Mcp 23017 and Pic 16f877

[CLAUDY]

Hello everyone,
I'm new to this forum and I would use a MCP23017 with a 16F877.
Not knowing the MCP23017, I initially ordered the release of the MCP23017 in PBP.
Could you please give me an example of code that could unlock and so I understand.

In the wait for your answers, thank you in advance.

Claudy

[Dave]

CLAUDY , Is the MCP23017 not some sort of I/O expander from MicroChip? You do not need PBP to use it. It is not a processor but an mearly a periferal to allow more I/O pins to be added to an exsisting processor. If the processor you are using is indeed from MicroChip then you can use some of the I2C or Shiftin/Shiftout commands to access it's registers provided if the PBP language.

Dave Purola,
N8NTA

[CLAUDY]

Hello DAVE,
Yes, I know that the MC23017 is not a processor but I forgot to say that I wish I could control the outputs from the 16F877 I2C. It would be very nice if I could have an example in PBR so I can start with that component which is new for me.

A big thank you for your reply

Claudy

[Darrel Taylor]

Here's an example from DaveC3.
Just ignore the interrupt and USB portions.

The constants and I2CREAD/I2CWRITE statements are what you need.

http://www.picbasic.co.uk/forum/show...5720#post55720

[CLAUDY]

Hi DT,

Thank you for taking the time to answer me.
I'll try this code as soon as Monday this weekend I'm in France.

Again thank you and good weekend to all.

Claudy

[CLAUDY]

Hello,

OK I managed to contact the MCP23017en I2C.
For now, the test program the LED lights of the 2 ports following a counter made ​​by the software in the 16F877.
I have a small question.
When feeds it off, the LED of port B are all lit and the rest for + / - 30sec (while the PIC is running without problem) and then everything works normally.
If I reflash the PIC everything is OK, the problem happens only when switched on.
As if it would take some time for it to MCP23017 syncronise with I2C.

Here is the code used

adcon1=7 'pour que les entrées analogique du port A soient digitale

trisa=0 'pour que RA3 soit une entrée et les autres des sorties
trisb=0 'pour que tout le port B soit des entrées
trisc=0 'pour que tout le port C soit des entrées

'initialisation de LCD

'debut code specifique 877


DEFINE LCD_DREG PORTD 'LCD data port
DEFINE LCD_DBIT 4 'LCD data starting bit 0 or 4
DEFINE LCD_RSREG PORTD 'LCD register select port
DEFINE LCD_RSBIT 0 'LCD register select bit
DEFINE LCD_EREG PORTD 'LCD enable port
DEFINE LCD_EBIT 1 'LCD enable bit
DEFINE LCD_BITS 4 'LCD bus size 4 or 8
DEFINE LCD_LINES 2 'Number lines on LCD '
DEFINE LCD_COMMANDUS 2000 'Command delay time in us
DEFINE LCD_DATAUS 50 'Data delay time in us




'***definition des variable pour I2C***
DEFINE I2C_HOLD 1
DEFINE I2C_SCLOUT 1
define I2C_SDA PORTa.0
DEFINE I2C_SCL PORTa.1
Define I2C_SLOW 1 'At this clock speed this needed to be added to make it work.


sda var porta.0 'sda est sur le porta.0
scl var porta.1 'scl est sur le porta.1


Addr var byte
j var byte
i var byte

'================================================= =====================================

'************* DEFINITION DU PORT A ET DU PORT B ***************
' Port A Registers (8 Bit Mode)
IODIRA con $00 ' Port A Direction Bit Control Reg (suivant datasheet)
IOLATA con $0A ' Output Latch Reg (suivant datasheet)
'Port B Registers (8 Bit Mode)
IODIRB con $10 ' Port B Direction Bit Control AReg (suivant datasheet)
IOLATB con $1A ' Output Latch Reg (suivant datasheet)
debut:
lcdout $FE,1, "TEST MCP23017 " ;affichage ecran acceuil
lcdout $FE,$C0, "V 1.3" ;affichage ecran acceuil
pause 1000 ;pause pour ecran acceuil

i2cwrite sda,scl,$40,[IODIRA,0]


i2cwrite sda,scl,$40,[IODIRb,0]
pause 100
'i2cwrite sda,scl,$40,[Ipola,255]
lcdout $FE,1, "Debut"
for i=1 to 255

i2cwrite sda,scl,$40,[iolata,i]
i2cwrite sda,scl,$40,[iolatb,i]
lcdout $FE,$C0,"i=",#i
pause 100
next i
lcdout $FE,1, "Fin"
goto debut




end


Thank you for your help

Claudy

[Demon]

Would a pull-down resistor at each LED solve the problem?

