I2C Slave problem

rightbak · Joined: 16 Sep 2011 Posts: 10

I'm using a 16F15376 with compiler version v5.070.

I'm having problems setting up the device to configure the I2C module as a slave. The problem I am seeing is that it works correctly if I use MSSP1 but fails to work if I use MSSP2.

So for example, this code works as an I2C Slave using I2C1

temtronic · Posted: Tue Oct 17, 2017 6:48 am

There are several reserved addresses and pretty sure 0x10 is one of them..though that may not be a concern, yet....

Have you the correct pullups on the 2nd I2C bus ??

Have you tried PCM P's I2C scanner program from the code library ?

Might be a compiler bug, version please !

Jay

rightbak · Joined: 16 Sep 2011 Posts: 10

The physical I2C lines come in to RB1 (Clock) and RB2 (Data) and I'm using the pin select #use statement to redefine which MSSP module is 'connected'. MSSP1 appears to work but MSSP2 doesn't - see code above. The I2C pullups don't change so must be correct for MSSP1 to work.

Compiler version is v5.070.

I haven't seen PCM's I2C scanner program. I'll see if I can find it.... thanks

newguy · Joined: 24 Jun 2004 Posts: 1908

I haven't used that PIC but from the PICs I have used that have featured the PPS functionality, sometimes not all peripherals can be remapped to all pins. Have you checked that MSSP2 can actually be remapped to those two particular pins?

rightbak · Joined: 16 Sep 2011 Posts: 10

MSSP2's default location is RB1 and RB2 anyway so perhaps I don't need to re-map it at all? I did try without the #PIN_SELECT statements but it didn't compile..

jeremiah · Joined: 20 Jul 2010 Posts: 1349

Since I2C1 works on the same pins but I2C2 does not, I would start out by dumping the listing (LST) file for both versions (working and not working) and see how the assembly compares. Perhaps a wrong address is used for I2C2. The datasheet should have all the addresses listed so you can quickly tell.

Also did you check the errata on the chip to make sure there wasn't a problem with I2C2?

rightbak · Joined: 16 Sep 2011 Posts: 10

Problem resolved...

I still had the interrupt enabled for INT_SSP rather than INT_SSP2!

Once changed over, all working ok! Thanks for all your prompting...

Ttelmah · Joined: 11 Mar 2010 Posts: 19518

We have all done something similar at some point.

It is the little things that become vital....

alan · Joined: 12 Nov 2012 Posts: 357 Location: South Africa

Which once again highlights the request that a small compilable program with the fault present be attached.

Ttelmah · Joined: 11 Mar 2010 Posts: 19518

That is also why often the act of producing the 'cut down' version will make you realise what is wrong....