Post by thread:[pic]: Problem with PortB

strange Hi to all, when I set break point on portB bit 4 other pins turn on too and shift_left doesn't work I worked on it whole day and I gave up. I appreciate if any one can see what I am doing wrong. I am not sure if what I put together is real SPI interface what I wanted to do is to make spi interface. here is partial code. I am using 18f452 with mplab ccs plugin thanks Andre #include <18F252.h> #device ADC=8 #fuses HS,NOWDT,NOPROTECT,NOLVP #use delay(clock=4000000) // #use rs232(baud=1200, xmit=PIN_C6, rcv=PIN_C7,errors) // #define MICROSOFT TRUE #define MOUSE_SYSTEMS FALSE #DEFINE LEFT_BUTTON PIN_B6 #DEFINE RIGHT_BUTTON PIN_B7 #DEFINE X_CHANNEL 0 #DEFINE Y_CHANNEL 1 #define PS2_CLK_OUT PIN_B4 // #define PS2_DATA_OUT PIN_B5 // #define PS2_CLK_IN PIN_B1 // #define PS2_DATA_IN PIN_B0 // #define LED PIN_C3 void ps2_out(BYTE data) { BYTE i; output_low(PS2_CLK_OUT); // start bit on for(i=1;i<=8;++i) { output_bit(PS2_DATA_OUT, shift_left(data,1,0)); output_high(PS2_CLK_OUT); // <= I set break point here output_low(PS2_CLK_OUT); } output_low(PS2_DATA_OUT); delay_ms(11); }

Hello, Not sure if I can help much but it looks like shift_left should be output_bit(PS2_DATA_OUT, shift_left(&data,1,0)); data needs to be a structure or array shift needs the address of data &data Not sure about the port. Single step the program and watch for changes? I'm just starting with CCS too. If you post some more code maybe I can help. If you need SPI you should use the built in stuff very easy + lots of examples. If your rewriting PICC\Examples\EX_MOUSE.C maybe you should get that working first. http://www.ccsinfo.com/forum/ is a good place to hang out. A very good book is PICmicro MCU C by Nigel Gardner ISBN 0-9724181-0-5 Hope this helps. Pete {Original Message removed}

Pete, thaks for your input yes that was the problem with shifting. what is the purpose of & ? Andre peteHVAC wrote: {Quote hidden}>Hello, >Not sure if I can help much but it looks like shift_left should be >output_bit(PS2_DATA_OUT, shift_left(&data,1,0)); >data needs to be a structure or array shift needs the address of data &data >Not sure about the port. Single step the program and watch for changes? > >I'm just starting with CCS too. If you post some more code maybe I can help. >If you need SPI you should use the built in stuff very easy + lots of >examples. > >If your rewriting PICC\Examples\EX_MOUSE.C maybe you should get that working >first. > >http://www.ccsinfo.com/forum/ is a good place to hang out. > >A very good book is PICmicro MCU C by Nigel Gardner ISBN 0-9724181-0-5 >Hope this helps. > >Pete > >{Original Message removed}

On Nov 28, 2005, at 4:42 PM, Andre Abelian wrote: > output_bit(PS2_DATA_OUT, shift_left(data,1,0)); > I'm not familiar with the specifics of CCS, but... What is shift_left supposed to do, beyond "data = data<<1", and are you sure it is appropriate here? And does output_bit() output the low order bit, high order bit, or boolean of its second argument? (unless it's the high order bit, which seems unlikely (not C-like), then this isn't going to do what it probably needs to do for serial-like communications...) Finally, though I guess you're not there yet, it seems to me that this code is somewhat likely to execute too fast for reliable communications with most PC-like PS2 devices/computers. You'll probably need a delay before toggling the clock bit... BillW

Andre, & means "The address of" It gives you the memory location of the variable data. When you make portb all low you are writing 0b00000000 to address F81h(portb on 18f452) If you declare a pointer variable int *a; and a variable int b=3; and you then say a=&b; a will contain the address of b not the value of b. Pete {Original Message removed}

William, thanks for your replay William Chops Westfield wrote: {Quote hidden}> On Nov 28, 2005, at 4:42 PM, Andre Abelian wrote: > >> output_bit(PS2_DATA_OUT, shift_left(data,1,0)); >> > > I'm not familiar with the specifics of CCS, but... > > What is shift_left supposed to do, beyond "data = data<<1", and > are you sure it is appropriate here? it will shift bits out thru one of pic pin if you do data=data<<1 then you are shifting to left only then you need to add more code to shift out thru the pin. > > > And does output_bit() output the low order bit, high order bit, or > boolean of its second argument? output_bit() // will specify single bit only high or low > (unless it's the high order bit, > which seems unlikely (not C-like), then this isn't going to do what > it probably needs to do for serial-like communications...) > > Finally, though I guess you're not there yet, it seems to me that > this code is somewhat likely to execute too fast for reliable > communications with most PC-like PS2 devices/computers. You'll > probably need a delay before toggling the clock bit... > my spi routine is messed up. Since I am not there yet deffinetly needs more work to be done. thank you for your time > BillW Andre

Andre Abelian wrote: > when I set break point on portB bit 4 I wasn't aware any tool let you set a breakpoint on a particular bit, whatever that means. Do you mean write to the bit? Read from it? Any access? Port pin externally changed? ****************************************************************** Embed Inc, Littleton Massachusetts, (978) 742-9014. #1 PIC consultant in 2004 program year. http://www.embedinc.com/products

Olin Lathrop wrote: All I mean is when I am setting break point on this line instead of just setting bit 4 it sets other bits too. that was the problem. I do not think there is a way to set a break point on single bit. Andre {Quote hidden}> Andre Abelian wrote: > >> when I set break point on portB bit 4 > > > I wasn't aware any tool let you set a breakpoint on a particular bit, > whatever that means. Do you mean write to the bit? Read from it? Any > access? Port pin externally changed? > > > ****************************************************************** > Embed Inc, Littleton Massachusetts, (978) 742-9014. #1 PIC > consultant in 2004 program year. http://www.embedinc.com/products

On Tue, 2005-11-29 at 07:17 -0500, Olin Lathrop wrote: > Andre Abelian wrote: > > when I set break point on portB bit 4 > > I wasn't aware any tool let you set a breakpoint on a particular bit, > whatever that means. Hi Olin, I'm not sure to which tool Andre refers, but gpsim (and probably other 3rd party simulators) can break "on a particular bit". > Do you mean write to the bit? Read from it? Any > access? Port pin externally changed? gpsim can do all of these. For example, Break if RB4 if a '1' is written to portb: break w portb & (1<<4) == (1<<4) Break if RB4 is read as a zero: break r portb & (1<<4) == 0 Or if you want an assertion type break that checks portb while the code is running at a particular address: break e loop_start ( (portb & (1<<4)) == (1<<4)) This last one sets an execution break point at the label 'loop_start' and will halt whenever the simulation reaches this address and RB4 is high. Scott

At 10:31 PM 11/28/2005 -0500, you wrote: I'm not that familiar with CCS C (except to know that it is not very much like the C language from doing a translation), but you should also consider inserting a delay between where you set PS2_CLK_OUT high and set it low or the strobe duration may not be long enough to work, depending on the efficiency of code the compiler emits. Worst (or best) case it could emit two succeeding bsf/bcf instructions that might not even reach the correct logic levels with a 20MHz clock, let alone be long enough to meet the requirements of whatever is connected to your interface. Data setup and hold times (probably more the former) could also be an issue, so look at that too. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" spam_OUTspeffTakeThisOuTinterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com ->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff