Possible issue with the legs. Just received Bittle this week, pre-assembled. During initial testing, Bittle seems to fall over and act a little crazy during walk and runs.
I did some physical calibrating per the instructions, and finally did some more precise ones via the IDE software yesterday. Everything looks aligned and at right angles in Calibration Mode.
Turn it on and it's crouched. And when I press RUN, it's more of a walk with the left legs not using a full range of motion.
When I press Calibration from the remote, and then press RUN, it runs very fast when I hold it up and the legs seem to be using their full range. But when I place it on a surface, it struggles and flips over.
Not sure what to do with it. Seems like it wants to do everything, but something's off. Here are a few videos to see what anyone thinks. Thanks in advance!
Thanks Knut. We'll do some playing and keep you posted. Seems like this should be pretty straight forward - but there's always a nuance it appears. Thanks again...
Re: RPI-zero, the idea is that gaits/skills (like running) are under the bittle board control, and yes instead of receiving high-level commands like "run forward" from the IR, the board gets them from the RPI, over the serial port.
I'm playing with something similar - take a look at this post, maybe you can help too: https://www.petoi.com/forum/software/driving-bittle-from-raspberry-pi
It works but not reliably. You can drop/run that code in the serialMaster folder.
Two Questions:
First one: The OpenCat upload seems to possibly have a glitch in the Run command. When I press it on the remote, the following video is what occurs:
Everything else seems ok. But this clearly isn't "Running" anymore. The servos ran much quicker in Run mode previously. But this is what happens now with the current firmware. Any ideas what's happening or how to fix it?
Second question - on adding a Raspberry Pi Zero-W:
If we install a RPI Zero-W using the serial connector; do the commands, (Run, Forward, Sit, etc.), remain under control of the Nyboard?
Could the RF/IR commands can now be delivered via the serial interface connection between the Pi and the Nyboard?
The intent is to allow the PI to replace the command set that is currently being delivered via RF/IR, and tell Bittle to move from a smartphone app via the Pi instead. Is that possible?
Thanks as always!...
Thanks for the response and direction on correcting things. I reinstalled an entirely new version of Arduino IDE, added the libraries carefully, and followed all your other instructions. Things finally are back up and running and seem to be ok.
The battery got low, so I haven't run it through all the functions yet until it's recharged. I also wasn't turning the Gyro back on after calibrating, so that should correct a lot of it - great tip!
I'll keep you posted, but I think we're on the right path.
Now that I have a better understanding of uploading the OpenCat, I'll keep up on the latest versions for continued stability too.
Appreciate your prompt communication and support!
I'm going to be adding a Pi Zero-W with some functions later. If I run into any issues, you may hear from me again. :-) :-) Thanks again...
Hi, the first two videos show that the dog is turning. One side of the legs is moving faster at a larger amplitude. You can press the forward button (CH on the remote) to make it walk straight.
There's a switch (CH+) for the gyro, which detects the body angle and calculates adjustments. By turning it off, the movement will be faster. When you enter the calibration mode, the gyro will be turned off automatically to avoid disturbance when you move the robot. So after you calibrate, you need to press the switch to turn on the gyro.
Please get the newest OpenCat code on GitHub to improve the stability of walking gaits.