6 replies to this topic

[Nevyn]

I'm just trying to get to grips with I2C and I've either got a problem with code or interpreting teh logic analyser output. My code looks like this:

device = new I2CDevice(new I2CDevice.Configuration(0x48, 400));
byte[] command = new byte[] { 1, 0 };
I2CDevice.I2CWriteTransaction[] transaction = new I2CDevice.I2CWriteTransaction[] { I2CDevice.CreateWriteTransaction(command) };
int result = device.Execute(transaction, 1000);

So on the LA I'm expecting to see
Start command
Device code with the lower bit set to 0 (for write) followed by ACK
Two bytes 1 and 0 (each followed by an ACK)
Stop command

What I'm actually seeing is
Start command
Device code and lower bit set correctly followed by NAK
Stop command

The LA is Saleae and I've set the system to process I2C data.

Have I missed something in the code as it appears that the 1 and the 0 bytes are not being output to the I2C device

Thanks in advance,
Mark

Edit: Aditional Information
The chip is a lone chip on the bus and has the specified pull-up resistors.
According to the spec the data transfers do not require repeated start bits.

Edited by Nevyn, 29 April 2011 - 09:03 PM.

[Chris Walker]

Hi Nevyn, When you say that you're getting a NAK, is your device actively refusing the message? If there's no ACK from the device, then data won't be sent... What device are you trying to talk to? Does it have a datasheet? Chris

[Nevyn]

When you say that you're getting a NAK, is your device actively refusing the message?

That's the conclusion I'm coming to. I've modified the code and config back to the default address of 0x64 and I'm still getting the same results.

What device are you trying to talk to? Does it have a datasheet?

Device details can be found here. The Functional Description document talks about the I2C interface on pages 8 & 9. I'm wondering of the debug monitor (discussed on page 10) is on and so the chip is refusing a connection - however I've got the serial input pin held low so this should be disabled.

Thanks for any help,
Mark

[Michel Trahan]

Would that solve our Math library problem ? LOL

[Nevyn]

Would that solve our Math library problem ? LOL

Maybe if I could get the little begger to work - an old solution to an old problem. Off to get a serial cable and try and talk to the debug monitor as even SPI is playing up now (well more than it was a day or two ago). Public holidays here so it'll be Wednesday before I can get the parts - drat.

Regards,
Mark

[Nevyn]

When you say that you're getting a NAK, is your device actively refusing the message?

If there's no ACK from the device, then data won't be sent...

Finally got the FTDI cable and have manged to talk to this little beastie. I have also found a horrible little comment on the chip and it relates to the 3.3V power supply. Apparently, the chip should be clocked at 15MHz if using 3.3V power and not the full 30MHz (the default). Using the debug mode I reconfigured it and I managed to get the thing synchronising with the N+.

Next step, add 1 to 1 and hopefully get the right answer.

Thanks,
Mark

[Nevyn]

Next step, add 1 to 1 and hopefully get the right answer.

YES: 1 + 1 = 2 and even better, 1 + 1 + 1 = 3 !

Now on to hard sums

Mark