LM337 Voltage Regulator

[otpowell]

Hi all, I bought an LM337 and I was under the impression that I could input a positive DC voltage and get a negative DC voltage out.  I hooked up this circuit many times but always get a positive voltage out.  I'm hooking the circuit up exactly like the schematic and made sure my capacitor polarities are correct.  Am I using this wrong?  Am I really supposed to be feeding a negative DC voltage in?  That's what the data sheet illustrates.  I guess I'll have to use a bridge rectifier and feed an AC voltage in to get a negative DC voltage out.. 





O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

[kjs]

this is a linear regulator and not a switchmode inverting one.
You input a negative voltage and it outputs a regulated negative voltage.

[Alex Eisenhut]

Yes, absolutely wrong. It's a linear regulator, it can't do anything about the direction of the current. And your capacitors will surely fail if they're reversed electrolytics.

What's a "beige" rectifier?

[cpt.charlie]

Note that the schematic says -Vin not Vin it can't change the polarity.

[AG6QR]

Others have it correct, as the datasheet says.  It may be worth pointing out that the LM337 is exactly analogous to the more popular LM317.  The LM317 takes a positive voltage as input and outputs a lower magnitude, but regulated, positive voltage.  The LM337 takes a negative voltage as input and outputs a lower magnitude, but regulated, negative voltage.

Both of them behave as though they are "smart" variable resistors between Vin and Vout, as though there's a magical elf inside watching a voltmeter and quickly dialing in just the right amount of resistance to drop the voltage to exactly the desired value.  I don't think they really put elves inside, but the point is that, as a linear device, the current in is equal to the current out (with a small correction factor due to the current through the ADJ pin), and there is no voltage boosting nor polarity reversing.

[otpowell]

this is a linear regulator and not a switchmode inverting one.
You input a negative voltage and it outputs a regulated negative voltage.

Great,  thank you!  I'll get a switch mode regulator. !


O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

[otpowell]

Note that the schematic says -Vin not Vin it can't change the polarity.

I noticed this later, I'm not sure why I thought I could input a positive voltage and get a negative out which he data sheet clearly states.  Thanks for the info! 


O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

[otpowell]

Yes, absolutely wrong. It's a linear regulator, it can't do anything about the direction of the current. And your capacitors will surely fail if they're reversed electrolytics.

What's a "beige" rectifier?

I meant bridge rectifier.  I found this circuit where
I can use AC and get negative DC out, or so I think. 


O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

[otpowell]

Others have it correct, as the datasheet says.  It may be worth pointing out that the LM337 is exactly analogous to the more popular LM317.  The LM317 takes a positive voltage as input and outputs a lower magnitude, but regulated, positive voltage.  The LM337 takes a negative voltage as input and outputs a lower magnitude, but regulated, negative voltage.

Both of them behave as though they are "smart" variable resistors between Vin and Vout, as though there's a magical elf inside watching a voltmeter and quickly dialing in just the right amount of resistance to drop the voltage to exactly the desired value.  I don't think they really put elves inside, but the point is that, as a linear device, the current in is equal to the current out (with a small correction factor due to the current through the ADJ pin), and there is no voltage boosting nor polarity reversing.

Thank you very much!  I actually just learned that in class about linear devices.  Guess I should go back and read that chapter a bit.  Thanks guys I really appreciate it. 


O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

[Alex Eisenhut]

Yes, absolutely wrong. It's a linear regulator, it can't do anything about the direction of the current. And your capacitors will surely fail if they're reversed electrolytics.

What's a "beige" rectifier?

I meant bridge rectifier.  I found this circuit where
I can use AC and get negative DC out, or so I think. 


O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

Oh of course, bridge. Well, you could do it with just one diode too, it depends on what you need.

[paulie]

What's a "beige" rectifier?

Much lower Vf compared to teal. Chartreuse are worst of all in that respect.

