Re: [tlug] EE question (was: VPN?)

[Curt Sampson <cjs@example.com>]

On 2018-07-06 19:14 -1000 (Fri), David J Iannucci wrote:

> On another note... a question for the EEs in the audience.

I am not an EE, but I play one in my fantasies. Thus, take my
speculations here here with a largish grain of salt. (And I look
forward to any corrections.)

> I've recently acquired a sweet little "tiny form-factor" PC that I'm
> looking forward to throwing some distro on and playing with.

It would help if you could tell us what is is. Better yet, find the
spec sheet on line and give us a link to it.

> It apparently takes a laptop-style PS, and the specs are 19V @3.42A.
> As it turns out, I have one in my junk box that seems to fit the
> bill but the specs are 18.5V @<something >= 3.42A, can't recall
> now>. Should I worry about this thing being rated 0.5V below the
> machine?

Almost certainly not. The chances that it internally uses anything
more than 12 V internally are very close to nil, and it's likely that
all voltages within the system are 5 V and less. It will use internal
voltage regulators to bring the voltage from the PSU down to what it
needs, and these almost invariably use feedback loops that will
maintain the appropriate lower voltages so long as the input voltage
is a minimum amount above them (say, at least 6 V for a 5 V
regulator). You could likely run it off a 13 V (or perhaps even 6 V)
PSU, but see below.

The reason for using higher voltages between the PSU and the PC is to
reduce the amperage flowing across the cable between the PSU and the
PC so that the cable can be thinner. Using the standard P = VI (I is
amperage, for historical reasons) equation you can see that a 65 W
draw at 19 V will be 3.42 A, but the same 65 W draw at 13 V would be
5 A and at 5 V would be 13 A. You need a rather thicker (and thus both
more inconvenient and more expensive) cable to handle 13 A than you
do for 3.5 A.

So problems can appear in two circumstances here. First, the amperage
draw over the cable is high enough to cause it to heat, melt, catch
fire, whatever, which is not going to happen given your specs. The
second is that the computer will try to draw more power than the PSU
can supply which can lead to the same issues with the PSU. There are
often some protection mechanisms in place for this, some rather good.
For example, my ThinkPad T510 came with a 90 W 19 V PSU, but if I use
my smaller X200 laptop's 65 W 19 V PSU, when I boot the BIOS will pop
up a screen informing me that I'm using a lower-wattage PSU and it's
going to restrict the power usage. (Likely it's going to do something
like not let the CPU and/or GPU go into higher power modes. I don't
even use the GPU for any more than very minimal computing normally, if
at all, so I suspect that there's basically no effect here.)

Also note that the system will be spec'd for a current draw based on a
"fully loaded" system. One of the best ways to make sure it's not
fully loaded is to use an SSD instead of an HDD, which one would
normally want to do for performance reasons anyway.

Beyond that, Kalin covered some pretty important stuff.

cjs
-- 
Curt J. Sampson      <cjs@example.com>      +81 90 7737 2974

To iterate is human, to recurse divine.
    - L Peter Deutsch