From: Manuel Dejonghe <limbus_at_gmail.com>
Date: Fri, 11 Feb 2005 23:35:27 +0100

Hi Tom,

I didn't expect feedback from that article so soon :-)

There are three reasons why I post this on the rockbox-mailinglist as well:
- documentation. I would not like to write the same info a second time
if somebody else asks next month :-)
- knowledge: maybe somebody else has another idea.
- correction: maybe somebody else on this list reveals I got no clue
about electronics.

I strongly recommend you to get subscribed on the list. You will be
able to reply to everybody, and you will get anybodies answer,
independently of the fact wheter they forgot to add you to the list of
recipients or not.

Well, in the meantime, I read the papers (specs of the 4940) more
throughfully, and thought I should've taken the V12 one, as the low
voltage behaviour is better than I expected, but thats another story.
I first wanted to wait to see how mine behaves before adding info to
the article. My fear was that 10V could be a small bit to weak to
charge properly, but it does just very well.

So, now to your problem:
No, mine is not getting very hot at all. I tested a few times by
holding the small package in my hand after having an half an
hour-travel (that the most I am doing at the moment) and it was so
cold I could not determine if it comes from ambient temperature or
not. The black brick I mounted on it is merely a buffer than a
heatsink as well. Else I wouldn't have it packaged (and thus
preventing from cooling) with the thermo-shrink tube. I just mounted
it (the black block) for security as IF the unit drains more power
than expected, it would only be for a short while like hard disk power
up/spin up.

I know from experience that such voltage regulators are allowed to get
hot. I've once been told to check the temperature by touching it with
a wet finger. If it fizzes, it would be too hot. So that shows you
what temperature regulators are allowed to get, and up to which
temperature regulators are supposed to work without getting damaged.

There is two things that may heat up the regulator in regular
operation: to much power drain or input voltage different from what we
expected (I assume below 8 volts, or even less, and above 17 volts, as
the data sheet says)

Now there are two things that is disturbing me:
- The first is that yours is getting hot, and not mine. Temperature is
usually rised if your device drains a lot of power/current (ampere) or
if the input voltage is higher.
- The second is, that the voltage regulator is known (or at least
advertised) to have a very decent temperature controlling. It would
rather stop delivering that current than burning itself to death. I'm
just checking the spec sheet to see what may be the cause. I assume
"too hot to touch it" is above 75 °C
Unfortunately, the data sheet does not have a figure showing the
result temperature of output current or something like that.

Now I can think of two things:
- Either you did not pay attention (I admit, I did not warn in the
article about that, should add that) to not grill the regulator by
assembling it. I absolutely don't know how long it should be exposed
to which temperatures during soldering. The data sheet says the max is
150 °C, but this is certainly for a longer time, as I know I soldered
it with 300-350°C. I've been teached to always pay attention with
diodes, transistors, triacs and so on (in general, all
semi-conductors) to not over-heat them, but I've seen people handling
them I would've bet they broke it, and it worked, so maybe it's not
that sensible my vater taught me. I actually never killed any
component during soldering by accident.
- There is more power drain as expected. If there would be a
short-circuit between regulator and archos-device, you would not tell
me it would be working very well, as there would be no current
arriving at the archos anymore. So the only one thing I can imagine
of, is that something in the archos broke so in trains more power
somewhere, most probably a loose contact then.

So I definitely recommend you to measure the load/draining current of
the archos as well. Most probably, it will not drain more than 600 mA
in ANY case.

Unfortunately, I got not enough practical experience, to be able to
predict anything.

I hope that helps, Manuel (lImbus on irc)

On Fri, 11 Feb 2005 14:02:11 -0600, Tom Clemmer <t-clemmer_at_comcast.net> wrote:
> Manual,
>
> Thanks for the write up in the car adapter using an l4940V10. I finished
> mine and everything measures good. When I power the Archos, the L4940 gets
> very very hot, too hot to touch. Now I do not have a heatsink on it but was
> wondering if your was getting very hot as well?
>
> Tom

_______________________________________________
http://cool.haxx.se/mailman/listinfo/rockbox
Received on 2005-02-11