RPI 4 USB-C supply

[wildfire]

Some of the official retailers that sell Pis also sell there own power supplies. For example I have the following:
* CanaKit Raspberry Pi 3 Power Supply (Micro USB) - 2.5A
* CanaKit Raspberry Pi 4 Power Supply (USB-C) - 3.0A
Shouldn't these power supplies also be considered just as safe as buying an official power supply? Can't I assume that these power supplies meet technical requirements RPF has established?

If it's a trusted reseller and they describe the PSU as suitable for whatever model of Pi then you shouldn't have a problem. Here in the UK/EU you could return the goods as unfit for purpose (other jurisdictions may vary).

In the case of CanaKit, I don't shop with them but they are well known and I doubt you would have an issue with them.

Adding to my previous post, if you have a charger that's unofficial but works then great, use it but if it doesn't work don't complain to the RPF about it.

[bjtheone]

Actually the whole point of rigorous standards, such as USB, USB 2.0, USB 3.0, and USB C, is to provide interoperability between vendors. A good thing, no?

Why wouldn't I buy a RFP/RPT power supply for each Pi? Because I've already got plenty of 5V supplies sitting around doing nothing. Many of them, specially the cheaper ones, work fine with the Pi.

3 potential issues:

Rigorous adherence to standards and generic no name adaptors from China are not terms I would put in the same sentence.

Chargers/adaptors are not power supplies, and may not deliver the voltage regulation required. This is likely the biggest issue people face with trying to use phone chargers as Pi power supplies.

Does a standards compliant adaptor meet the Pi's requirements in all cases? Certainly USB 1.0 came out before such high current loads were envisioned. Not sure about USB 2.0 either. For USB 3.0 the requirement of +- 5% at the source's connector against the 5V nominal which puts a legal adaptor at 4.75V. Add in a connection adaptor if you are using a micro USB cable and you are likely going to have issues.

[davidcoton]

Does a standards compliant adaptor meet the Pi's requirements in all cases? Certainly USB 1.0 came out before such high current loads were envisioned. Not sure about USB 2.0 either. For USB 3.0 the requirement of +- 5% at the source's connector against the 5V nominal which puts a legal adaptor at 4.75V. Add in a connection adaptor if you are using a micro USB cable and you are likely going to have issues.

The voltage standard on USB is 5V +/- 0.25V, AFAIK the same for the three USB generations. The Pi is designed to work with that, though it can be marginal at the lower end (bottom limit voltage in will not guarantee compliant voltage out). This only affects peripherals, the Pi itself will happily work with input voltage well below the USB spec.

The minimum current standard even for USB3 (0.9A) is not enough for the newer Pis (probably OK for a Pi0). So any power supply designed for Pi2, 3 or 4 will need to exceed the minimum specification.

The other issue is the design error on the Pi4B USB-C power input, which means that active (e-marked) cables will not be compatible. This is where adherence to standards is a good idea, and the problem is listed for correction on some future board revision.

[W. H. Heydt]

The minimum current standard even for USB3 (0.9A) is not enough for the newer Pis (probably OK for a Pi0). So any power supply designed for Pi2, 3 or 4 will need to exceed the minimum specification.

Agree about the default 0.9A. Since the Model B originally spec'd 700mA, and the Pi0/Pi0W use the same SoC and don't have the LAN chip, 900mA is *plenty* for those boards.

[rpdom]

The minimum current standard even for USB3 (0.9A) is not enough for the newer Pis (probably OK for a Pi0). So any power supply designed for Pi2, 3 or 4 will need to exceed the minimum specification.

Agree about the default 0.9A. Since the Model B originally spec'd 700mA, and the Pi0/Pi0W use the same SoC and don't have the LAN chip, 900mA is *plenty* for those boards.

Don't forget that on the 1B, 280mA of that 700mA were for the two USB ports with 140mA fuses on each port. That means the Pi was expected to use up to 420mA at the original 600MHz clock speed.

I've measured a Pi 0W at a little over 300mA when loaded, and about 120mA when idling.

[jcyr]

Adding to my previous post, if you have a charger that's unofficial but works then great, use it but if it doesn't work don't complain to the RPF about it.

It is the Pi4 USB C interface (designed by the RPF) that is non compliant, not the other way around. In fact I wonder how the RPF got USB certification for that design?

As for the legitimate complaint about it, it has already been made, noted, and will be corrected in future designs.

[jcyr]

@jcyr @brasta78

As I emphasized in my post, a charger is not the same as a power supply unit, they may look the same and in many cases a charger can replace a PSU but, one has regulated voltage when under load the other does not.

Actually, most if not all, brick type chargers are in fact power supplies. They consist of a transformer, a full wave rectifier, a few capacitors, an inductor, and a buck regulator. A USB C supply adds the ability to sense the current requirement by measuring the resistance of a few lines to ground. The circuitry required to manage the charging rate and voltage limits for lithium ion batteries is typically part of the device being charged (such as your cell phone or laptop), not the charger.

If you're talking about a lab grade calibrated power supplies, then yes there are major differences between that and a common brick supply. But otherwise, no. There is no difference between a brick charger and a brick power supply. The terms are synonymous.

[jdb]

The clue is this line:

The circuitry required to manage the charging rate and voltage limits for lithium ion batteries is typically part of the device being charged (such as your cell phone or laptop), not the charger.

With the charging control managed by the device that a charger brick is connected to, there is only a vague requirement to regulate the output voltage of this type of power supply.

When Pi 1 was released, many micro-USB chargers sold with phones of the time were found to be grossly inadequate, as at rated load they would deliver only about 4.2V due to sloppy feedback implementations.

The set of "good" devices encompasses both types, but you are more likely to get problems with a device that is advertised as a "charger".

[jcyr]

With the charging control managed by the device that a charger brick is connected to, there is only a vague requirement to regulate the output voltage of this type of power supply.

Sure, if a 5V 'charger' doesn't regulate to 5V at advertised max. load then it is out of spec. The same would be true for a 'supply'.

I was talking about the misleading statement emphasizing that brick chargers and brick power supplies are significantly different.

[drgeoff]

Sure, if a 5V 'charger' doesn't regulate to 5V at advertised max. load then it is out of spec.

I disagree. If the charger supplied with a device has adequate performance to charge the device's battery then it is not out of spec. That it might have "5V xA" on the label does not necessarily mean it is supposed to give 5 Volts when delivering that x Amps.

[jcyr]

Sure, if a 5V 'charger' doesn't regulate to 5V at advertised max. load then it is out of spec.

I disagree. If the charger supplied with a device has adequate performance to charge the device's battery then it is not out of spec. That it might have "5V xA" on the label does not necessarily mean it is supposed to give 5 Volts when delivering that x Amps.

Seriously? IANAL, but in the United States at least a label must accurately describe the contents and capabilities of the product.

[jamesh]

Seems to are repeating ad nauseum the same old stuff. Lock the thread?

[davidcoton]

Seems to are repeating ad nauseum the same old stuff. Lock the thread?

Yes please.