Hello,

I'm sure this is a pretty frequent topic...but here's my spin on it.

Like a lot of people, I want to develop a power supply for my lab. Eventually I would like it to be a dual supply, but for now I just want to get the theory down for one.

I would like to be adjustable constant voltage and if too much current is drawn for it to change to adjustable constant current. By adjustable, I mean potentiometers setting the CC point and the CV.

The voltage regulation part is pretty bog standard. Non-iverting type op amp configuration. The constant current part I guess can be achieved through an op-amp current source.

My main question is...What can I use to determine whether CC or CV should be enabled? And then how can it switch seemlessly between them. I've been trying to decipher the schematic of Agilent E361x Power Supply for ideas but I've very confused.

Any ideas?

Thanks.

 

Study the datasheet for the LM723 chip.
Seamless crossover is not a problem in that all you have to do is collapse the reference voltage when the voltage on a low side current sense resistor crosses a threshold. This can be done as simply as a 317 chip with a current sense resistor that turns on a single base emitter junction of an npn, and its collector dumps the reference voltage on the 317 feedback loop.

More sophistocated is to develop a much more adjustable current limit by using an op-amp to compare the voltage across a current sense resistor to an adjustable voltage.

Gotcha started?

 

Here are a few links..
The LM723...
http://www.phoenixcomputerlabs.com/All-About-Mixers/LM723.html
..
An LM7812 and LM7912 Dual polarity supply...
http://www.phoenixcomputerlabs.com/All-About-Mixers/7812-7912.html
..
And a dual supply using LM317 and LM337s
http://www.phoenixcomputerlabs.com/All-About-Mixers/LM317-LM337-Dual-Polarity-Regulators.html
...
Hope this helps...

 

CC or CV doesn't really matter.

For CV, you monitor the output voltage and feed this back into the inverting input of the op-amp driving the output.

For CC, you monitor the voltage across a current sense resistor and feed this back into the same inverting input of the op-amp.

The circuit will automatically switch into CV or CV depending on which threshold is reached first. The switch over is automatic and seamless.

 

Hello,

I'm sure this is a pretty frequent topic...but here's my spin on it.

Like a lot of people, I want to develop a power supply for my lab. Eventually I would like it to be a dual supply, but for now I just want to get the theory down for one.

I would like to be adjustable constant voltage and if too much current is drawn for it to change to adjustable constant current. By adjustable, I mean potentiometers setting the CC point and the CV.

The voltage regulation part is pretty bog standard. Non-iverting type op amp configuration. The constant current part I guess can be achieved through an op-amp current source.

My main question is...What can I use to determine whether CC or CV should be enabled? And then how can it switch seemlessly between them. I've been trying to decipher the schematic of Agilent E361x Power Supply for ideas but I've very confused.

Any ideas?

Thanks.

Here you go.

 

CC or CV doesn't really matter.

For CV, you monitor the output voltage and feed this back into the inverting input of the op-amp driving the output.

For CC, you monitor the voltage across a current sense resistor and feed this back into the same inverting input of the op-amp.

The circuit will automatically switch into CV or CV depending on which threshold is reached first. The switch over is automatic and seamless.

I appreciate your answer, however I feel it has been somewhat oversimplified for someone who doesn't quite understand all the assumed nuances.

I have been searching for an elegant and simple solution, which I think you are are offering. Would you be able to point me in I direction so that I could better understand your suggestion?

 

Treat CV and CC as maximum limits because that is what they are.

When you set the voltage, you are setting the maximum voltage that the power supply must never exceed.
When you set the current, you are setting the maximum current that the power supply must never exceed.

In a CV power supply, a voltage is tapped off the output voltage (i.e. you monitor the output voltage) and is fed back into the inputs of an op-amp (or similar control circuit) to control the output VOLTAGE of the power supply.

In a CC power supply, a voltage is tapped off a current monitoring resistor (i.e. you monitor the output current) and is fed back into the inputs of the same control circuit to control the output VOLTAGE of the power supply.

No, I did not make a typo error. You are always controlling the output VOLTAGE.

The switch over from CV to CC is seamless. The control circuitry will do it for you. The power supply is in CV or CC mode depending on which limit is reached first.

If CC limit is reached, the output voltage will fall because Ohm's Law must be satisfied, i.e. V = I x R.

If CV limit is reached, the current will seek a value lower than the set limit because Ohm's Law must be satisfied, i.e. I = V/R.

Another way of putting it, if both voltage are current limits are reached simultaneously,
then the load must be R = V/I.

