12.6 V RMS * 1.414 - 2Vd ~ 16.4 Vdc at 1st C so you have 4.4 V extra
split it between 2 resistors each with 0.8 A
R= 2.2V/0.8 = 2.75 use 2.4 ohm
power = 0.8^2* 2.4 ~ 1.5 Watts each
buy 2 x 2R4 5W generic power resistors
place both in between 2 X 2200 uF caps should be clean enough
Thanks Ken, is it applicable for CLC though ?
Infinia ,why does it need to use two resistors? I guess in series?
800mA
What kind of mA?
Continuous, or maximum, or average?
It makes a big difference to the designing and selecting of components.
What kind of mA?
Continuous, or maximum, or average?
It makes a big difference to the designing and selecting of components.
2 resistors to handle the heat (wasted power) and give some symmetry.
actual R values are strongly dependant on your choice of transformer or wall wort and accuracy and stability of your current draw. I assume the regulator has a somewhat wide input
I used XFMR = 12.6VAC under load (very common part)
your current draw = 0.80 A
If you find any changes > change the calcs I used in post 21 to find the new resistor
overrate the power dissipation by a factor of 3 to 5 or you can burn something
Last edited:
While the tool doesn't compute load ripple for an CLC filter, you could simply apply the capacitor value computed for the LC filter for both capacitors in an CLC filter.
Infinia thanks for the detailed reply, I got an idea how to do it. Why did you pick 12.6VAC as the number to base your calculations.I thought after bridge rectifier it is DC with some ripple. Are you defining as 12.6 VAC because it is still AC after bridge rectifier?
no AC RMS voltage is what comes out of the transformer. the first cap tries to charge to the peaks, so you'll also see RMS*1.414 conversion used after the full wave bridge rectifiers, and -2Vd is two diode drops each are conducting for any AC half cycle.
the 12.6 # is a legacy ( generic) part left over from the tube filament days.
edit > BTW who suggested a Salas regulator for an entry digital main bd use?
doubtful you'd see any benefit on D/A analog outputs over a normal 3T series part. The digital parts create a high noise floor, and so many other strict measures have to be taken on a CPU board design to make separate analog supplies matter. at least you still have PI CRC filter for yer raspberry PI main bd. so that sounds good. LOL on second tht, maybe id call it a mirrored PI design (top and bottom R's ).
>IMO you should rethink your whole strategy to get 5V, cuz yer wasting a lot of power and parts here.<
Last edited:
Infinia,do you have anything better than Salas to suggest ?Diy or assembled?may be I am reading it wrong .you are suggesting Salas is not going to change anything since the raspberry itself makes a lot of noise.
Last edited:
Design for 500mA continuous and assume the extra for the peak is supplied from the last capacitor.
Then measure your final assembly and make adjustment if needed.
correct
use a LM317 or equivalent
IDK isn't there another RP forum and ask a broader question based on your end user or system goals
edit> there must billions of 5V power supplies already made or on their way to a landfill near u > anything from a wall wort from an old printer to a half frame linear to a open frame switcher.
Last edited:
- Status
- Not open for further replies.