Standardization of DC power supplies (chargers) for CE

On a sunny day (7 Jul 2008 11:33:14 GMT) it happened n08W28+ snipped-for-privacy@cl.cam.ac.uk (Markus Kuhn) wrote in :

An act of parliament should be passed creating a standard committee whose approval would be needed for the introduction of any new style of connector.

You must be joking right?

With the toys (celphones, stuff) getting smaller every day nw connectors will be needed. Such a system as proposed would slow down development. I am not against standard connectors, but there arre already a few of those, problem is that those do not have standard voltages.

All my adaptors carry a sticker for which toy it is intended. And better too, the 12V adaptor has the same connector as the 9 V, and one has some sort of cable with all possible connectors. Using in the wrong adaptor could be very expensive.

For example the eeePC 9V adaptor has the same connector as the Alecto battery charger 12V adaptor etc. Boom.

So, a negotiation protocol, 2 $ you say? Many of those adaptors cost only 2$ to make (or less) in China. Maybe a connector with keys in it, so you could only insert it in the right socket, but that conflicts with 'as small as possible'. And it would take years for a negotiation protocol to make it into all toys.

Then there is + and -, inverted too on some equipment.

So a new connector, but then high current and low current requires different size pins.... no, I think I will not sign that petition, ever.

"Bewildering variety"? Folks like you, or anyone that thinks it is a "bewildering variety" are "easily bewildered".

Bullshit.

Count me out. The makers should be allowed to build whatever they want. It's their product they are offering for sale, their power source they've intimately specified and in turn I can choose to buy it or not. Put a Euro committee on the job and we'll end up with expensive, abortion connectors and even more complications. I remember the original USB connector was designed by a committee of (software?) suppliers as 'standard'. Seems we're now up at about six varieties. I'd happily petition 'against' forced CFL usage, NHS privatisation and faith schools, or even petition 'for' a common Euro mains plug and socket etc but not for the technical can-o-worms mess, that 'standardised' wall warts would offer up.

Things may get a little better. The mini-USB connector is now obsolete for new designs and is replaced by the micro-USB. At the same time the USB standard has been updated to include a dedicated USB charger specification which is now coming into use for newer mobile phones and other products. (This was almost certainly the result of the Chinese initiative.)

The new USB charger will deliver 5V at up to around 1A. I suspect that this charger will become the preferred one for many small devices. Out of all the connector variants, only two are allowed for USB chargers. The large original type A connector is to be used where the power outlet connector is in the charger body and the type A micro- USB plug is to be used if a flying lead is fitted.

Devices detect the new charger by looking for a short circuit between D

• and D- contacts. This replaces previous unofficial methods such as that used by Apple where a network of resistors was connected to the D

• and D- contacts.

John

My junk-cables box has standard type A and B connectors, and five different variants of miniature USB tips. So it's at least seven.

Fourteen, if you count male/female as different types. More, if you allow internal (motherboard) USB connectors.

Standardization is wonderful. Proprietary connector schemes bedeviled SCSI for years, and now it's USB that is the major headache. The standards organizations didn't LIKE the design of connectors, and only did it grudgingly, because interoperability would die if one and only one manufacturer could make connectors. Some equipment makers had the idea that their 'parts' department would turn a profit on replacement cables (laptop modem dongles being the shining precursor example), so my Nikon Coolpix sports a nonstandard USB socket.

Oh. I think they do have a point.

A quick glance around my desk reveals that the older laptop PSU is an

18v 6.3A whereas the new one is 15v 5A - the lengths of the concentric plugs are very slightly different, but based on diameters I reckon one would fit into the others socket. They each have a fully decorated xmas tree of standards compliance stickers on.

Others lying around are all for different voltages and currents and connectors for modem, router, scanner, phone. Out of the random sample of about 20 in my office only 3 share the same nominal voltage.

Most are centre positive which is standardisation of sorts. And a couple are simple transformers with AC output. By far the most common connector is the concentric cylinder plug of various dimensions.

Some will certainly plug into entirely the wrong piece of kit if you were to try it.

It would be nice if the manufacturers did get their act together and had fewer random voltages and currents and/or standardised connectors so that if the connector fits it would guarantee that the supply was suitable. The units all seem to be made in China. Cylinder plugs with current encoded by OD, voltage by length and ID perhaps.

Consumer items should be a bit more nearly bullet proof wrt plugging in the wrong external psu in these days when almost everything comes with one. Various TV set top boxes and games being an obvious example.

