Moin!
On Sun, 15 Jan 2012 00:56:05 +0000
"gonzo ." <[email protected]> wrote:
> On the schematic the inputs are listed as "supply" (ie. "+15V_SUPPLY" and
> "+5V_SUPPLY"), where as the outputs of the regs are marked with their actual
> voltages.
> The boards silk screen on the other hand lists the inputs as 7V and 17V.
Ah.. ok.. overlooked that.
> In my original post I said I'd used the LM317 from the Eagle library, but
> infact use the pin compatible LM1084 (Eagle doesn't know about the LM1084).
> The LM1084 is a 5A device (vs 1.5A:LM317) with 1.5V LDO (vs 2.5V:LM317), so
> 17V is an acceptable minimum input.
Ok.
> Heatsinking of the LM/TS/AMS1117 is not generious and it does run hot when
> SJ1 is shorted and the board is running with a single supply (15V input to
> the 5VReg).
> I have added more heatsink ares to the V1.1 board, but the '1117is rated to
> 800mA and is only loaded to 10% so I don't think it is a problem.
Please be aware of the fine print: These chips can handle that much current
if and only if you can keep the junction temperature below the rated
maximum (usually 130-150°C). This means, you have to be able to disipate the
approx 1W while keeping the case temp below 120°C. Natsemi/TI say that their
SOT-223 case has a thermal resistance of 10°C/W (junction-case), so this
should be easily done if you have enough copper around or have a pad where
you can solder a small metal fin on (i guess any metal larger than 1cm^2
should be enough). Fortunately the 1117 has a themal shutdown that will
prevent it from frying.
And also a little warning here: There are many companies producing the "same"
device, but the specs have little differences, especially when you get
the device to its limits. In this case, when you let the *1117 run very
hot, you should keep in mind that different versions behave differently.
> It is worth remembering that the boards I'm offering are simply the left
> overs from my own use (V1.0 board at least) and was never intended to be
> fully idiot proof. Fortunatly, this list seems to be an idiot free corner of
> the internet
*g* not 100% idiot free, but the ratio of people who can think is very high :-)
Attila Kinali
--
Why does it take years to find the answers to
the questions one should have asked long ago?
_______________________________________________
time-nuts mailing list -- [email protected]
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
