3-pin Variable regulator with low ref voltage?

[Lostgallifreyan]

An LM317 uses 1.25V as reference, which in a current regulator circuit puts
that voltage plus the pass transistor's own voltage drop as the total drop,
which is quite large. I looked up low dropout regulators (Wikipedia), and saw
that the same 1.25V bandgap reference was generally used. If that could also
be lowered, perhaps to a tenth or even to as low as 50 mV via a resistor
divider onboard, as well as using a FET as pass transistor, total drop in
current regulators could be lowered further than usual. 50 mV seems small,
but I think many newer PSU and LED driver circuits with small sense
resistances imply that accuracy can still be had with low sense voltages.

Is a very low reference voltage used in 3-pin variable voltage regulators? If
so, any specific devices I should look at, and if not, why not?

 

[Lostgallifreyan]

Lostgallifreyan <no-one@nowhere.net> wrote in
news:Xns9B57C58939235zoodlewurdle@216.196.109.145:

An LM317 uses 1.25V as reference, which in a current regulator circuit
puts that voltage plus the pass transistor's own voltage drop as the
total drop, which is quite large. I looked up low dropout regulators
(Wikipedia), and saw that the same 1.25V bandgap reference was generally
used. If that could also be lowered, perhaps to a tenth or even to as
low as 50 mV via a resistor divider onboard, as well as using a FET as
pass transistor, total drop in current regulators could be lowered
further than usual. 50 mV seems small, but I think many newer PSU and
LED driver circuits with small sense resistances imply that accuracy can
still be had with low sense voltages.

Is a very low reference voltage used in 3-pin variable voltage
regulators? If so, any specific devices I should look at, and if not,
why not?

Correction to previous post: I think that if the total dropout is to be much
less than 1.25V, a 1.25V bandgap reference, onboard divider or not, isn't
going to work. But if the 50 mV reference can be derived some other way,
the main question still stands.

 

[Spehro Pefhany]

On Sat, 15 Nov 2008 13:25:08 -0600, the renowned Lostgallifreyan
<no-one@nowhere.net> wrote:

An LM317 uses 1.25V as reference, which in a current regulator circuit puts
that voltage plus the pass transistor's own voltage drop as the total drop,
which is quite large. I looked up low dropout regulators (Wikipedia), and saw
that the same 1.25V bandgap reference was generally used. If that could also
be lowered, perhaps to a tenth or even to as low as 50 mV via a resistor
divider onboard, as well as using a FET as pass transistor, total drop in
current regulators could be lowered further than usual. 50 mV seems small,
but I think many newer PSU and LED driver circuits with small sense
resistances imply that accuracy can still be had with low sense voltages.

You can have excellent accuracy with a very low sense voltage, but
it's difficult to make an analog circuit work well from less than a
volt or so. So, that means that *two terminal* current regulators will
tend to drop a lot of voltage.

Is a very low reference voltage used in 3-pin variable voltage regulators? If
so, any specific devices I should look at, and if not, why not?

Look at solutions that are not two-terminal if you want low voltage
drop. A simple op-amp/transistor/reference circuit can get you good
accuracy at 10-100mV voltage drop.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

 

[Lostgallifreyan]

Spehro Pefhany <speffSNIP@interlogDOTyou.knowwhat> wrote in
news:dl9uh4tua498iklk2dis472a1v2h3eop71@4ax.com:

Is a very low reference voltage used in 3-pin variable voltage
regulators? If so, any specific devices I should look at, and if not,
why not?

Look at solutions that are not two-terminal if you want low voltage
drop. A simple op-amp/transistor/reference circuit can get you good
accuracy at 10-100mV voltage drop.

Interesting. My own solution is similarly not two-terminal, though it does
adapt an LM317 current regulator circuit. It uses ground as a source of
relative negative voltage to dupe the regulator into seeking its 1.25V with
only part of that across the sense resistor. It's probably very similar to
your idea because the three main elements are the same, (plus a few
resistances). It makes a variable current regulator but I haven't yet adapted
it to try the lower volts on the sense resistor. I predict it will work well
if I only want to drop the Vf to around half of standard 1.25V, but I won't
know how much better ratios can be had till I spend time trying that (and a
protection against wiper failure in the pot, which connects to the Adj pin,
so must not go open). If I can get 10 mV sense voltage and open-wiper
protection with the modified LM317 circuit I'll be happy. Though I'll maybe
be even happier if it works with an adjustable LDO regulator too.