1792 réponses à ce sujet

[Mike118]

You can just update ins and outs

[firened]

But then driving forward with only very little turning input will make it turn too fast.

[firened]

Easiest is to just split the servo cables and connect to GPIO instead of writing code to use the i2c registers.

However the speaker and mic would be the most interesting to test. I wasn't able to get it working yet. I wonder if it has a gpio conflict with the Vigibot client.

@Garfield2142 if you get it, if you want you could try to get the speaker working on an fresh Raspbian install without Vigibot.
You'll have to edit the sunfounder install script because it tries to do some things that fail, like installing picoTTS over apt....

[ralph_62550]

I couldn't find a file with a STEPY value inside. However I could make some adjustments in the hardware configuration with the "OUTS" values for "Left wheels" and "Right wheels".
But it is still a bit frustrating with those values for several reasons: frist you have to sacrifice the lowest possible speed and second there is even a dependency of the battery voltage. I had a running configuration. But after the voltage dropped a bit the problem reappeared, one of the motors wasn't starting with the lowest speed.
It is clearly a mechanical problem of at least one of the motors. Maybe I try to apply a bit of oil in the gear box?
You wrote that the roboter is rarely controlled using the pulse key. In my opinion it doesn't which input device is used to control the speed (at least if the slowest possible speed is the same for all input methods).

[firened]

STEPY and STEPZ are here

[firened]

Adjusting STEPn might be the best option in my opinion.

Scenarios to consider:
- robot may be on an in- or decline. Driving forward may need a very small amount to move.
- when driving forward with the joystick, it must be possible to apply tiny left and right commands or driving will become frustrating. -> Preferably no INS or OUTS adjustments.
- i use the pulse key as autopilot. I click the pulse key multiple times forward and then add 1-2 clicks left or right depending on terrain (floow may not be level). If you increase the minimum PWM, using it like that would not be possible anymore, since it'll turn too fast.

[ralph_62550]

That's how it looks like. It makes a difference how long you wait between different tries. It looks like when the motors just have turned before the chances are bigger that they turn with lower steps. Motor 2L is the one with the biggest starting difficulties.

[ralph_62550]

[firened]

Ok interesting.
I'll check when mine start

[Mike118]

You can also check the solder. This can have a visible impact.

[firened]

On low battery the last motor starts turning on click 8

Screen_Recording_20260224_205236_Brave.mp4   2,94 Mo

[garfield2142]

Remove the bird's nest from the axles for extra power

[firened]

*Tactical bird nest. It's mission critical

[ralph_62550]

I tried resoldering but the behaviour of the motor is still the same.

[ralph_62550]

Thanks for testing. It is good to know the effect of low voltage. What is your defintion of "low battery"?

[firened]

The video was taken at around 3,1 - 3,2V Vbat.

Though i consider a battery to be low below 3,5V usually.

[garfield2142]

I just received mine.

[garfield2142]

@fire_ned Did you manage to install the fusion-hat module? It fails for me even on a fresh bookworm install. I can't even get the example speaker test to work.

[firened]

@Garfield2142 did you remove the picoTTS install from the script? That for one fails almost surely

[firened]

I removed more and more from the install script until it was able to run through