Hi Folks
Been browsing semi manufacturer's websites today and I see a lot of stuff about LDO voltage regulators. Can anyone explain in simple terms that my feeble brain can understand what the difference is between these and ordinary regulators, and would there be any sonic advantage to using them?
Been browsing semi manufacturer's websites today and I see a lot of stuff about LDO voltage regulators. Can anyone explain in simple terms that my feeble brain can understand what the difference is between these and ordinary regulators, and would there be any sonic advantage to using them?
LDO stands for Low DropOut. The dropout voltage is the minimum difference between input and output for maintenance of voltage regulation.
Non-LDO regulators typically use emitter follower output devices, which require base drive of Vbe above the output, and whatever drives the base will also drop some voltage, so the dropout voltage is usually at least 1 V. LDO regulators use common-emitter outputs. So a positive LDO uses a PNP (or in some cases a PMOS) and so the dropout voltage is essentially Vce(sat) for the device.
I don't think there is any benefit to using LDOs except in portable or battery powered devices, where efficiency is very important. Unlike emitter followers, which natually have low output impedance, and are relatively speaking easier to stabilize, and tolerant of almost any output capacitance you use, LDOs have high impedance outputs, relying only on the feedback loop inside the regulator and the output capacitor for low output impedance, and LDOs are finicky about output capacitance and often there appear in LDO datasheets tedious pages of charts of forbidden combinations of output capacitance and ESR (effective series resistance) which must be carefully avoided in order to avoid oscillation.
So I'd avoid them ...
Non-LDO regulators typically use emitter follower output devices, which require base drive of Vbe above the output, and whatever drives the base will also drop some voltage, so the dropout voltage is usually at least 1 V. LDO regulators use common-emitter outputs. So a positive LDO uses a PNP (or in some cases a PMOS) and so the dropout voltage is essentially Vce(sat) for the device.
I don't think there is any benefit to using LDOs except in portable or battery powered devices, where efficiency is very important. Unlike emitter followers, which natually have low output impedance, and are relatively speaking easier to stabilize, and tolerant of almost any output capacitance you use, LDOs have high impedance outputs, relying only on the feedback loop inside the regulator and the output capacitor for low output impedance, and LDOs are finicky about output capacitance and often there appear in LDO datasheets tedious pages of charts of forbidden combinations of output capacitance and ESR (effective series resistance) which must be carefully avoided in order to avoid oscillation.
So I'd avoid them ...
Low Drop-Out regulators require less voltage across them to work.
Eg. A 78XX regulator needs about 4 volts between it's I/P & O/P to work. LDO's need considerably less.
The first sign that a regulator is "dropping out" is when ripple starts comming through. This happens when the trough of the ripple causes there to be less than the dropout voltage across the device.
So, to make sure this doesn't happen, we make sure that there's plenty of voltage drop across the regulator. But this means HEAT.
LDO's need less drop for the same performance so less heat.
Are they sonically better? I don't know. Best look at the noise figures.
BTW Some LDO's have special shutdown circuitry within them. This can make them fail to startup if used as direct replacements.
Eg. A 78XX regulator needs about 4 volts between it's I/P & O/P to work. LDO's need considerably less.
The first sign that a regulator is "dropping out" is when ripple starts comming through. This happens when the trough of the ripple causes there to be less than the dropout voltage across the device.
So, to make sure this doesn't happen, we make sure that there's plenty of voltage drop across the regulator. But this means HEAT.
LDO's need less drop for the same performance so less heat.
Are they sonically better? I don't know. Best look at the noise figures.
BTW Some LDO's have special shutdown circuitry within them. This can make them fail to startup if used as direct replacements.
Thanks chaps, just what I was looking for🙂
It all makes perfect sense now, and should have been obvious from reading product spec. sheets, but sometimes these technical authors manage to obscure the obvious with great skill when writing...
