They raised max USB output voltage from 5.25V to 5.50V. Retroactively.

[ivan747]

The USB Implementers Forum raised max USB voltage from 5.25V to 5.50V, retroactively.

I took this personally. They got away with literally raising the maximum voltage of the USB port. With billions of devices in the field. Those devices were designed to receive 5.25V maximum. And now they can receive 5.5V from a new port.

In manufacturing, my deviation requests used to get denied all the time. So many hours getting everyone in the plant to agree, so many hours writing good arguments on the request forms... Most of the time, if they got denied it was for good reason. But I still had to request them to make management happy.

I have to laugh because the other option is to cry 

USB ENGINEERING CHANGE NOTICE

Title: USB 2.0 VBUS Max Limit
Applies to: Universal Serial Bus Specification, Revision 2.0
Summary of ECN
Update the USB2.0 specification to align with the expanded VBUS voltage definition in the USB-IF Type-C
specification which defines VBUS as having a Max supply voltage of 5.50 V DC.
The Type-C cable and connector definition specifies a minimal current carrying capacity of 3A. This increased
capacity creates a situation where losses in cables, connectors, PCB traces and power supply components will result
in the VBUS level available at the consumer end of the cable that is below the minimum acceptable level for
operation. As a result, in order to have an implementable Type-C specification it is necessary to boost the upper limit
for VBUS at the supply side to 5.5V DC from the present spec value of 5.25 V DC.
Reasons for ECN
OEM’s can design products that support the higher power levels of the Type-C connector and provide VBUS levels at
or above 5.25V to ensure that the VBUS voltage that is available at the far-side connection is within a suitable working
range for USB products, and is also at a high enough level to charge batteries in as short a time interval as possible to
ensure customer satisfaction.
Impact on Existing Peripherals and Systems:
The impact is difficult to assess [you don't say!?   ] as there is a wide range of component vendors supporting USB, and also a wide
variation of design methodologies.
As a result, there are certainly going to be a subset of existing USB products that may be sensitive to operation at
VBUS levels of up to 5.5V. The sensitivity will be mitigated in many cases by cable losses, so the devices themselves
may never experience the 5.5V.
Hardware Implications:
New designs adhering to this ECN will need to consider the new 5.5V VBUS max level during the component
selection process.
Software Implications:
There are no known software implications.
Compliance Testing Implications:
Compliance will need to make the small modifications necessary to accommodate the new upper limit.



Emphasis added by myself.

Note: if anyone at the USB-IF is reading, I imagine this was not an easy decision. It's just funny how all this strict compliance is thrown out the window when the approvers are also affected and there are no other realistic solutions.

[SiliconWizard]

That's good to know.

[magic]

In a world without Apple there would be no reversible connectors, no USB C, none of that nonsense.

Just sayin'

[DavidAlfa]

Absolutely no worries.

[SiliconWizard]

That shouldn't cause much worries, but still something to keep in mind.
Many "5V" ICs, such as regulators, are rated up to 5.5V only. So, instead of having a small margin, now we'll be running with no margin. Sure those ICs have usually an absolute max of 6V or 7V, so nothing special should happen here, but having a reduced effective margin is never that good news.

They didn't bother to adjust the minimum Vbus as it didn't fit the rationale - AFAIR it was symmetric (4.75V to 5.25V), now it's going to be asymmetric.

[thm_w]

As a result, there are certainly going to be a subset of existing USB products that may be sensitive to operation at VBUS levels of up to 5.5V. The sensitivity will be mitigated in  many cases by cable losses, so the devices themselves may never experience the 5.5V.

I would think the person who wrote this should understand USB and would know this statement is BS. USB devices are required to support a sleep state of less than 2.5mA. So any time a device enters sleep state it would be subjected to the full 5.5V. Of course, during full load, you won't see that, so maybe that is more what they were getting at.

They could put in somewhere to only supply 5.5V at output currents of >1A, but that would create a complex regulator.