Regards, Martin Brown

"Martin Brown"

** Which is an argument for MORE variety of such plug and socket types. ** Given the huge variety of possible voltages, current needs , regulated voltage or not - that means a HUGE increase in the number of plug and socket types. ** More proof that a greater variety of types is a absolute MUST !!

** Marvellous how some people can inadvertently produce such a strong cases for the OPPOSITE of their mad beliefs.

....... Phil

The obvious standardisation is an input voltage of 10-15VDC, which would allow everyone to use their computers etc. in the car, caravan, boat etc. direct off the battery (with "of course" suitable protection for load dumps etc.)

But these things are best left to the market (along with goldplate like EMC susceptibility), rather than standards committees. The consumer has to learn that you get what you pay for. There's no point in having laws that are applied and policed only sporadically if even that.

JS

In article , "Phil Allison" writes: |> |> > Some will certainly plug into entirely the wrong piece of kit if you |> > were to try it. |> |> ** Which is an argument for MORE variety of such plug and socket types.

You are making the unwarranted assumption that the different types are used to distinguish incompatible electrical properties. How nice that would be - but believing that it is so (or even would be) is merely naive.

There are exceptions but, in general, there is little correlation between plug/socket type and electrical properties. And, where such correlation exists, it is often imposed by standards organisations (and quite often ignored, even so).

Regards, Nick Maclaren.

What did they make wrong before, when that much power that close together was considered as impossible?

Maybe - as long as two/four contacts are enough. Sometimes USB casings where abused for e.g. 4+1 pins.

I'd welcome 5 V and micro USB as one standard, while there might be another one for 12 V (maybe up to 5 A).

However, step-up/down converters (or simple resistive voltage dividers) maybe inferior, burning energy for the desired internal voltage.

I hope the future might offer smarter devices which request a certain voltage from a smart supply when plugged in.

... and I expect smart (and cheap) battery packs which would manage to log the current charge/discharge status e.g. for less smart supplies (portable photovoltaic, bicycle generators).

iGo is one of the current companies which do provide smarter supplies and adapter solutions. At least a step in between...

- Martin

"Nick Maclaren" "Phil Allison"

** No - but they could be so used. ** You need reading lessons - fool.

** Which is just what is being lamented here - you over-snipping fool.

..... Phil

uct

ed

True, they should be ALLOWED to do any silly irreproducible thing they want. But, not ENCOURAGED.

The standard only relates to three, the A and B and one 'mini'.

I find seven (five different mini connectors, no less) represented in my box o' cables. Back in the old days, SCSI connectors numbered twenty variants, and it was not pleasant. Four standard connectors for SCSI were eventually specified, and things got... better.

For USB, I predict another three years of chaos. Has anyone seen the USB-3 connectors with fiber optic elements?

AFAIK, the micro USB connector addresses the only real engineering need (that being a physically thinner/smaller connector) than wasn't addressed by the regular USB A, B, and mini-B connectors: All the other USB connectors were cooked up due to wanting to save a few pennies on cost or to purposely restrict which peripherals could be used with the device (see, e.g., the Xbox -- all the controllers are standard USB, but with a purposely incompatible connector). I know of a defense contract at where the client specifically requested, "USB interface, but with a non-standard connector," so the manufacturer bought regular USB A->B cables, whacked off the B end, and replaced it with a DIN connector instead. Bizarre... plus of course the cables are now considerably spendier than if regular cables had been spec'd.

High-power USB peripherals are perhaps a bit of a grey area -- one can make something like "side by side" connectors with a standard USB socket and a high-power connector of some sort (laptop docking stations seem to do this fairly regularly), although I would accept that it raises prices enough and looks dorky enough that sometimes it perhaps is just better to use a single connector with beefier contacts/cabling.

Well there is one standard that's almost universally adhered to, but it might not be practical... it's the cigar lighter plug for cars. 12-18V DC at as many amps as you can take before the fuse blows.

How about an alternative solution... all devices should include their own power conditioning module/circuit. You throw them N volts AC at 'sufficient' current, where N is within some range (maybe 3-18V) and it's up to them to rectify and regulate this to whatever's wanted internally. The wall-wart is just a transformer and a fuse. If you want the wall-wart to be an efficient SMPSU, just feed DC into the low voltage AC input. (The rectifier diodes will have to be specced for this). There would probably need to be some kind of brownout detection in the regulator circuit.

This isn't too different from what happens already, it's mostly a documentation change. Rather than saying '9V DC, 1A' on some input that already has reverse protection and an internal switching regulator, just specify it to take all comers.

The really cheap stuff has a transformer-in-a-box and some bad linear regulation. I'm not sure of the best approach for this - make switching modules that are cheap enough to put inside?

Theo