Power Interruption with Multi-Port Wall Chargers (GaN & Non-GaN) When Plugging in or Unplugging from Another Port

johnwickspuppy on 30/04/2024 - 11:45
Last edited 30/04/2024 - 16:22 by 1 other user

This is a PSA for people researching wall chargers that don't exhibit power interruption. There seems to be a bit of confusion on this.

You may require uninterrupted power for devices like:

• Raspberry Pi
• USB fan
• Baby camera monitor

Here is a conversation on the need for uninterrupted power if you want to be entertained.

There are claims that all GaN chargers are unable to deliver uninterrupted power. It has been called a feature of GaN but I found this to be false.

Another claim is that all non-GaN chargers are able to deliver uninterrupted power. This is also false.

Here is a Reddit discussion looking for GaN chargers that deliver uninterrupted power. The newest comment claims that such a charger does exist and provides evidence.

It seems that manufactures don't clearly specify whether their charger can deliver uninterrupted power. Maybe because this is not a big issue for most people? But it is to some and it would be nice if they could find out before purchasing.

One workaround is to plug in your device to a separate USB wall charger for devices that require constant power.

I am not a techie or wall charger expert. Feel free to add your knowledge on this topic if you know more.

• 

  bio on 30/04/2024 - 12:48

  One workaround is to plug in your device to a separate USB wall charger for devices that require constant power.

  …or a small powerbank between the charger and the device? There is a reason we invented something called "Uninterruptable" Power Supply decades ago.

  I would do that for a Raspberry Pi anyway, with or without a GaN charger.

  • 

    johnwickspuppy on 30/04/2024 - 12:52

    Great idea! I haven't heard of this one. Do all powerbanks supply power while charging? I'm guessing most modern ones would but I'm not sure.

    • 

      bio on 30/04/2024 - 13:05

      Not all of them can do. It needs to support pass-through charging.

    • 

      trongy on 30/04/2024 - 13:46

      Raspberry Pi users may want to look at a UPS designed specifically for the Pi.

      example 1

      example 2

• 

  trongy on 30/04/2024 - 13:43

  What you are talking about is the difference between a USB (battery) charger and a power supply designed for continuous power delivery.

  Most battery powered devices can tolerate short interruptions when charging the batteries.

  The interruption occurs due to the connection being reset as the maximum charging rate on the primary port is lowered when the a second device is connected (or disconnected). This is done so that maximum power can be delivered when only a single port is in use, while also allowing for safe charging of multiple devices on the same charger.

  Since most chargers are not suitable as continuous power supplies a simple strategy would be to use separate chargers - one for devices that are never removed and one for temporary battery charging.

• 

  CodeXD on 30/04/2024 - 16:14

  It sounds like the lesson here is if you require uninterrupted power for devices on a multi-port charger, make sure that all the devices on that charger requires it. In which case there shouldn't be any power re-negotiation as you won't be constantly plugging/unplugging into the port. Ofc this means you'll need all your devices in a central location which might not be the case.

• 

  Namidairo on 30/04/2024 - 19:50

  I've seen a couple of the pricier ones being reviewed wherein they don't do the PD renegotiation if conditions are met, but it's quite obviously in the minority and limited to… not cheap ones.

  Eg. https://youtu.be/qKY5SLwCwd0?t=166

• 

  Prong on 01/05/2024 - 22:21

  +1

  USB-C PD is still a bit of a mess, because there is enough variance in the possible implementations that it can cause issues.

  The USB-C PD spec doesn't actually require charger disconnection for power level renegotiation. But many chargers do so because it is the surest way to make sure the renegotiation works as expected every time. But as the electronics get smarter, we should see more in the future that don't disconnect.

  If you have a low power device like a Pi etc then Limeta EB40 is a great 10,000 mAh UPS power bank option for $30. It does 5V 3A, 9V 2A, 12V 1.5A.

  https://www.ozbargain.com.au/node/836987

  Each USB-C input / output is fully independent, so you can use it to do pass through charging and it works as a UPS with zero interruption. It also has a Bluetooth app that allows the max capacity to be set to 80% for longevity. Plus it uses Samsung INR21700-50E cells.