[otpowell]

Well guys, what's really the easiest way to get a negative voltage?  By using AC, a rectifier, linear regulator and some caps and resistors or the voltage divider method, or using a inverting regulator?   Which method uses the least amount of components?   I'm surprised that not many DC power supplies provide a negative voltage.  I know if you switch your polarities with a positive voltage the multimeter will read negative but that isn't the same as a negative voltage supply, is it? 

[cpt.charlie]

The polarity depends of what point you consider as ground, that's all. The best method to get a negative voltage depends of your requirements (current, voltage...)

[paulie]

Well guys, what's really the easiest way to get a negative voltage?

By far the easiest and simplest with an AC source is a single diode and a single cap. "Bridge rectifier" and "voltage divider" are not capable of generating negative voltage. You should also be aware most internet suggestions will involve the most expensive and complicated method, not the simplest.

[otpowell]

The polarity depends of what point you consider as ground, that's all. The best method to get a negative voltage depends of your requirements (current, voltage...)

Mostly using it for labs and biasing op amps.  Usually just minimal current. 


O.T.
Electrical & Electronics Engineering
Sent from my iPhone using Tapatalk

[Kalvin]

Well guys, what's really the easiest way to get a negative voltage?

I guess the easiest way is to create virtual ground between the positive and negative rails. This is quite handy if the current requirements are modest or the efficiency is of no concern.

With a transformer you can create positive and negative rails quite easity either with center tapped transformer or using a half-wave rectifier circuit. If the negative voltage requires only modest current, a simple negative charge pump using diodes and capacitor may do the trick.

In microcontroller systems one may even use a PWM output of an MCU to create needed negative voltage for op amp circuit from the positive voltage.

[otpowell]

Well guys, what's really the easiest way to get a negative voltage?

I guess the easiest way is to create virtual ground between the positive and negative rails. This is quite handy if the current requirements are modest or the efficiency is of no concern.

With a transformer you can create positive and negative rails quite easity either with center tapped transformer or using a half-wave rectifier circuit. If the negative voltage requires only modest current, a simple negative charge pump using diodes and capacitor may do the trick.

In microcontroller systems one may even use a PWM output of an MCU to create needed negative voltage for op amp circuit from the positive voltage.

May I ask, How would one use PWM to create a negative voltage?  I have an Arduino and a UDOO, so I have PWM capability.  I also have a function generator that supplies up to -10 volts.  I would need to know how to amplify that -10 in case I need -15 for an op amp circuit. 

[Kalvin]

Well guys, what's really the easiest way to get a negative voltage?

I guess the easiest way is to create virtual ground between the positive and negative rails. This is quite handy if the current requirements are modest or the efficiency is of no concern.

With a transformer you can create positive and negative rails quite easity either with center tapped transformer or using a half-wave rectifier circuit. If the negative voltage requires only modest current, a simple negative charge pump using diodes and capacitor may do the trick.

In microcontroller systems one may even use a PWM output of an MCU to create needed negative voltage for op amp circuit from the positive voltage.

May I ask, How would one use PWM to create a negative voltage?  I have an Arduino and a UDOO, so I have PWM capability.  I also have a function generator that supplies up to -10 volts.  I would need to know how to amplify that -10 in case I need -15 for an op amp circuit.

You can create a simple negative (or positive) charge pump using capacitors and diodes. And, using the MCU's A/D-input to measure the output voltage, you can even regulate the output by varying the PWM duty cycle.

With a single stage, you should be able to get -4V from the +5V PWM output. With a dual stage, you should get -5V quite easily. Best of all, you can hook the charge pump to the -10V power supply and get the -15V of few milliamps.

Here's a few circuits using 555 to generate positive an negative voltages, but the idea is similar when using Arduino:
http://www.eleccircuit.com/the-many-dc-to-dc-converters-using-ic-555/

See this Dave's video (dealing with the positive voltage pump, but it is applicable also to the negative charge pumps as shown in the 555 circuit collection) for further explanation:


Here's Dave's more fundamental video talking about the practical limitations:

[otpowell]

Well guys, what's really the easiest way to get a negative voltage?