If the resistance of the load is greater than R as calculated above, then the power supply will be in CV mode, i.e. current limit not reached.

If the resistance of the load is less than R as calculated above, then the power supply will be in CC mode, i.e. current limit has been reached.

 

What maximum current and voltage for the supply?

 

@MrChips I am curious to see a working circuit of your description, the theory makes sense, Im' not just sure how to implement it.

Lets say we want to setup a supply that limits 10V/10A.

I've come up with a circuit using a Diff Amp and a Summing Amp:

FOR CV:
Vdiff = Vset - Vsense (with the Diff Amp)
Vdrive = Vsense + Vdiff (With the summing Opamp)

Worked Example:
10Vset - 11Vsense = -1Vdiff
11Vsense + (-1Vdiff) = 10Vdrive
This circuit will autocorrect any over/undervoltage.

FOR CC (and CV combined):
Vdrop = I / Rsense (second Diff Amp)
VdriveFinal = VdriveCV + (-Vdrop) (second Summing Amp)

Worked Example:
10A / 1Ohm = 10Vdrop
VdriveFinal = 10VdriveCV + (-10Vdrop) = 0VdriveFinal

So. If our sense resistor is 1Ohm, in theory we would get 0V at 10A. And if our load current was 0Amp, our voltage would be 10V.

If I understand you correctly, then you're saying that I should design for something more like 20V and 20A so that I could achieve 10V AND 10A simultaneously?

 

If I understand you correctly, then you're saying that I should design for something more like 20V and 20A so that I could achieve 10V AND 10A simultaneously?

Where did I say that? No, I did not say that.

If your supply is delivering 10A @ 10V then your load is 1Ω @ 100W.

Your current sense resistor does not have to be 1Ω. It can be 100mΩ.
Yes, your supply has to have some headroom, i.e. it must supply more than 10V in order to account for the voltage drop across the pass transistor and the current sense resistor (assuming we are designing a linear PSU).

You do not need 20A output if you only need 10A max.

 

If you would like to see a 30V/10A linear PSU design, take a look at the design of the late Tony van Roon (VA3AVR), circuit and full construction details:

Part 1
Part 2
Part 3
Part 4

 

What maximum current and voltage for the supply?

What he said. Besides necroposting and hijacking (both way overly harsh descriptors), what do you, SS, require for your supply to deliver?

There are tons of working schematics on the innergoogle under 'lab supply circuit'. Step 1, limit the field.

ak

 

Here you go.

Can you give more information on this cirvuit i will like to build it

 

Can you give more information on this cirvuit i will like to build it

That member was last seen here in 2014, so I doubt you will get a reply.

 

That member was last seen here in 2014, so I doubt you will get a reply.

Thats why i made a new thread

 

Sounds like he’s passed on but left us with this

http://www.tonyvanroon.com/

 

Tony's site includes a "precision power supply" based on the MC1466 regulator chip, which was specifically designed for lab power supplies (the data sheet is very informative). Those chips have long been out of production, but a couple of people have replicated them using discrete components:
http://chemelec.com/Projects/MC1466/MC1466.htm
http://tubetime.us/index.php/2014/09/28/a-miniaturized-discrete-mc1466/
It was an expensive chip when I bought some in the '80s, probably around $30 in today's money, so making one from discretes may not cost any more, and would be cheaper to fix if something goes horribly wrong.

The app note for the LM10CH includes a regulator suitable for a lab power supply; last time I checked, those chips were readily available.

If you have an Astron linear power supply lying around, those can be easily converted to variable voltage. There's a mod that allows some adjustment of current limiting, but not a real constant current mode like a lab supply. They could be a basis for building a lab supply by replacing or improving the regulator.

 

Tony's site includes a "precision power supply" based on the MC1466 regulator chip, which was specifically designed for lab power supplies (the data sheet is very informative). Those chips have long been out of production, but a couple of people have replicated them using discrete components:
http://chemelec.com/Projects/MC1466/MC1466.htm
http://tubetime.us/index.php/2014/09/28/a-miniaturized-discrete-mc1466/
It was an expensive chip when I bought some in the '80s, probably around $30 in today's money, so making one from discretes may not cost any more, and would be cheaper to fix if something goes horribly wrong.

The app note for the LM10CH includes a regulator suitable for a lab power supply; last time I checked, those chips were readily available.

If you have an Astron linear power supply lying around, those can be easily converted to variable voltage. There's a mod that allows some adjustment of current limiting, but not a real constant current mode like a lab supply. They could be a basis for building a lab supply by replacing or improving the regulator.

I will see this it looks intresting