Sustainability of switching power supplies

[freebsd-chat mailing list]

Hi,

does anybody know about papers related to the sustainability of
switching power supplies?

In my very limited experiences they are seemingly less sustainable than
old school power supplies.

I'm curious, if any paper does exist related to energy efficiency of
usage of a power supply, but also taking into account how much energy
is needed to repair or replace those switching thingies and classic
stone age power supplies. And apart from energy efficiency, how much
other resources are required to produce and maintain different kinds of
power supplies?

My experiences with modern computer power supplies aren't bad, but
my impression about switching power supplies in almost everything and
the sustainability of this approach ... voltage undersized caps etc.
isn't good.

Regards,
Ralf
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-chat
To unsubscribe, send any mail to "[hidden email]"

[freebsd-chat mailing list]

On Sat, 21 Nov 2020 09:24:44 +0100
Ralf Mardorf via freebsd-chat wrote:

> Hi,
>
> does anybody know about papers related to the sustainability of
> switching power supplies?
>
> In my very limited experiences they are seemingly less sustainable
> than old school power supplies.
>
> I'm curious, if any paper does exist related to energy efficiency of
> usage of a power supply, but also taking into account how much energy
> is needed to repair or replace those switching thingies and classic
> stone age power supplies.

In theory the difference in losses can be huge. You can see this more
starkly with  audio power amplification. Linear amplifiers typically
have big transistors bolted to big heatsinks. Switched amplifiers may
just use a small plastic chip with no heatsink at all.

In practice the typical switched computer supply is not hugely
efficient, but it there's a significant advantage that it's easy to
build them with a near perfect power-factor. This doesn't much matter
in the home, but commercial electricity customers pay extra for a poor
power-factor because it would create higher losses in the the supply
network if not compensated.

I think that meeting power-factor regulations was the main reason for
the switch.


> My experiences with modern computer power supplies aren't bad, but
> my impression about switching power supplies in almost everything and
> the sustainability of this approach ... voltage undersized caps etc.
> isn't good.

Generally things are getting less over-engineered than they used to be.

One problem with linear supplies is that they tend to have much larger
smoothing capacitors and so contain more corrosive liquid to drip onto a
more expensive circuit board.


_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-chat
To unsubscribe, send any mail to "[hidden email]"

[Brett Glass-2]

At 01:24 AM 11/21/2020, Ralf Mardorf via freebsd-chat wrote:

>My experiences with modern computer power supplies aren't bad, but
>my impression about switching power supplies in almost everything and
>the sustainability of this approach ... voltage undersized caps etc.
>isn't good.

Not sure what you mean by "sustainability" here. If you mean their
environmental
impact, well, it depends. If properly designed, switchers are more
efficient than
linear supplies, generate less heat, and waste less energy.
However, because they're
more complex and contain more toxic metals and more solder, they
have the potential
to generate more harmful waste when they're retired.

If you mean reliability, they're actually pretty close. Switchers
are more resistant
to power surges (because they can withstand higher input voltages
with no damage)
but suffer due to their complexity; there are more parts to fail.
And the part that
fails most often in power supplies - the electrolytic capacitor -
is present in both
types. (The "capacitor plague" of the last decade is still haunting
us today, because
even now millions of power supplies made with the faulty capacitors
are still in use
and/or still in supply chains. But even good electrolytics are essentially
quick-charging batteries and do fail more often than other components.)

That's why Glass's Law of Electronic Diagnosis states: Whenever you
are asked about the
failure of an electronic device, simply say, "It's the power suppply."

You will be correct about 99% of the time, and since you came up
with the right answer
before you even looked, you'll be considered to be either psychic
or a bloody genius. ;-)

--Brett Glass



_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-chat
To unsubscribe, send any mail to "[hidden email]"

[soralx]

> [...]
> and/or still in supply chains. But even good electrolytics are essentially
> quick-charging batteries and do fail more often than other components.)
>
> That's why Glass's Law of Electronic Diagnosis states: Whenever you
> are asked about the
> failure of an electronic device, simply say, "It's the power suppply."

These days, the law does not hold. Power supply failures are rather
infrequent, and will be encountered even less in the future. Most
products utilize switching PSUs nowadays (due to low cost, size &
& weight, and efficiency). Thanks to improvements in solid-state
electronics, switching frequencies in MHz range are now the norm;
fast-switching PSUs require relatively little capacitance in high-
-current paths, which can be satisfied with reliable MLCC (ceramic)
and polymer types. Semiconductor devices used in modern mainstream
PSUs run much cooler that before, too, so have exponentially longer
life (in accordance with Arrhenius equation).

> You will be correct about 99% of the time, and since you came up
> with the right answer
> before you even looked, you'll be considered to be either psychic
> or a bloody genius. ;-)
>
> --Brett Glass

--
[SorAlx]  ridin' VN2000 Classic LT
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-chat
To unsubscribe, send any mail to "[hidden email]"

[freebsd-chat mailing list]

On Sat, 21 Nov 2020 17:30:08 -0700, Brett Glass wrote:
>At 01:24 AM 11/21/2020, Ralf Mardorf via freebsd-chat wrote:
>
>That's why Glass's Law of Electronic Diagnosis states: Whenever you
>are asked about the failure of an electronic device, simply say, "It's
>the power suppply."
>
>You will be correct about 99% of the time

That's exactly my experiences with gear using switching power supplies,
but not with gear using classic power supplies. And yes, when I ask for
"sustainability" everything should be taken into account. The better
efficiency, but also other variables, such as special waste, if the
device is borked.
_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-chat
To unsubscribe, send any mail to "[hidden email]"