From: Adam M. <mcl...@gm...> - 2006-06-13 22:15:40
|
On 6/13/06, Dave Hylands <dhy...@gm...> wrote: > Hi Adam, > When you were testing, do you have the programmer connected? I began to think that might have been an issue, so no; in the last test I did to confirm that what I was saying was correct, I pulled the programmer out before testing the pins. I had been doing that in the past, though. > Other items to check (I don't recall whether these have been mentioned > already or not): > - Make sure you have a common ground between the gumstix and the ATMega8 They share a ground through the I2C connector. I just changed the setup such that they now share also share the power (they're now fed by the same regulator), but this doesn't change anything. > - Make sure that you have pullup resistors on the SDA and SCL lines > that pull them upto 3.3v. I'm using the level shifting circuit that Philips reccomends to go from 3.3v to 5v (application note AN97055). It's just two FETs and four pull-up resistors. I isolated that section of the circuit and made sure each side could pull the bus low. If I run i2c-io a few times on the gumstix, I can occasionally see the burst on the oscilloscope, so I'm pretty confident that the bus circuitry is fine. Adam |