I guess the easiest way is to create virtual ground between the positive and negative rails. This is quite handy if the current requirements are modest or the efficiency is of no concern.

With a transformer you can create positive and negative rails quite easity either with center tapped transformer or using a half-wave rectifier circuit. If the negative voltage requires only modest current, a simple negative charge pump using diodes and capacitor may do the trick.

In microcontroller systems one may even use a PWM output of an MCU to create needed negative voltage for op amp circuit from the positive voltage.

May I ask, How would one use PWM to create a negative voltage?  I have an Arduino and a UDOO, so I have PWM capability.  I also have a function generator that supplies up to -10 volts.  I would need to know how to amplify that -10 in case I need -15 for an op amp circuit.

You can create a simple negative (or positive) charge pump using capacitors and diodes. And, using the MCU's A/D-input to measure the output voltage, you can even regulate the output by varying the PWM duty cycle.

With a single stage, you should be able to get -4V from the +5V PWM output. With a dual stage, you should get -5V quite easily. Best of all, you can hook the charge pump to the -10V power supply and get the -15V of few milliamps.

Here's a few circuits using 555 to generate positive an negative voltages, but the idea is similar when using Arduino:
http://www.eleccircuit.com/the-many-dc-to-dc-converters-using-ic-555/

See this Dave's video (dealing with the positive voltage pump, but it is applicable also to the negative charge pumps as shown in the 555 circuit collection) for further explanation:


Here's Dave's more fundamental video talking about the practical limitations:

That's awesome, thank you so much! I have a bunch of the 556 chips too! 

[IanB]

Well guys, what's really the easiest way to get a negative voltage?  By using AC, a rectifier, linear regulator and some caps and resistors or the voltage divider method, or using a inverting regulator?   Which method uses the least amount of components?   I'm surprised that not many DC power supplies provide a negative voltage.  I know if you switch your polarities with a positive voltage the multimeter will read negative but that isn't the same as a negative voltage supply, is it?

There's no such thing as a "negative" voltage. That's why you don't find any DC power supplies that output a negative voltage (and also why you don't find any left-handed screwdrivers).

Voltage is measured (and defined) as the difference between two potentials. If you have two potentials "A" and "B", then "A - B" has one sign, and "B - A" has the opposite sign. The difference has the same magnitude either way.

It just so happens that for convenience people often choose to assign a value of 0 to B and reference everything to that. If you have a voltage:

    V = A - B

Then by assuming B = 0 you have:

    V = A - 0 = A

This may lead you to think that "A" is a voltage, but it is not really. A voltage V is still a difference from some reference and you have to keep straight what your reference is.

Always just remember that in circuits, voltages are differences and currents are flows.

[otpowell]

Well guys, what's really the easiest way to get a negative voltage?  By using AC, a rectifier, linear regulator and some caps and resistors or the voltage divider method, or using a inverting regulator?   Which method uses the least amount of components?   I'm surprised that not many DC power supplies provide a negative voltage.  I know if you switch your polarities with a positive voltage the multimeter will read negative but that isn't the same as a negative voltage supply, is it?

There's no such thing as a "negative" voltage. That's why you don't find any DC power supplies that output a negative voltage (and also why you don't find any left-handed screwdrivers).

Voltage is measured (and defined) as the difference between two potentials. If you have to potentials "A" and "B", then "A - B" has one sign, and "B - A" has the opposite sign. The difference has the same magnitude either way.

It just so happens that for convenience people often choose to assign a value of 0 to B and reference everything to that. If you have a voltage:

    V = A - B

Then by assuming B = 0 you have:

    V = A - 0 = A

This may lead you to think that "A" is a voltage, but it is not really. A voltage V is still a difference from some reference and you have to keep straight what your reference is.

Always just remember that in circuits, voltages are differences and currents are flows.

Thank you for your explanation.  I understand the potential between two points concept.  So what is it then you are reading when you get a reading like -15 V when referenced to zero ground?  Also, what am I to do if a op amp circuit requires +15 volts on one DC input and -15 on the other?  Would that be the same as 30 V on the positive input and zero on the negative DC input?   Thanks IanB I appreciate it. 

