[NTLK] A 68kMLA forum member engineers a Newton power adapter

Andrei Chichak newton at chichak.ca
Mon Apr 14 02:24:55 EDT 2014

On 2014-April-13, at 6:35 AM, Morgan Aldridge <morgant at makkintosshu.com> wrote:

> A 68kMLA forum member has engineered a 3W Newton power adapter:
>    <https://68kmla.org/forums/viewtopic.php?f=14&t=23453>
> I'm no electrical engineer, so I can't comment on its accuracy or quality,
> but thought that the rest of you who are might be interested.

I work with a bunch of EEs. After a whole bunch of years of wrapping code around their power supplies and associated gizmal-gazmals I can say that we would only use an LM317 regulator as a very quick and dirty, wasteful hack.

Why? Well, the way that these regulators work is by taking the input voltage and throwing away enough voltage to make the required output.

So, with an input voltage of 9V and an output voltage of 7.5 we have to drop 1.5V at whatever current you are drawing. So 1.5V at 1.2A means that we will have to get rid of 1.5*1.2=1.8Watts in the form of heat.

As your input mains voltage goes up, the voltage on the output of the transformer goes up, the voltage of the circuit goes up, the input of the regulator goes up, the output of the regulator stays the same, the rest is wasted as heat.

One of the niceties of the Newton power supply is that it is happy with 120V in this part of the world, 220V in the middle part of the world, 200V in the other parts, without catching fire in Europe or Japan (since the input to the regulator will be roughly doubled).

His comment, that his schematic "is literally the reference implementation from the ST datasheet"is worrisome. Datasheets do not contain best practice implementations, they are just for reference and the EE is expected to add appropriate protection circuitry such as correctly sized components, current limiters, fuses, static electricity protection diodes, bypass capacitors, and whatever is required to keep the circuit safe when something fails. In this case, the author is not selling you a device, he is giving you enough information that you, as an EE, can put in enough stuff to make the circuit suit your installation, you make the kit, you assume the responsibility to make it safe and effective.

Remember, a couple of months ago, when some woman electrocuted herself with an iPhone? Apple's EEs designed their charger circuit to isolate the iPhone from ground so that there is no path for mains current to reach ground. No path, no current, no electrocution. It's a safety thing, isolating the hand-held device's ground from mains ground, you get the same thing with shavers and other things that might electrocute you when plugged into the wall. Well, the woman bought a knock-off iPhone charger because Apple charges so much for their version, and it's just a transformer and a voltage regulator, well apparently not.

> He also
> requests a correction to one bit of related data in the Newton FAQ, but I
> didn't quite catch what he felt the correction should be.

He shows that the FAQ has the output voltage of the adapter as being 5V @ 400mA.

He figures that the real output is 7.5V @ 1.2A for the 2K series and the 1XX was rated at 7.0V @ 430mA.

More information about the NewtonTalk mailing list