Maker Pro
Maker Pro

5V to 5V regulators

P

panfilero

Jan 1, 1970
0
I have 5V from a regulator that's not very well regulated that I would liketo feed into a LDO regulator to clean it up a bit.... but I'm not sure howI can do this since I would have to be consistently above 5V in order for the LDO to do its thing.... does anyone know a good way to do this?

much thanks
 
G

George Herold

Jan 1, 1970
0
What does "...not very well regulated" mean?  Does it mean poor
stability of output voltage value; or does it mean that ripple and
noise is getting thru the regulator?

If he can afford a diode drop, then a capacitor multiplier will clean
up the noise.
(Assuming not too much current draw.)
And what is the headroom for the existing 5V regulator?  Maybe just
replace it with a modern unit?

Yup cascading regulators seems a bit silly.

George H.
 
G

George Herold

Jan 1, 1970
0
I kinda gathered that the one he has may dip below 5V -- but, I'm not
looking at his meter.




It's not uncommon to use a switcher into an LDO to get a really clean
supply.

OK, that makes sense.

Though I'm not sure I'd call the output from some LDO's 'really
clean'.

I made this 'low noise' power supply that's an LDO regulator followed
by a cap multiplier.
(Sorta ~2-3nV/rtHz at moderate currents (100mA) (Noise almost all from
the LDO.) I later needed a higher current (low noise) power supply
for the same project (~0.4A).
So I used a voltage reference into an opamp->transistor pass element-
cap multiplier. And it's much less noisy <1nV/rtHz. I wished I
hadn't used the LDO, but it was too late by then.

George H.
 
P

panfilero

Jan 1, 1970
0
I have 5V from a regulator that's not very well regulated that I would like to feed into a LDO regulator to clean it up a bit.... but I'm not sure how I can do this since I would have to be consistently above 5V in order for the LDO to do its thing.... does anyone know a good way to do this?



much thanks

I can't touch the first regulator, it's regulation can be off by 8%, I can only add something to it's output, the cap option sounds interesting I'm going to look into that

thanks!
 
P

panfilero

Jan 1, 1970
0
I have 5V from a regulator that's not very well regulated that I would like to feed into a LDO regulator to clean it up a bit.... but I'm not sure how I can do this since I would have to be consistently above 5V in order for the LDO to do its thing.... does anyone know a good way to do this?



much thanks

If I'm able to get my first regulator up to 5.5V is that good enough for anLDO to clean it up to 5V or am I too close to 5V? My load is pretty light.... 13mA or so... I've found some LDOs that have a dropout voltage of less than 100mV

I'm thinking this part lp2950

http://www.ti.com/lit/ds/symlink/lp2950-n.pdf
 
Regulate raw voltage into about the same clean voltage is pretty common
problem. Neat solution to that is one inductor convertor with switches on
both sides of the inductor. So it could work in buck or boost mode depending
on the input/output  voltage relation.
The only IC of this kind that I know is LTC3789; expensive and requires many
extra components.
I wonder why this good idea is not used very widely.

Vladimir Vassilevsky
DSP and Mixed Signal Consultantwww.abvolt.com


maybe one of the numerous boost converters meant for LiPo batteries
something like a tps62100, add a diode or two to keep the Vin
under the needed Vout

-Lasse
 
G

George Herold

Jan 1, 1970
0
I can't touch the first regulator, it's regulation can be off by 8%, I can only add something to it's output, the cap option sounds interesting I'm going to look into that

thanks!

panfilero, Is your problem that the 5V is not constant, it droops
under load? Or is it the noise on the 5V line? The cap multiplier
will do nothing for the droop.. in fact it might make it worse.
(Higher output impedance.)

George H.
 
J

Joerg

Jan 1, 1970
0
panfilero said:
I can't touch the first regulator, it's regulation can be off by 8%,
I can only add something to it's output, the cap option sounds
interesting I'm going to look into that

Easiest might be John's 2nd suggestion. A simple boost conversion that
adds 20% or so to the voltage, then a linear regulator to make a clean
5.00V. If it doesn't have to be so clean use only the step-up switcher.
If necessary with a Schottky up front to drop 400-600mV in case there
are phases where the voltage coming in is too high.
 
Easiest might be John's 2nd suggestion. A simple boost conversion that

adds 20% or so to the voltage, then a linear regulator to make a clean

5.00V. If it doesn't have to be so clean use only the step-up switcher.

If necessary with a Schottky up front to drop 400-600mV in case there

are phases where the voltage coming in is too high.

You mights as well suggest he improve the stabilzation control algorithm of SpaceX.
 
R

rickman

Jan 1, 1970
0
Though I'm not sure I'd call the output from some LDO's 'really
clean'.

I made this 'low noise' power supply that's an LDO regulator followed
by a cap multiplier.
(Sorta ~2-3nV/rtHz at moderate currents (100mA) (Noise almost all from
the LDO.) I later needed a higher current (low noise) power supply
for the same project (~0.4A).
So I used a voltage reference into an opamp->transistor pass element-
hadn't used the LDO, but it was too late by then.

