mosfet - clutch delay on release

[Vic]

hi,

I've implemented the LR7843 mosfet module for engaging the clutch on my raymarine drive. And when I put the pypilot in standby, it takes about 5 seconds to relaese the clutch. The activation light on the module is also still on for about 5 seconds. It seems that the mosfet module is still active for 5 seconds, after the standby command is given.

Has this something to do with the type of mosfet?

oh, by the way, when I engage the pypilot, it reacts instantly...

link to mosfet used: LINK

   

Regards,
Wouter

[seandepagnier]

Can you confirm where the delay comes from? Is there an led or? 5 seconds is very long. If you manually use a switch (no pypilot not software) does it have a delay?

[kniven]

Hi. 
I have the same issue. No big deal, really. But when I disengage the AP, it takes some seconds until the clutch opens. I have installed a manual "emergency switch" that cuts the power to the motor controller in case of communication/remote control issues. When I use this switch, the valve opens immediately.

I also do have the impression (although not systematically verified), that the delay time is influenced by the value of the servo.max_slew_slow and/or servo.max_slew_speed in the configuration menu.

[Vic]

Can you confirm where the delay comes from?    Is there an led or?   5 seconds is very long.    If you manually use a switch (no pypilot not software) does it have a delay?

I didn't do additional testing yet of were this comes from. Only that the mosfet module works without delay if I connect this to a switch.

I'll try to connect the pypilot to a separate arduino nano. And just put a led (only a led) on the clutch pin an ground. To see if this also occurs when nothing else is connected.

Keep you posted..

Regards, 
Wouter

[kniven]

I tried this on my tinypilot. When increasing the servo.period parameter to 2.5 seconds, it takes ages (>5 sec) for the clutch to disengage. I don't know if this is to be expected/normal, but. I do think that it is a software issue, since the "Sync Engaged" status remains active until the clutch releases.

[seandepagnier]

Thanks for the feedback.

There is a timeout... It is in the code "self.command_timeout > self.period.value*3"

This feature is not well optimized since I don't have a clutch on my own boats, sorry.

What it is doing, is enabling the clutch for manual control, or autopilot control. It only disengages the clutch if in manual control 3 times the period from the time the last movement was made.

I agree that this is obviously bad logic and wrong. I will update in future versions to immediately disengage when the pilot is disabled. I guess this could potentially interfere if you tried to make manual control immediately after disabling the ap (to dodge something or whatever) as it would have to re-engage, so this is the main issue, and it needs to at least not click the clutch off and on when using the dodge buttons. So a simple change is not simple, but I will find a way to make it work better.

[kniven]

Sounds good, Sean.
And yes. There has been some seconds of nervousness, waiting for regaining manual control...

[seandepagnier]

It will be greatly improved in the next version.

I recommend you install a manual switch on the clutch regardless for safety reasons.

[kniven]

It will be greatly improved in the next version.

I recommend you install a manual switch on the clutch regardless for safety reasons.

I have a manual switch. In case of communication/remote control issues. My switch is basically cutting power to the motor controller, but leaves the Pypilot computer on.

[Vic]

It will be greatly improved in the next version.

I recommend you install a manual switch on the clutch regardless for safety reasons.

I have a manual switch. In case of communication/remote control issues. My switch is basically cutting power to the motor controller, but leaves the Pypilot computer on.

Indeed, this seems like a good failsafe anyway. I already ordered somethin waterproof to fit besides the helm..