Why are there chips in USB chargers?

[Leo Liu]

I learned that USB chargers had built-in processors when I was browsing the details of a USB charger. After doing some searches on Google, I discovered that the computing performance of a chip in a modern charger is probably better than that of OBCs aboard the Apollo 11. I wonder what the purposes for putting chips in USB charges are.

 

[berkeman]

https://www.st.com/en/interfaces-and-transceivers/usb-charging-controllers.html

 

[rbelli1]

https://www.ti.com/product/TPS65994AE

Here is another for the newer generation of USB charge technology. ST makes this kind too. See the menu on the left in berkeman's link.

These devices have state machines inside them to determine what should be happening with the power line. Calling that a processor is rather optimistic. I have no way of comparing them to the complexity of Apollo computers but the capability of the Apollo computers would be far superior as they can execute user configurable stored software.

However, if the charger has any features more than simple charging then the "better than Apollo" statement is likely true. The Apollo AGC was a bleeding edge computer that was part of what today would be a trillion dollar plus military project. That was over 50 years ago. Today even the slowest micro-controller would be faster. Making a programmable chip slower than the AGC would be massive undertaking that would result in a much more expensive product.

I don't know exactly what that capability is used for. The cost of a programmable chip is so low low that simply blinking a light is sometimes done with them. The difference in price between a blinking light circuit (555?) and a computer is so small that it is just easier to use a computer. Do you want the light to blink differently in different circumstances? Now you need more circuitry and it is more expensive than the computer.

BoB

 

[anorlunda]

If you want a cynical answer, the processor allows manufacturers to query the charger to find out if it is their own brand, then if not say, "Sorry, not supported."

The non-cynical answer is that it allows them to exclude cheap knock off chargers that may lack key safety features.

Welcome to the world of "smart" everything.

 

[Filip Larsen]

I haven't tracked the USB standards in any details, but I would expect the USB charging specification to include capabilities that requires some digital signal processing in addition to act as a power source. The specs themselves are not freely readable, but this short USB charging overview also seems to indicate that this is so. Perhaps someone with fresh knowledge of electronics can verify?

Of course, even if the above assumption is correct this does not mean the "processors" you see in your charger does not have functions beyond charging or even that they have nothing to do with standardized USB charging on that charger.

 

[Borek]

Also note, that while in principle you probably don't need much computing power to control the charger, it is often simpler and much cheaper to use a ready, universal microcontroller than to design a specialized chip. These microcontrollers are much faster and much more powerful than the Apollo computer, but their price in bulk can be in cents range.

From what I recall USB-C standard requires some initial handshake between devices, they exchange some data about their capabilities ("I can supply power", "I can only consume power" , "I can do both", "I am happy with 2A currents", "0.5A is my max"). So the charger has to actually be capable of being kind of a "network" device capable of sending and receiving data over the USB cable (no idea about the exact protocol involved, but judging from the variety of USB devices existing it must be quite flexible). That's quite a challenge - unless you use a ready, specialized chip, that technically is just a small computer. And a powerful one, just quite cheap due to the scale of the production.

 

[Rive]

the computing performance of a chip in a modern charger is probably better than that of OBCs aboard the Apollo 11.

Please note that that 'computing power' is pretty much negligible by modern standards.

 

[jsg2021]

If you want a cynical answer, the processor allows manufacturers to query the charger to find out if it is their own brand, then if not say, "Sorry, not supported."

The non-cynical answer is that it allows them to exclude cheap knock off chargers that may lack key safety features.

Welcome to the world of "smart" everything.

No. That's not what happens.

READ the USB-3/USB-C charger standard! There is a negotiation to define who is the power source vs. power load, how much voltage and current can be supplied, etc. This is how these interfaces are able to charge devices faster than before. Either side can be the source or load in these newer standards. NOT possible with older USB. E.g. I have charge my MacBook through its USB-C or I can charge my iPhone from the exact same port - older USB can not do that.

Older USB standards used resistors to define available current (always the same 5V and always one end is the source vs. the load). These are backward compatible with USB-3/USB-C standards but charging will be slower. If the minimum standard isn't followed for the newest USB-3/USB-C chargers or devices.

You only get "problems" when the USB charger vendor fails to follow the standard which is common with cheapo chargers or devices. USB interfaces "speaking the correct language" results in misunderstanding and failure.