[IanB]

Thank you for your explanation.  I understand the potential between two points concept.  So what is it then you are reading when you get a reading like -15 V when referenced to zero ground?  Also, what am I to do if a op amp circuit requires +15 volts on one DC input and -15 on the other?  Would that be the same as 30 V on the positive input and zero on the negative DC input?   Thanks IanB I appreciate it.

Well zero ground doesn't have any real existence; it is just a choice made by you or the circuit designer when deciding where to measure voltages from. It's a bit like "putting a stake in the ground". If you put a stake in the ground and take your tape measure in one direction you might go a distance of 100 ft. If you went the same distance in the opposite direction you could call it -100 ft. But always your distances are relative to where you put the stake. You could put the stake anywhere...

(It turns out the power companies like to put a real stake in the ground and attach their transformers to it, making the earth beneath our feet their chosen ground reference. But this has little to do with signal processing and electronic circuits.)

So for the op amp, it doesn't know the difference between +15 V on one power rail and -15 V on the other. It could as well be 30 V and 0 V. This is because (+15) - (-15) = (+30) - (0).

The reason for using +/- supplies with op amps is not for the op amp, it is for your signal. Op amps are signal amplifiers, and signals are usually arranged to be small differences around a reference point. The op amp will work much better if the reference point is somewhere in the middle of its range and not out near the edges. So if you arrange a power supply as -15 -- 0 -- +15 V you get a nice convenient 0 right in the middle, which is a perfect place to put your signals. There are other ways to put a zero reference in the middle of the range, but they are not always so convenient.

[otpowell]

Thank you for your explanation.  I understand the potential between two points concept.  So what is it then you are reading when you get a reading like -15 V when referenced to zero ground?  Also, what am I to do if a op amp circuit requires +15 volts on one DC input and -15 on the other?  Would that be the same as 30 V on the positive input and zero on the negative DC input?   Thanks IanB I appreciate it.

Well zero ground doesn't have any real existence; it is just a choice made by you or the circuit designer when deciding where to measure voltages from. It's a bit like "putting a stake in the ground". If you put a stake in the ground and take your tape measure in one direction you might go a distance of 100 ft. If you went the same distance in the opposite direction you could call it -100 ft. But always your distances are relative to where you put the stake. You could put the stake anywhere...

So for the op amp, it doesn't know the difference between +15 V on one DC input and -15 V on the other. It could as well be 30 V and 0 V. This is because (+15) - (-15) = (+30) - (0).

The reason for using +/- supplies with op amps is not for the op amp, it is for your signal. Op amps are signal amplifiers, and signals are usually arranged to be small differences around a reference point. The op amp will work much better if the reference point is somewhere in the middle of its range and not somewhere near the edges. So if you arrange a power supply as -15 -- 0 -- +15 V you get a nice convenient 0 right in the middle, which is a perfect place to put your signals. There are other ways to put a zero reference in the middle of the range, but they are not always so convenient.

Ok I get it, thank you!  That was the best explanation!  So all I need is a 30v power supply, stake it and add a voltage divider and I'll have my +15 and -15.  Perfect l! 

[AG6QR]

Ok I get it, thank you!  That was the best explanation!  So all I need is a 30v power supply, stake it and add a voltage divider and I'll have my +15 and -15.  Perfect l!

If you use a voltage divider to split 30V in half, generating +15 and -15, just remember that the voltage divider can't sink or source any current.  If you try to put current into or out of a voltage divider, the voltage will change.  The voltage divider equation just comes from Ohm's law, but with the assumption that the current in the top and bottom half of the divider is equal, which it won't be if current is entering or leaving.

There are "rail splitter" ICs built specifically for this purpose -- they can source or sink current, within their limits.  A voltage divider followed by an appropriate buffer (maybe an op amp) will do the trick, and that's more-or-less what's inside a "rail splitter" IC.

[cpt.charlie]

Use the output of a linear regulator LM78xx as the circuit "ground"