SSD Booting : low power issue

[HeviiSailor]

I am using this one, its adjustable:  https://amzn.to/3rvaHif...

I am also working on an engine sensing project like Boatingbaileys...

Wow. The diagram is a tremendous asset, Jason. Thank-you very much!

With regards to your initial concern about connecting power directly to the RPi's GPIO pins, I had/have the same concern. Since, apparently, some protection is afforded by powering via the USB C port, this is my preferred route. After having had looked at a couple of Bailey's videos again, I realised that I can use a breakout board for this purpose. Even if connecting via the USB port isn't really necessary, it still provides an ability to quickly swap-out a Pi rather than having to open things up!

Here's a shot of typical board which I plan to use with my variable power supply: 

---

my schematic is attached ...

Once again, thanks for posting the schematic, Jason. 

Now that I've had a chance to go through it in detail, I have lots of questions!  Have you (or do you plan to) described/describe the ciruit and its rationale somewhere?

[Techstyle]

[quote pid='24680' dateline='1673281635']
my schematic is attached ...

Once again, thanks for posting the schematic, Jason. 

Now that I've had a chance to go through it in detail, I have lots of questions!  Have you (or do you plan to) described/describe the ciruit and its rationale somewhere?
[/quote]

I go through each circuit in this video:  https://youtu.be/enPFlHpmJVA

there is an older video that focusses on the power circuit, its a bit raw though: https://youtu.be/45IxpEa0ZcM

[HeviiSailor]

Once again, thanks for posting the schematic, Jason. 

Now that I've had a chance to go through it in detail, I have lots of questions!  Have you (or do you plan to) described/describe the ciruit and its rationale somewhere?

I go through each circuit in this video:  https://youtu.be/enPFlHpmJVA

there is an older video that focusses on the power circuit, its a bit raw though: https://youtu.be/45IxpEa0ZcM

And once again, "thank-you", Jason!

I'm quite impressed with your power supply system. As a mechanical kinda guy rather than a Sparky I can clearly see its logic. But, perhaps precisely because of my thick nuts'n'bolts brain, I'm having difficulties to see what's happening behind the scenes. The video makes it appear that the relay cuts the power to the RPi almost immediately after the panel switch is thrown. As I'm under the impression that the point of the 3.3V circuit and relay is to allow for a graceful shutdown of the RPi rather than an abrupt pull-of-the-plug, the video suggests (to me, at least) that this may not be the case. Currently (hah!), this circuit looks to be superfluous. At risk of looking an already very generous Gift Horse in the mouth, could I ask you to please expand on this as well? I really like your concept and I'd like to use it. However, like a child with a loaded gun, the results of my attempts could be rather damaging if I go at it half-cocked!

[Techstyle]

Once again, thanks for posting the schematic, Jason. 

Now that I've had a chance to go through it in detail, I have lots of questions!  Have you (or do you plan to) described/describe the ciruit and its rationale somewhere?

I go through each circuit in this video:  https://youtu.be/enPFlHpmJVA

there is an older video that focusses on the power circuit, its a bit raw though: https://youtu.be/45IxpEa0ZcM

And once again, "thank-you", Jason!

I'm quite impressed with your power supply system. As a mechanical kinda guy rather than a Sparky I can clearly see its logic. But, perhaps precisely because of my thick nuts'n'bolts brain, I'm having difficulties to see what's happening behind the scenes. The video makes it appear that the relay cuts the power to the RPi almost immediately after the panel switch is thrown. As I'm under the impression that the point of the 3.3V circuit and relay is to allow for a graceful shutdown of the RPi rather than an abrupt pull-of-the-plug, the video suggests (to me, at least) that this may not be the case. Currently (hah!), this circuit looks to be superfluous. At risk of looking an already very generous Gift Horse in the mouth, could I ask you to please expand on this as well? I really like your concept and I'd like to use it. However, like a child with a loaded gun, the results of my attempts could be rather damaging if I go at it half-cocked!

I am a mechanical engineer, but specializing in hydraulics which has lots of parallels to electrical (pressure = Voltage, Flow = current) so, like you, the logic works of me.  In the video, you are correct it turns off near immediately, that is because it wasn't really busy.  In reality, I have seen it take up to 15 sec to switch off.  More usually, it is about 1 sec.  

The way I have it set up is the switched power goes to a breaker and the constant goes to the master 12V (so in my case it is still switched), but I can hear the relay from my breaker panel and find myself standing waiting for it to go off!!

I wired it this way because I was worried about something happening and Pi staying on and draining the battery while my boat is on the mooring.  It has been 2 years of use now and it has never took more than 15 sec to power off, so I might swap it to be truly constant (unless the master battery switch is off).  I have never had a corrupted SD card with this system.  Before this it happened frequently.

having said that, it is unlikely that the master gets switched by accident - it tend to be somebody throw the instrument breaker, while trying to find the lights!

[HeviiSailor]

I wired it this way because I was worried about something happening and Pi staying on and draining the battery while my boat is on the mooring.  It has been 2 years of use now and it has never took more than 15 sec to power off, so I might swap it to be truly constant (unless the master battery switch is off).  I have never had a corrupted SD card with this system.  Before this it happened frequently.

I appreciate the time that you've taken to explain this, Jason. It's ïnspired me to dig deeper into the rabbit hole of safe RPi shutdowns.

[Techstyle]

I started off just installing a power switch.  momentary switch was connected across pins 39 (GND) and 40 (GPIO21).  an LED was connected up between pin 38 (GPIO20) and the same GND (39).

In the file /boot/config.txt (use sudo nano /boot/config.txt)

Add the following:

#on/off switch
dtoverlay=gpio-shutdown,gpio_pin=21
dtoverlay=gpio-poweroff,gpiopin=20,active_low="y"

this will command a shutdown when pin 21 is grounded and then when power is off, pin 20 will go low - this is a simple demonstrator and works well (and you can prototype with simple components)

mine looked like this:
     
   
simple and easy to use, but breakers got switched off...

[HeviiSailor]

I started off just installing a power switch... mine looked like this:

The gold just keeps flowing. Thanks, Jason!

[Sailoog]

Techstyle, that is why we added shutdown/power off to settings:

https://openplotter.readthedocs.io/en/3....y-settings

[HeviiSailor]

Techstyle, that is why we added shutdown/power off to settings: 

It's all coming together now, in my soft mushy mind.

[HeviiSailor]

Although I've not yet tackled Jason's safe-shutdown circuit, I've got my RPI booting and running off of the SSD now. WOW, what a difference in speed!

- I used my existing non-variable buck converter to power the RPi.
- The 5V output of the buck is connected to a USB C whip that I had made (this avoids having to connect to the unprotected GPIO pins).
- The SSD is powered by one of the the RPi's USB ports
- All other USB connections go to a USB hub powered by the same supply delivering 12V to the buck.