Ok, so I’m not using the Raspberry Pi with the GoPiGo. I’m attempting to use the nVidia Jetson. So far I have not been able to get the two to talk over i2C.
From the Jetson to the GoPiGo I have connected wires between GND (Pin 6), 3.3 V (Pin 1), SCL (Pin 5), and SDA (Pin 3). I have not connected any other lines between the two boards.
I power the GoPiGo off of the battery pack and the Jetson off of a wall plug, but will move to a battery pack.
Running i2cdetect on the Jetson or using the SMBus python library I do not get a response from the GoPiGo. Unfortunately, I do not have any other i2c slaves other than the GoPiGo to test with, so I can’t say if my problem is on the Jetson or my connections between the two devices.
Are there additional lines that the GoPiGo needs? Perhaps some line is being used to tell the board that a Pi is connected or something?
Hey, not familiar with the nVidia Jetson. Googling showed me this: http://elinux.org/Jetson_TK1 Is that what you’re using?
So, I would try to connect a 5V line, not the 3.3V line. On the GoPiGo, the 3.3v pin doesn’t go anywhere. For now, until you want to upload firmware, you don’t need to connect the SPI lines.
Have you tried reversing the SCL/SDA lines, sometimes they’re easy to reverse.
Connecting the 3.3V pin does actually seem to make a difference in the voltage that I’m seeing on the SCL and SDA lines on the GoPiGo when I’m using the Jetson.
I have been able to talk to the board using the Raspberry Pi just by running 3 wires between the two, so I believe the problem is on the Jetson side.
Sorry we don’t have more answers, just sort of shooting in the dark here with the new system. This would be a really cool project to show off, when you get it working.
Do you have an o-scope or a multimeter? Is the data line running at 3v3 or 5v? I love using the Salae Logic analyzer, it lets me check that data is going through a bus like this pretty easily, helps troubleshoot problems. The new ones have analog analysis as well.
I have both an o-scope and a multimeter. My o-scope skills are rusty so I haven’t figured out how to make good use of it. Agilent has kindly informed me that to use the i2c analyzer for my scope I need to pay them an additional $1500
The Logic analyzer looks good and reasonably priced.
I have several i2c buses to choose from with the Jetson, some run at 1.8 V and some at 3.3 V. I’ve tried both (with a level converter when appropriate), but have had no luck.
I’m wondering if Dexter can make the circuit diagram for the GoPiGo available?
Yes, I’ve looked at that diagram, but it only contains a single image showing the various connections. It would nice to have a more detailed diagram. Maybe the PCB layout? Or at least a diagram that shows all the connections on the board.
For example, you said that the 3.3 V pin on the connector doesn’t go anywhere, yet connecting that pin does in fact make a difference to the voltage levels I read on the SCL and SDA lines so clearly it must be connected somewhere. But from the Eagle file I can’t tell what happens with the 3.3 V line.
The signals that you need to make the GoPiGo work are: 3.3V(for the level convertor chip),GND,SDA and SCL (comms). The GoPiGo can feed 5V to another board with the 5V pin. The SPI lines (SCK,MOSI,MISO and CE0) are also connected for uploading the firmware but the avrdude software used to do that is custom written for the Raspberry Pi so I am not sure if that would even work. The RX and TX lines are also connected to the Grove port on the GoPiGo with the level convertor but this wont be 'very important.
Also, it would be very difficult to work on this without the logic analyzer and I think it might be better to use https://www.saleae.com/ instead of paying $1500 for the licence for the Agilent scope. These analyzers are just great and you can even use the cheapest one for this.
