usb hub backpowering

[jakky567]

What would be a simple way to test for backpowering or backfeed from a usb hub on the pi 3? I'm not interested in powering my pi with a hub, as I have a good quality power supply, but would it interfere/supply less clean power?

[W. H. Heydt]

What would be a simple way to test for backpowering or backfeed from a usb hub on the pi 3? I'm not interested in powering my pi with a hub, as I have a good quality power supply, but would it interfere/supply less clean power?

The power control through the LAN9514 chip doesn't go active until the SoC "boots" the LAN chip, so you can not boot a Pi3B by backpowering. In theory, once the Pi3B is up and running, if backpowering is in place you can pull the main power, but you can't start out that way.

More importantly, I wouldn't recommend doing it because the nominal requirement for the Pi3B is 2.4A, and most 4-port hubs don't have PSUs that will provide that much for the hub, let alone send it upstream. In addition, hubs aren't supposed to back power. If the hub does, it is poorly designed and there is no telling what else is wrong with it. That said, I do use hubs that will back power if used with a power supply, but I always use them as passive hubs, not powered, save in one or two early experiments. When I want a powered hub, I have some good ones, and they don't feed power back on their upstream port.

[jakky567]

What would be a simple way to test for backpowering or backfeed from a usb hub on the pi 3? I'm not interested in powering my pi with a hub, as I have a good quality power supply, but would it interfere/supply less clean power?

The power control through the LAN9514 chip doesn't go active until the SoC "boots" the LAN chip, so you can not boot a Pi3B by backpowering. In theory, once the Pi3B is up and running, if backpowering is in place you can pull the main power, but you can't start out that way.

More importantly, I wouldn't recommend doing it because the nominal requirement for the Pi3B is 2.4A, and most 4-port hubs don't have PSUs that will provide that much for the hub, let alone send it upstream. In addition, hubs aren't supposed to back power. If the hub does, it is poorly designed and there is no telling what else is wrong with it. That said, I do use hubs that will back power if used with a power supply, but I always use them as passive hubs, not powered, save in one or two early experiments. When I want a powered hub, I have some good ones, and they don't feed power back on their upstream port.

Is there a simple way to test for it though?

[W. H. Heydt]

What would be a simple way to test for backpowering or backfeed from a usb hub on the pi 3? I'm not interested in powering my pi with a hub, as I have a good quality power supply, but would it interfere/supply less clean power?

The power control through the LAN9514 chip doesn't go active until the SoC "boots" the LAN chip, so you can not boot a Pi3B by backpowering. In theory, once the Pi3B is up and running, if backpowering is in place you can pull the main power, but you can't start out that way.

More importantly, I wouldn't recommend doing it because the nominal requirement for the Pi3B is 2.4A, and most 4-port hubs don't have PSUs that will provide that much for the hub, let alone send it upstream. In addition, hubs aren't supposed to back power. If the hub does, it is poorly designed and there is no telling what else is wrong with it. That said, I do use hubs that will back power if used with a power supply, but I always use them as passive hubs, not powered, save in one or two early experiments. When I want a powered hub, I have some good ones, and they don't feed power back on their upstream port.

Is there a simple way to test for it though?

Test for what?

[B.Goode]

What would be a simple way to test for backpowering or backfeed from a usb hub on the pi 3? I'm not interested in powering my pi with a hub, as I have a good quality power supply, but would it interfere/supply less clean power?

Just try it?

Create a spare SD card that you are prepared to sacrifice if things go wrong.

Boot your RPi from it using your conventional power supply on the microUSB connecter and with the hub powered up and connected to a full size USB port.

Once the RPi is running normally, pull out the power from the microUSB.

If the RPi keeps running the USB hub must be back powering it. It the RPi turns off, it is not. But in the latter case you may corrupt the SD card.

[davidcoton]

If the RPi keeps running the USB hub must be back powering it. It the RPi turns off, it is not. But in the latter case you may corrupt the SD card.

That does not prove there is no backpowering. Only that it is not adequate to power the Pi.
You need to measure the voltage on the power pins of the USB input to the hub, while the hub is powered but with no USB connections. You may need to sacrifice a suitable USB cable to make the measurement. It will be non-trivial to make sure you are testing the right connections, and it may be difficult to get good test connections if the cable has thin conductors.

[B.Goode]

If the RPi keeps running the USB hub must be back powering it. It the RPi turns off, it is not. But in the latter case you may corrupt the SD card.

That does not prove there is no backpowering. Only that it is not adequate to power the Pi.
You need to measure the voltage on the power pins of the USB input to the hub, while the hub is powered but with no USB connections. You may need to sacrifice a suitable USB cable to make the measurement. It will be non-trivial to make sure you are testing the right connections, and it may be difficult to get good test connections if the cable has thin conductors.

Agreed.

[jakky567]

Thank you all very much for telling me. I'll test it out this weekend for backfeed. It definitely doesn't backpower, but the pi 3 needs a bit more power, so that test really doesn't tell me much.

[Ronaldlees]

Thank you all very much for telling me. I'll test it out this weekend for backfeed. It definitely doesn't backpower, but the pi 3 needs a bit more power, so that test really doesn't tell me much.

Seems sensible. Otherwise, wouldn't there be a power supply tug-of-war going on?

[Ronaldlees]

1 Actor 1: This line is going to be 4.9 volts, dammit.

2 Actor 2 : Nooooo, it's going to be 5.3 volts (increase component current)

3 Actor 1: Ya think so? Nah - I DONT think so, 4.9 volts it WILL BE (increase component current).

4 Actor 2: Smart Aleck.

go to 1.

[W. H. Heydt]

Another way to test for a hub that back powers would be to connect to a Pi Zero.

[jakky567]

1 Actor 1: This line is going to be 4.9 volts, dammit.

2 Actor 2 : Nooooo, it's going to be 5.3 volts (increase component current)

3 Actor 1: Ya think so? Nah - I DONT think so, 4.9 volts it WILL BE (increase component current).

4 Actor 2: Smart Aleck.

go to 1.

Possibly the funniest thing I've read in a while.

[jakky567]

Another way to test for a hub that back powers would be to connect to a Pi Zero.

I forgot those were finally back in stock! I should pick one up.