Well to answer that:
Consideration 1 :accept 32V and gives 5V @ 0.2A nominal. ie. at times abit higher than 0.2A
worst case: P = 27v * 0.3A = 8.1W.
Consideration 2: this 32V comes from SMPS, so even after the series zener diodes, dont think linear regulator is able get rid dirty noise originated from the sources supply.
Oh, for the love of potato salad.
Take your SMPS output (32 volts).
Run it thru a small ferrite core having 32 loops wound onto it.
THEN, Run it in series thru an 8 Ω resistor.
Then Into a 1000 μF cap.
THEN to the chain of zeners
And another 1000 μF cap.
THEN to the low-dropout linear regulator.
The loop inductor gets rid of the RF chop.
The 8 Ω resistor is part of the capacitor HI-cut filter.
The 1,000 μF cap totally sucks up almost all AF chop.
The Zeners drop 20 volts.
The next cap is the "keep things steady" reservoir.
The LDO regulator sets a constant output voltage.
And there you are. Maybe what I take as “obvious” isn't. Whatev's. The idea-chain, the design thought process is ALWAYS the same when designing power supplies.
[1] prefilter
[2] prefilter some more
[3] adjust raw volts if need be
[4] stick in a pre-reg reservoir
[5] pass thru a regulator
and possibly
[6] follow on with a SMALL reservoir to give better instantaneous low impedance power.
Never varies.
GoatGuy
Take your SMPS output (32 volts).
Run it thru a small ferrite core having 32 loops wound onto it.
THEN, Run it in series thru an 8 Ω resistor.
Then Into a 1000 μF cap.
THEN to the chain of zeners
And another 1000 μF cap.
THEN to the low-dropout linear regulator.
The loop inductor gets rid of the RF chop.
The 8 Ω resistor is part of the capacitor HI-cut filter.
The 1,000 μF cap totally sucks up almost all AF chop.
The Zeners drop 20 volts.
The next cap is the "keep things steady" reservoir.
The LDO regulator sets a constant output voltage.
And there you are. Maybe what I take as “obvious” isn't. Whatev's. The idea-chain, the design thought process is ALWAYS the same when designing power supplies.
[1] prefilter
[2] prefilter some more
[3] adjust raw volts if need be
[4] stick in a pre-reg reservoir
[5] pass thru a regulator
and possibly
[6] follow on with a SMALL reservoir to give better instantaneous low impedance power.
Never varies.
GoatGuy
Dint have much former education on power electronics.
Really appreciate the shared experience... of idea-chain, the design thought on designing power supplies. Any idea what wld the end-game ripple noise range from ultra-low 1uV ~ to moderate 10mV?
As a comparison :
This china-made $1 LM2596 BUCK DC-DC switching regulator claims to be 5mv.
Listed specs:
V-in: 4V-35V
V-out: 1.25-30V
A-out: 3A
Efficiency : 92%
Switching Freq: 150KHz
Ripple noise: 5mV --- Is the 5mv valid, maybe other hidden concerns?!?
V-in varies: ±2.5%
Dim: 48*23*14mm
Where shld the potato salad be send to? 😀
Last edited:
- Status
- Not open for further replies.