Swordfish

I have posted a new user module to the Wiki. It provides read/write capability for the Melexis MLX90614 family of non-contact IR sensors. More on the sensor at http://www.melexis.com/ProdMain.aspx?nID=615.

I'm a self taught hobbyist programmer just learning SF so I welcome any comments on module improvements.

Tom

Edited to correct product link.

Hi
I've been trying to get some response from your module - no joy.
The pic (18F2620 @20M) seems to lock-up.
If I disable the I2C call, and just send out the USart.writes all is ok - obviously no values though.

I have the device connected as follows

Pin1 SCL to Pic C.3
Pin2 SDA to Pic C.4
Pin3 Vdd
Pin4 Vss

I have made no changes to you example code, except added Portb.0 to toggle its state, after every sample. When I disable the I2C call this addition functions as expected.

Any ideas ?

Do you have 4.7k pullup resistors on the clock and data lines?

Hi
Thanks for getting back to me, yes I do have pull-ups on the scl/sda lines.

look forward to hearing from you.

cheers

After what seems to be a long time testing the code, I'm getting no where using the MLX90614 from Sparkfun.

Again have not changed anything with Tom's orginal code.

Getting desperate !

Phil

Sorry but I'm away from my bench for several weeks but the code continues to work in an application I have running.

The only thing that comes to mind is that the MLX90614 is a 3.2V device and if you have applied 5V to it, it may be fried.

The pullups (I used 22k) must be tied to the 3.2V supply as well. The PIC will interpret the logic levels correctly when set up this way. Actually I used the LF version of the PIC so I could apply 3.2V to it as well - needing only one power supply.

Otherwise, there may be a silicon issue for the I2C hardware in the PIC. You might try using the software I2C or try getting I2C working with another less complex device.

Any help?

Tom