• 

  johnwickspuppy on 27/05/2024 - 16:40

  +1

  @Prong posted a comment in an Anker 323 charger deal where he provides a very detailed explanation on exactly how power interruption operates on this particular charger. I will reproduce it here to save you some clicks.

  So to add to this, I did some extra testing.

  As has been noted, the USB-A port is always interrupted when the USB-C port is plugged in or out.

  The USB-C port is interrupted any time it is not using the (5V 3A 15W) or (9V 2.22A 20W) PDOs, and the USB-A port asks for more than 5V 1.5W.

  EG, if you are running a laptop at 20V via USB-C, and plug something into the USB A port that wants more than 1.5W (like a phone) then the USB-C port will disconnect power, reconnect, and tell the laptop there is only 5V 3A or 9V 2.22A available. If you plus something that uses less than 1.5W (like a USB light, or charging some earbuds etc) then the USB-C does not disconnect.

  So if a phone is charging from the USB-C port using PPS (EG, 8.5V, 2A), then plugging something in to the USB-A port will cause the USB-C port to negotiate to charge at 9V and up to 2.22A (or 5V 3A). The switch from PPS to 9V 2.22A is extremely fast and seamless, so the phone does not indicate charging has stopped. I will have to log the voltage change with a high sample rate to see if it actually disconnects or just adjusts the voltage up as it changes PDO.

  If you are charging a phone from USB-C using the 9V 2.22A PDO (which is most common with phones) or the 5V 3A PDO then plugging something into the USB-A port will not cause the USB-C port to disconnect.

  This is because when both ports are in use, the USB-A port is limited to a maximum 5V, 12W. And the USB-C port is limited to 20W. So if the charger already knows the USB-C output is no more than 20W, it doesn't have to renegotiate it when the USB-A port is used.

  @johnwickspuppy In your case, the phones are likely using PPS (seamless switch), or already using the 9V 2.22A or 5V 3A PDOs.

• 

  rumblytangara on 10/06/2024 - 18:26

  I have never managed to understand why this is a big deal… any of the devices I have that require continuous power are either located in places far away from a centralised "group" charging station so they have their own cheap little USB power supplies. Or for the sake of argument if they were fans or cameras, I don't care if they are rebooted for 30 seconds occasionally.

  Anything that really sucks down a lot of wattage over USB-C is likely to have its own battery- it'll be a laptop or a phone.

  I've had a Pihole running on a Pi Zero dangling off a group charger for the better part of a year, but because it was one of two local DNS servers it didn't matter if it was rebooted every time someone else plugged in a device. If it was important, I could have just relocated it to a dedicated charger.

  Why is this power interruption thing such a big deal? Surely for any critical uptime device it's easy to just have a dedicated $20 power supply rather than start messing around with power banks acting as pseudo-UPSes.

  • 

    Prong on 11/06/2024 - 12:46

    +1

    Main thing for me is when using a charger with a USB-C dock, hub, or device doing pass through. If the charger cuts power when it renegotiates then everything downstream has power cut and has to renegotiate.

    EG when travelling / working on the go I often use a USB-C monitor that passes power through to the laptop via the same USB-C cable it uses for data. (Or a hub / dock doing similar).

    If I plug something into the second port of a charger that cuts power then the monitor / any peripherals turn off, the laptop moves any open apps into the one display. Then power reconnects, everything re-negotiates and the laptop changes video mode, moves apps back etc. Windows is pretty good at doing this without issue but my work Macbook will often get confused, not reconnect things correctly, not move things back to the screens they were on etc.

    Of course this issue is easily avoided - if I need to fast charge more than one device, bring two chargers (or use a different setup). But that's annoying too. Especially if running from a power bank.

    The other thing is if running USB lighting from USB-A ports. Not an issue for most people, but I have all sorts of USB lights I use and many default to off after power is cut and restored. So they need their own charger if I want to avoid having to manually turn them back on, which means I often have USB charger ports and charge capability going to waste.

    • 

      rumblytangara on 11/06/2024 - 13:40

      Yeah, monitors were the one thing I could think of where it would be very noticeable and annoying.

      Still, I think that it's an edge case affecting a very small number of users, but every thread for multi port chargers seems to have someone chiming in with the disconnection thing.