm sure you can tell, voltage, current, temp and rpm (through the index pulse of the encoder). The power supplies are 2 HP ESP120 (51.4V, 57A) 3KW server supplies wired in series. Right now, you're probably thinking WTF! Why would he run those? Well, all I can say is, they're the cheapest power you'll ever find . I've juiced 'em in series, and they haven't missed a beat. The only tricky part is isolating the high-side supply from earth as well as humans. I have them both in a fully insulated 1/4" polycarbonate enclosure. As far as wanting the code for the pic finalized (or almost) before the first run is nothing more than it's kinda a pain in the butt doing the remove, burn, reinsert routine (the LCD is mounted on stand-offs above the chip carrier as well as no ICSP on my board). If you think I've got workable code, I'll go for it. Do you have any slicker ways to get finer RPM from my encoder scheme? I have to admit that I have brain fried over that part!

Thanks again Henrik,
Chris

[HenrikOlsson]

Hi Chris,
As long as it works I have no problems and, like I said, "erasing" the individual characthers sounds like the best solution.

Sounds like a fun project! Be careful, lots of power there!

Since your chip has a USART and you're laready using it to communicate with a PC (by the looks of it) I'm assuming the board has a MAX232 or similar device so why not use a bootloader to get the code into the PIC? (And next time, put a ICSP header on there...)

As for the RPM, since you apparently aren't interested in direction and/or position why not run one of the A/B channels of the encoder to a timer/counter input and configure the appropriate timer in the PIC as a counter. Read it at every TMR1 interrupt and you have motor speed.

/Henrik.