3.3V on Pin of Unpowered Raspberry Pi 4B

[vne147]

Yes, the idea being that a 1k resistor will provide a suitable load to stop the supply voltage from getting above the brown out voltage of the CPU (anyone know what the BOV is?) and prevent it being "back powered"

Mike.

Got it. That makes sense to me now.

BTW, to have identical boards means repeating the circuit for however many PIs you have on every PI. Assuming each slave will listen on a different pin and the master will output on many pins to address each slave. Seems excessive.

Could there be a software solution so that you send all communication to all PIs with an address identifying which PI it's for?

Mike.

Yes the circuit will be repeated as many times as there are RPis networked together. They will all use the same GPIO pin regardless of whether that RPi is the master, or one of the slaves. The master RPi will configure it's GPIO as the output, and set a global enable/disable bit for the entire network. All other RPis will be slaves and configure their GPIOs as input.

As long as they see the master telling them to enable, they go about their business. If the master pulls that bit low, everyone halts and modes to a safe condition.

All the communication that requires the master to individually address a slave with be done via ethernet.

This discussion that started this thread is more or less a backup safety measure that's just intended to tell the system as a whole that things are OK, or to stop everything when it's not. There will also be redundant signals sent through ethernet.

 

[rjenkinsgb]

That'll still drive the entire supply bus up to nearly 3.3V and "back power" it. Still nothing should happen with the 100k resistor limiting current, and with the added diode now at least it'll prevent the GPIO from going far above the supply, but I feel like we're back to square one accepting that portions of the unpowered RPi will be driven up to 3.3V.

No!
As long as there is a protection diode at the input (whether internal or external), you are in effect trying to power the entire computer via a 100K resistor and diode.

The few microamps available through that resistor will not give any significant voltage across the device power system, with all the components connected to it and potentially drawing current.

On the Pi 4 I just checked it with, feeding 4.4V (5V minus a normal diode drop) to the 3.3V power via 100K gives just 72mV on the 3.3V power system.

 

[vne147]

No!

Whoa, buddy. I know we’re all excited about electronics here, but there’s no need to shout. We’re having a civilized discussion.

As long as there is a protection diode at the input (whether internal or external), you are in effect trying to power the entire computer via a 100K resistor and diode.

I agree.

The few microamps available through that resistor will not give any significant voltage across the device power system, with all the components connected to it and potentially drawing current.

On the Pi 4 I just checked it with, feeding 4.4V (5V minus a normal diode drop) to the 3.3V power via 100K gives just 72mV on the 3.3V power system.

I also agree. If the 3.3V pin went open circuit when the RPi was powered off (which we all know it doesn’t), then that pin would be driven up to 3.3V from the GPIO pin if it was still seeing a logic 1 from the master RPi. However, that pin doesn’t go open circuit. As you stated there are other components connected to it that are drawing a really small amount of current (micro amps) and pulling it down to the 72 mV you measured.

My concern all along has been that I don’t know what exactly it is that’s drawing that admittedly small amount of current. Maybe the answer is that it’s so small I shouldn’t care. I don’t know. That’s what I’m wondering about.

At any rate, thank you to everyone who offered their thoughts and advice.

 

[ChrisP58]

A few questions-
How many RPis could one system have?
How far apart are they? (wire distance not physical distance)
What would be the total wire distance end to end of the largest system?
What currents could there be between different nodes? Powered and un-powered.
What size is the ground wire between nodes? (could different nodes have different ground potential)

 

[vne147]

A few questions-
How many RPis could one system have?
How far apart are they? (wire distance not physical distance)
What would be the total wire distance end to end of the largest system?
What currents could there be between different nodes? Powered and un-powered.
What size is the ground wire between nodes? (could different nodes have different ground potential)

Chris,

Sorry, I didn't mean to ignore you. I just thought the conversation was finished. However, I'd appreciate any input you'd like to offer. In answer to your questions:

How many RPis could one system have? 10

How far apart are they? (wire distance not physical distance) 0.5m

What would be the total wire distance end to end of the largest system? 0.5m x 9 = 4.5m

What currents could there be between different nodes? Powered and un-powered. I don't have a hard answer for this one, but I expect very small (mA or less). All connections between nodes are for data only.

What size is the ground wire between nodes? (could different nodes have different ground potential) The nodes will be connected to one another by ethernet cables which I believe use 24AWG conductors. I'll only be using 4 of the conductors. 2 will be used for data D+/D-, one for ground, and the last to provide the discreet enable/disable signal that has been the subject of this thread. All RPis will be powered from the same mains which will supply one non-isolated AC power adapter outputting 24VDC for each RPi. So one 24VDC power supply per RPi. The 24VDC will be input power for another non-isolated DC/DC converter that steps it down to 5VDC to power the RPi. So, there may be small differences in ground potential, but I don't anticipate anything outside of the mV range.