[ivan747]

That shouldn't cause much worries, but still something to keep in mind.
Many "5V" ICs, such as regulators, are rated up to 5.5V only. So, instead of having a small margin, now we'll be running with no margin. Sure those ICs have usually an absolute max of 6V or 7V, so nothing special should happen here, but having a reduced effective margin is never that good news.

They didn't bother to adjust the minimum Vbus as it didn't fit the rationale - AFAIR it was symmetric (4.75V to 5.25V), now it's going to be asymmetric.

Also those SOT-23 linear voltage regulators. They have very small power dissipation margins. For example, with a TPS7A0533PBDZ that extra voltage 250mV at 200mA will raise the junction temperature by 13 degrees Celsius.

[langwadt]

As a result, there are certainly going to be a subset of existing USB products that may be sensitive to operation at VBUS levels of up to 5.5V. The sensitivity will be mitigated in  many cases by cable losses, so the devices themselves may never experience the 5.5V.

I would think the person who wrote this should understand USB and would know this statement is BS. USB devices are required to support a sleep state of less than 2.5mA. So any time a device enters sleep state it would be subjected to the full 5.5V. Of course, during full load, you won't see that, so maybe that is more what they were getting at.

They could put in somewhere to only supply 5.5V at output currents of >1A, but that would create a complex regulator.

isn't the sorta implied but not spelled out, so ..

"boost the upper limit for VBUS at the supply side to 5.5V DC from the present spec value of 5.25 V DC"

[Someone]

They could put in somewhere to only supply 5.5V at output currents of >1A, but that would create a complex regulator.

Some USB PD sinks and sources already implement cable compensation, its a bit of a wild west.

[dferyance]

USB is a LiPo charging and pocket-lint collecting port. Don't try to make it into something it isn't. Someone should invent a sort of serial bus to allow for computer peripherals to connect to the computer. When that happens, we'll have to be careful on the voltage levels.

[Siwastaja]

The sensitivity will be mitigated in many cases by cable losses, so the devices themselves may never experience the 5.5V.

Are these guys serious, they don't understand Ohm's law and the fact that USB powers digital electronics which totally have a tendency of varying load from zero to maximum, so that voltage ratings would be exceeded at small load?

[magic]

They had to justify it... somehow...

[thm_w]

Are these guys serious, they don't understand Ohm's law and the fact that USB powers digital electronics which totally have a tendency of varying load from zero to maximum, so that voltage ratings would be exceeded at small load?

Yes thats what I said. It looks like this is a very old change, from 2014: https://web.archive.org/web/20211103140823/https://www.those.ch/designtechnik/wp-content/uploads/2014/08/USB-Type-C-Specification-Release-1.0.pdf

The actual spec says it more reasonably:

4.4.2 VBUS
The allowable default range for VBUS as measured at the DFP receptacle shall be as defined by the USB 2.0 Specification and USB 3.1 Specification. Note that due to higher currents allowed, legacy devices may experience a higher voltage (up to 5.5V maximum) at light loads.
The DFP’s USB Type-C receptacle VBUS pin shall remain unpowered until a UFP is attached. The VBUS pin shall return to the unpowered state when the UFP is detached. See Table 4-15
for VBUS timing values. Legacy hosts/chargers that by default source VBUS when connected

Surprised that most people still quote the 5.25V number since that date.

[magic]

They didn't bother to adjust the minimum Vbus as it didn't fit the rationale - AFAIR it was symmetric (4.75V to 5.25V), now it's going to be asymmetric.

Stop giving them ideas

Besides, that's the voltage at the root port, and various losses are permitted so that a compliant device actually needs to at least enumerate at supply as low as 4.4V. Chapter 7.2 of USB 2.0 specification.

I'm pretty sure the original spec was simply copied from ATX, which is 5V ±5%.

[SiliconWizard]

I'm pretty sure the original spec was simply copied from ATX, which is 5V ±5%.

Yes, that's likely.

[amyk]

I'm pretty sure the original spec was simply copied from ATX, which is 5V ±5%.

Yes, that's likely.

...and that spec itself came from 5V TTL logic.