Unfortunately I was not able to get the single_finger_test running. I just want to provide a quick overview of what I have done to see if there are any obvious issues that you see. This is a long post so please feel free to take some time to respond!!
The current situation:
I have a fully built single finger:
(Obviously, there’s a bit of wire management that I need to do, but I’ve been moving everything around trying to get it all working)
The “upper leg” module is connected to the bottom microcontroller (the microcontroller with only 1 Booster Pack). The “hip fe” module is connected to Booster Pack A on the top microcontroller, and the “hip aa” module is connected to Booster Pack B on the top microcontroller. On both microcontrollers, only jumpers JP3, JP6, and JP7 are installed. Both microcontrollers are set to the “ON-ON-OFF” position. Both microcontrollers have been successfully flashed with the provided software using the instructions within the Github page. Both motors are connected to a nearby lab power supply supplying around 3.3V at around .3 A:
When turning the power supply on, the green D1 LEDs are illuminated on both boards:
Playing with the amperage supplied can also turn on the blue D10 LEDs as well:
If I set the voltage and/or amperage too high, the red D9 LEDs will illuminate:
I don’t know what any of these mean, but I generally assumed that the green was the best and left it in that configuration.
On all three Booster Packs, the red “nFault” LED, and the green “D3 3.3V” LED are illuminated:
I assumed the “nFault” LED was not supposed to be illuminated, but the official documentation says that “A fault may appear [after turning on your power supply] on the nFAULT LED. This is normal and should be cleared when the status registers are read or EN_GATE is taken HIGH.”, so I didn’t look too far into it.
Both microcontrollers are connected via CAN to a temporary computer that I built for this project:
The components on this computer are very bad and around 10+ years old (core i3-2100, 8gb ddr3, no gpu, etc.) but I assume this does not matter. On the SSD is Ubuntu 20.04. The CAN ports have been initialized by running the “initialize_can_bus.sh” script provided in another Discourse thread, which allows me to see that they are being recognized by the computer.
With your help, on the computer, I’ve been able to fully get the Singularity image working and the demo to run as seen here:
You can see the warning “Warning this thread is not going to be real time.” which I assume is a warning notifying me that I have no yet configured my Ubuntu install to be an RT_PREEMPT installation – I attempted doing this but had some issues which led me to stop trying for the time being. Seeing as the demo worked, I assumed this meant that it would be possible to run the single_finger_test, just without real-time results.
Next, I try to run the single_finger_test and see the following:
One thing to note here is that there was a code update on Mar 1 which set the value of “homing_with_index” to “false” which is not compatible with my codebase – it threw some error that I needed to use “homing_method” instead, which I set to “none” and it seemed to work.
After this, I get the message “waiting for board and motors to be ready” as seen in the screenshot. I’ve tried various things that I don’t really know if they matter such as: changing the CAN ports for each microcontroller, swapping the microcontrollers out for different ones (I have 6 fully prepared ones), connecting the different modules to different power sources (not sure if there is a required order for this, but I assume not), and various other software things (restarting my computer, etc.) and I just can’t seem to get it working…
I’m not sure what I’m doing wrong, and what should do next but I would love your input.
Once again, thank you very much for designing this awesome robot and helping me build it!!