[Bug] Sudden Prime Hub Shutdown in Beta

[zoligyenge]

Hello, recently we expereinced repeated sudden shutdowns of two different PRime HUbs when using Pybricks Beta (in the block coding interface - if it matters).

Interestingly teh same program did never causes a shutdown in Pybcricks Code, despite the hub still having the beta firmware (BTW: Is using teh stable interface but the beta firmware recommended or rather not?)

Any help much appreciated.

[laurensvalk]

Thanks for reporting! Did the hub shutdown fully and stay off or restart? Did the usual rotating animation get stuck for a while before it did that or is it sudden?

Did you see any patterns that made this occur? Was there a particular program that causes it? For example always during driving or something else entirely?

It's fine to use the beta firmware with the older interface.

To rule out hardware issues: some people have reported that their battery is a bit loose, moving around in the compartment. You could see it that might be the case. If so, pushing a piece of paper or cardboard behind it to keep it in place can help.

[zoligyenge]

Thanks for the quick reply. There is a fully shutdown (not a restart) with the screen, start button and bluetooth light all off.

Could not find a pattern so far. I don't think it is a hw issue since we switched same program, same robot to Pybricks stable (only interface, not the FW) and it works fine.

BTW: Is it expected behavior to have a small rectangle on the left bottom corner of the Hubs screen (in standby) with the beta?

[laurensvalk]

If it turned off all the way, my bet would be on the loose battery. Does it happen if you push or shake the battery manually?

But please do keep an eye on it if you happen to see it due to a certain combination of software actions :)

Switching to the other interface is likely a coincidence, since these are identical at this point.

The pixel at the bottom is the slot indicator. You can save and run multiple programs now!

[zoligyenge]

Hi. We did some more testing:

1. Loose Battery we excluded
2. It is dependent on the firmware not the programming environment 
3. We confirmed that 100% same code runs perfectly with stable FW and causes sudden full shutdown with Beta.
4. My suspicion is that the drive with stop-hold in the multitask starts, but never gets to hold (see video)
   Here is the respective task:

[laurensvalk]

Thanks for investigating further!

Could you share the full program, or even better, an independent small program with the same issue?

[zoligyenge]

This code 1-1 (re)produced the same sudden comlete shutdown. BTW. after restart, the hub did not start the progarm - apprantly it wa snot saved (as there was no proper shutdown).

from pybricks.hubs import PrimeHub
from pybricks.parameters import Axis, Direction, Port, Stop
from pybricks.pupdevices import ColorSensor, Motor
from pybricks.robotics import DriveBase
from pybricks.tools import StopWatch, multitask, run_task, wait


# Set up.
prime_hub = PrimeHub(top_side=Axis.Z, front_side=Axis.X)
motorRight = Motor(Port.B, Direction.CLOCKWISE)
motorLeft = Motor(Port.A, Direction.COUNTERCLOCKWISE)
drive_base = DriveBase(motorLeft, motorRight, 62.4, 160)
Color_Sensor_Bottom = ColorSensor(Port.E)
watch = StopWatch()
reflection_black = 35
reflection_white = 70
reflection_dark_white = 70
drive_acceleration_slow = 1100
drive_acceleration_fast = 1000
drive_speed_extrem_fast_ = 1000
drive_speed_fast = 500
drive_speed_slow = 150
turn_acceleration_fast = 1200
turn_acceleration_slow = 500
turn_speed_fast = 750
turn_speed_slow = 550


async def subtask(Distance):
    await drive_base.straight(min(Distance, 100) + 100)
    await drive_base.straight(-100)


async def subtask2(Distance):
    # warten auf schwarz
    if Distance != 0:
        while not await Color_Sensor_Bottom.reflection() > reflection_white:
            await wait(0)
    while not await Color_Sensor_Bottom.reflection() < reflection_black:
        await wait(0)
    drive_base.straight(0)


async def _go_to_black(Distance):
    await wait(0)
    # vorwärts
    await ___drive_FAST___()
    await drive_base.straight(max(Distance - 100, 0), then=Stop.NONE)
    await ___drive_slow___()
    await multitask(
        subtask(Distance),
        subtask2(Distance),
        race=True,
    )
    await ___drive_FAST___()


async def ___drive_FAST___():
    await wait(0)
    drive_base.settings(straight_speed=drive_speed_fast)
    drive_base.settings(straight_acceleration=[drive_acceleration_fast, 1000])
    drive_base.settings(turn_rate=turn_speed_fast)
    drive_base.settings(turn_acceleration=[turn_acceleration_fast, 1000])


async def ___drive_slow___():
    await wait(0)
    drive_base.settings(straight_speed=drive_speed_slow)
    drive_base.settings(straight_acceleration=drive_acceleration_slow)
    drive_base.settings(turn_rate=turn_speed_slow)
    drive_base.settings(turn_acceleration=turn_acceleration_slow)


async def main():
    await ___drive_FAST___()
    # Initialize NIE löschen!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!


    await _go_to_black(500)
run_task(main())

[zoligyenge]

Here is the Py file

debug.py

[dlech]

Probably not related, but drive_base.straight(0) is missing await in subtask2().

[zoligyenge]

The code is generated from this block code:

[zoligyenge]

Probably not related, but drive_base.straight(0) is missing await in subtask2().

Interesting comment. So going by this I should never use the drivebalse-stop(hold) wordblock but instead use a straight(0)?

[laurensvalk]

Thanks - you've found a bug. We'll fix this. You should be able to use any block at any time, block mode is supposed to be simple 😄

(This should not cause the shutdown by the way, but it's good to fix nonetheless.)

[zoligyenge]

Thanks - you've found a bug. We'll fix this. You should be able to use any block at any time, block mode is supposed to be simple 😄

(This should not cause the shutdown by the way, but it's good to fix nonetheless.)

Could the shutdown be due to two commands accessing the same resource (drivebase) in a multitask? Even though strange that it worked before.

[laurensvalk]

I'll give this a try next week when I'm near a robot.

So I should just build a drive base with one color sensor and run that one python file? And drive up to a black line?

Does the issue always trigger, or only sometimes? Just to prepare my test case.

Thanks!