Splash boards 2

In the next episode of the battle of the splash boards

https://www.rcgroups.com/forums/show...-waterproofing

The mane problem was they disintegrated in backflips. Step 2 was the most extensive use of TPU to PLA welding so far. They were adhered, but to be adhered, they needed to be printed on buildtak. The dirt wasn't going to easily wash off anymore.

Stepped the thickness from .4 to .64mm. It was definitely heavier. The race is on to get rid of the last of the clear TPU. It proved quite useless since it turns orange & the paw controllers all ended up having to charge without their protective socks. The final use was the tail lights. Those are going to get redesigned to use scotch.

The final reckoning is clear PLA. It's too brittle for any structural applications. Its only use is seeing the LEDs in the controllers. White PLA might do the job.
------------------------------------------------------------------------------------
It desperately needs heading hold in reverse leash mode. The steering code is such a rat's nest, there will be many malfunctions in the field.

A minimal change to get it to work was quite effective. The trick with the minimal change is the heading doesn't time out after 10 seconds of being stationary like RC driving. The heading only resets when the leash is driven forward. It's not clear why this is. target_heading might not be getting reset or steering_timeout might not be getting decremented. It is in brake mode when it's not moving & the leash is pointing backwards so steering_timeout isn't being reloaded.

The reverse speed can't be increased above minimum because it starts oscillating even at minimum.

---------------------------------------------------------------------------------------------------
Leash control is 1 of those oddball things that's really hard & no-one else is doing, like digital video editing in 1997. Still pondering some kind of toy that uses a 360 deg angle sensor. Perhaps a photovore could follow a leash in super slow motion or a walking robot would be more anthropomorphic. None of the robot dogs have leashes. A 360 sensor requires limiting the payload height.

The most practical need now is a minimal marker for camera work. It's not possible to drive more than a few ft in front of the lion. It desperately needs to tow a marker or project some kind of visual marker. It could be an unpowered 3 wheeler with just a leash & steering servo. It could be a fixed laser pointer.