[sdiy] Power input for 2xD / generic digital module [was: SMD part availability, manufacture]
Seb Francis
seb at burnit.co.uk
Thu Nov 13 12:10:22 CET 2008
John Luciani wrote:
> On Wed, Nov 12, 2008 at 7:31 PM, Seb Francis <seb at burnit.co.uk> wrote:
>
>> For the power supply regulators, I've been looking at low dropout ICs and
>> these seem pretty cool. Am I really understanding it right that they can
>> work with a output-input of down to like 0.2V. This would mean only a tiny
>> amount of heat dissipation so no need for any heatsinking. Are there any
>> disadvantages to these compared to standard linear regulators which need
>> typically 2.5V output-input differential?
>>
>
> Be careful when calculating the power dissipation. You need to take the ripple
> voltage into account. The trough may be at 0.2V but the peak could be much
> higher. The big LDO advantage is that you can size your filter caps a
> lot smaller
> since your ripple voltage can be 2.5V larger.
>
>
Good point! And anyway I guess it would need a really accurate input
voltage to ensure minimum power dissipation.
Which brings me to another thought on power. In making a generic
digital module, what is the most useful power arrangement for people.
Do many people have +5V available inside their modulars?
Is it useful to have the option that the whole board can run off only
+/-12V or +/-15V?
It would need clean +/-12/15V power anyway for the buffers and to be
regulated down to +5V for the analog side of the CODEC.
But also a power input of +5V would be useful, to be regulated down to
+3.3V to run the dsPIC and digital side of the CODEC (and hopefully also
the SRAM if I can track down a suitable and available 3.3V IC).
The same regulator could be used to get +3.3V from +5V or from +12/15V,
but obviously in the latter case the heat dissipation would be much
more, which probably restricts me to using a thru-hole TO-220 package.
Might be that I will use 2 regulators - one for pure digital, and one
for the digital side of the CODEC.
Any thoughts?
Seb
More information about the Synth-diy
mailing list