[CLAUDY]

In fact the Pcb is done simply for the moment to get used to the MCP23017 that will serve me to something else after.

The PCB is composed of 8 LED simply placed on Port A and 8 LED placed on port B. The communication with the Pic I2C is done.
Pic also controls an LCD 2 * 16.

The test program is to enable a counter from 1 to 255 that will be displayed on the LCD (i = 1 until i = 255) and the 2 * 8LED (port A and port B).

At the first power, while port A is activated (LED lit) and port B is disabled (LED off). When the counter reaches + / - 130 (viewed from the LCD), Port A (which was fully activated) starts from the port B but it does not change state. And only when the counter recomm early as 2 port (A and B) fonctionnnent normally (they have at the same time as the LCD).
And it works very well.
The problem is powering up.

The problem must come from the Start Bit of the I2C.
I already tried some codes for the Start bit but I can not.

Aurra you an idea?


Again thank you for your reply

Claudy

[Demon]

In fact the Pcb is done simply for the moment to get used to the MCP23017 that will serve me to something else after.
...

Are you saying the MCP23017 is NOT used at the moment? Why issue an I2C command to something inactive? Do you have LEDs connected on the MCP23017?

The PCB is composed of 8 LED simply placed on Port A and 8 LED placed on port B.
...

sda var porta.0 'sda est sur le porta.0
scl var porta.1 'scl est sur le porta.1

So you have 16 LEDs on ports A and B, but you also use Port A for I2C communication?

Why not use Port C.3 for SCL and Port C.4 for SDA (PIC 16F877 datasheet pinout p.3) to avoid interfering with your LEDs?

[mister_e]

Plus ça change, plus c'est pareil...

KeyWord: Datasheet

Not ALL PORTA pins are true output... and on PORTB there's this PGM pin who could do nasty thingy if your CONFIG FUSES aren't set properly...

About them (Config fuses) Where are they?!?

Are we talking about a 877 or 877A ?

And finnally... you want to manually assign value to your PORT BEFORE doing anything else... say BEFORE the PAUSE 1000. Usually

PORTA=0
PORTB=0
...
TRISA=0
TRISB=0
....
Variable and all other code here.

Bonne chance

[mister_e]

sda var porta.0 'sda est sur le porta.0
scl var porta.1 'scl est sur le porta.1

So you have 16 LEDs on ports A and B, but you also use Port A for I2C communication?

Why not use Port C.3 for SCL and Port C.4 for SDA (PIC 16F877 datasheet pinout p.3) to avoid interfering with your LEDs?

Actually If I understand properly, it is worst than this. If the LED are connected to Ground on the I2C line, they Screw-up the line as they have to be Pulled-high (hi-z) but the LED LOAD it to gnd.

Get rid of the LED on those I/O or, like BobyBert said, move the I2C line elsewhere (it doesn<t need to be on a specific SCL, SDA PIC line as you're using I2CREAD/I2CWRITE who's not using the built in SSP module). You want the I2C line on their own I/O.

[CLAUDY]

Hello,

Thank you for your answers.

When I speak of the port A and B have ports MCP23017 and not PIC.
The PIC used is a 877A.
The PIC controls only the LCD and I use the port A of the PIC to bond with the I2C MCP23017.

A big thank you for your help.

Soon

Claudy

[Acetronics2]

Hi, Claudy

Following your adventures on many forums for a moment ... I still miss a small thing : YOUR USED SCHEME ...

Alain

[CLAUDY]

Hello Alain,

I still have not found the solution.

Here's a quick diagram as requested.Printing TEST MCP23017-sch.pdf

Thank you

Claudy

[Acetronics2]

Looks there's also a missing register named IOCON in the settings.

Alain

[CLAUDY]

Hello everyone,

Ok it works!

Bit 7 of Port A and B of the MCP23017 must be 1 before the startup of the program.

And in this case each power program is operating normally the first

Here is the code added.

i2cwrite sda,scl,$40,[iolata,128] 'Output Latch Reg
i2cwrite sda,scl,$40,[iolatb,128] 'Output Latch Reg


If you have comments please.

Thank you

Claudy

[rauloj]

Bit 7 of Port A and B of the MCP23017 must be 1 before the startup of the program.

And in this case each power program is operating normally the first

Here is the code added.

i2cwrite sda,scl,$40,[iolata,128] 'Output Latch Reg
i2cwrite sda,scl,$40,[iolatb,128] 'Output Latch Reg


If you have comments please.

Thank you

Claudy

Hi Claudy!

It's not OLATA/OLATB you are setting 1, but IOCON.
After power-on BANK bit = 0 and IOCON's adr = 0A/0B not OLATA/OLATB's.
After setting IOCON bit 7 to 1 you can use address mapping as described in datasheet table 1-5.

R