I'm not following what you are describing. Can you explain in a little
more detail? I guess I'm not getting how you use a cap multiplier...
oh! I guess you are saying a cap multiplier rather than a cap *voltage*
multiplier! Is it that difficult to just use a large cap rather than a
cap multiplier? Or do I completely not understand what you are saying?

Rick
 
R

rickman

Jan 1, 1970
0
It's not uncommon to use a switcher into an LDO to get a really clean
supply.

I'm not sure that works so well. Most LDO's have noise rejection that
rolls off somewhere in the 10's to 100's of kHz which is typically below
the noise range of the switcher. So they don't really do much at all to
clean up switching noise. Much better to minimize the noise in the
first place with a proper layout, etc.

Rick
 
R

rickman

Jan 1, 1970
0
The problem I can see in the spec of the LP2950 is that the datasheet
gives a dropout voltage of 150mV at 100uA and 600mV at 100mA. The OP's
last post suggest he may have 500mV headroom and a load of 13mA, the
datasheet doesn't guarantee this performance. It depends if this a
one-off product or large-scale production. If the latter I would avoid
the LP2950.

TI used to make a switched cap regulator that was pretty sweet. It was
good up to 100 mA or so and regulated the switches like an LDO to
maintain a given output voltage. But I don't think they make it anymore.

Rick
 
T

Tim Williams

Jan 1, 1970
0
Cap multiplier: an emitter follower where the base is supplied by
RC-filtered power. Effectively, the capacitance as seen from the emitter
is C 1* hFE, which improves HF performance considerably. It also acts as
a reversed grounded-base amplifier, so the C-E isolation is large (only
Early effect working into the equivalent emitter impedance plus additional
bypass).

I recall Phil Hobbs had an application that required two or three stages.
Nice part about that is, subsequent stages can be biased from the
already-filtered base supply, which means an additional Vbe + Vce(sat) is
not incurred, only another Vce(sat).

Tim
 
G

George Herold

Jan 1, 1970
0
I'm not following what you are describing.  Can you explain in a little
more detail?  I guess I'm not getting how you use a cap multiplier...
oh!  I guess you are saying a cap multiplier rather than a cap *voltage*
multiplier!  Is it that difficult to just use a large cap rather than a
cap multiplier?  Or do I completely not understand what you are saying?

Rick

Oh Sorry, the cap multiplier is just a RC and transistor
circuit fragment,
http://en.wikipedia.org/wiki/Capacitance_multiplier
(R2 is part of the load.)
I learned it from Phil H.'s book.... (For years I called it a Hobbs
filter.)

George H.
 
P

panfilero

Jan 1, 1970
0
I have 5V from a regulator that's not very well regulated that I would like to feed into a LDO regulator to clean it up a bit.... but I'm not sure how I can do this since I would have to be consistently above 5V in order for the LDO to do its thing.... does anyone know a good way to do this?



much thanks

I've found a part that appears to reccommend a way of doing what I'm tryingto do, but I'm confused...

the part is DCP0105

the datasheet is here

http://www.ti.com/lit/ds/symlink/dcp0105.pdf

on page 16 it starts talking about using an LDO after the 5V to 5V isolator

"It is possible to post regulate the 5VOUT DCP0105 and still guarantee a minimum VOUT of 4.75V. This still gives the benefits of isolation in reducingthe power supply noise to 5V digital circuitry."

I don't understand what they mean... guarantee Vout of 4.75V? That's terrible... that's 0.25V away from 5V.... are they talking about something else?I thought adding a post regulator requires a higher votlage than 5V at the input of the regulator and that the point is to have the 5V LDO provide anice clean 5V at its output.... will this do that?

thanks!
 
P

panfilero

Jan 1, 1970
0
On Tue, 15 Jan 2013 13:25:23 -0800 (PST), panfilero








You could buy a little CUI/Murata SIP dc/dc converter, say 5 to 12

volts, and regulate after that. Or get a 5 to 3.3 dc/dc and stack the

output on the input to get 8.3, and linear regulate that. The SIPs are

cheap and easy to use.



Or make a charge pump booster, cheap in production and more fun.





--



John Larkin Highland Technology, Inc



jlarkin at highlandtechnology dot com

http://www.highlandtechnology.com



Precision electronic instrumentation

Picosecond-resolution Digital Delay and Pulse generators

Custom laser drivers and controllers

Photonics and fiberoptic TTL data links

VME thermocouple, LVDT, synchro acquisition and simulation

much thanks for all the replies, this is the solution I'm going with, I'm going to end up just boosting up the voltage and then regulating after that
 
J

Jamie

Jan 1, 1970
0
John said:
Yeah, like a polyphase buck switcher.

(Which I invented. As did lots of other people.)
Where I come from, there is only one inventor and
many copy cats.

Of course, if you talk to a lawyer, they may have a different
view of things.:)
Jamie
 
Top