It all makes perfect sense now, and should have been obvious from reading product spec. sheets, but sometimes these technical authors manage to obscure the obvious with great skill when writing...
low drop out regulators
With one PNP transistor as Passing Device
and 2 NPN to control the voltage
it is very easy to build a low drop regulator
I use a capacitor and some diodes to give a start up pulse,
because I want the reference to be feed by the regulated voltage.
With normal TO92 you can build a regulator with 50 mA output
and a drop out below 0.200 volt
In principal you can use Op-amp style without output buffer
or a simpler variant where reference is set by the 0.65 drop of the feedback receiving transistor
I built one for 5.00 volt which only consumes 500uA and have drop of something like 0.150 volts (25mA output) and have fairly constant output voltage precision
With one PNP transistor as Passing Device
and 2 NPN to control the voltage
it is very easy to build a low drop regulator
I use a capacitor and some diodes to give a start up pulse,
because I want the reference to be feed by the regulated voltage.
With normal TO92 you can build a regulator with 50 mA output
and a drop out below 0.200 volt
In principal you can use Op-amp style without output buffer
or a simpler variant where reference is set by the 0.65 drop of the feedback receiving transistor
I built one for 5.00 volt which only consumes 500uA and have drop of something like 0.150 volts (25mA output) and have fairly constant output voltage precision
Halo,
Thanks for the reply, but could you possibly post a rough schematic of your concept?
This would help me greatly work out the principal behind your idea, sometimes it takes a while for things to settle in my head😉 🙂
Thanks for the reply, but could you possibly post a rough schematic of your concept?
This would help me greatly work out the principal behind your idea, sometimes it takes a while for things to settle in my head😉 🙂
okay I will try to draw it
It must be perfect - or the opponents will attack me at once
I am not so good with this "Eagle.cad" I downloaded
I will get me a scanner and use pen and pencil and calculator,
like I always do.
But I have no scanner, so I try to manage with Eagle
It must be perfect - or the opponents will attack me at once
I am not so good with this "Eagle.cad" I downloaded
I will get me a scanner and use pen and pencil and calculator,
like I always do.
But I have no scanner, so I try to manage with Eagle
No rush, Halo, it's just for info, so take your time🙂
BTW, the Fun Sweden sounds good, but where is the Unfun part of Sweden?🙂
BTW, the Fun Sweden sounds good, but where is the Unfun part of Sweden?🙂
a lowdrop regulator, principally
So we have to rely more on export of all "the unfunnyness".
You can have it. We have plenty more here for you.
I am quite satisfied, it was not that hard to do. 😉
R2-resistor can be exchanged
for some LEDs and/or diodes in series to set the voltage.
R1 sets an almost constant current flowing through R2 or LEDs
We have some export, even though Ericsson mobiles are overpowered by Nokia.pinkmouse said:
So we have to rely more on export of all "the unfunnyness".
You can have it. We have plenty more here for you.
I am quite satisfied, it was not that hard to do. 😉
R2-resistor can be exchanged
for some LEDs and/or diodes in series to set the voltage.
R1 sets an almost constant current flowing through R2 or LEDs
Attachments
Halojoy,
Why doesn't that circuit oscillate?
It is 3 inverting stages with gain in a loop, with two dominant poles: the output pole (I assume there is a capacitor from the output to ground) and the pole from the Miller cap.
Looks like you have to be careful with the design so that the transmission zero from the Miller cap happens before the unity gain frequency, otherwise it will sing.
-- mirlo
Why doesn't that circuit oscillate?
It is 3 inverting stages with gain in a loop, with two dominant poles: the output pole (I assume there is a capacitor from the output to ground) and the pole from the Miller cap.
Looks like you have to be careful with the design so that the transmission zero from the Miller cap happens before the unity gain frequency, otherwise it will sing.
-- mirlo
Well, build it and see if it oscillates.mirlo said:
I guess you know how to stop it.
I usally can make it stop. If it likes to swing.
mirlo - Status
- Not open for further replies.