Voltage and Current sources

[[email protected]]

Hi,

Is anyone here familiar with any circuit's for voltage and/or current
sources programmatically controllable? Voltage ranges max or fixed of
around 500-750VDC with a power rating from 5W to 200W. Regulation and
repeatability of about 1% would be needed.

My initial approach was to use motorized potentiometer with a fixed
voltage source with feedback as this seems relatively straight
forward. This is not the approach I want to take for various reasons.

I've looked at some of the basic current source circuits but all tend
to be low power and low current. I believe it is possible to design a
larger version by using higher capacity components and/or paralleling
or serializing several smaller circuits.

Any suggestions?

 

[[email protected]]

Sounds like a job for a string of power FETS. What is the load? Ie
if its a arc, or something similar, there are more details we need to
know.

Steve

 

[[email protected]]

 Sounds like a job for a string of power FETS.   What is the load?  Ie
if its a arc, or something similar, there are more details we need to
know.

Steve

Essentially it is just resistive but of course a little capacitive and
inductive. The power needs to be applied continuously. Actually in one
case it would need to supply power to a transformer for voltage
translation into a highly inductive load.

In this case something like PWM would not work or at least I believe
it would not be a good option because of certain complicating
factors.

I need to emulate a *real* voltage/current source as closely as
possible that can supply continuous current to the load. By continuous
current I mean as if the supply was ideal.

The supplies will mainly power passive components but some power may
need to siphoned off to power some integrated circuitry.

A similar application which a variable source is needed is in
computing IV characteristic curves. This is almost what I am doing but
not quite. In my case the device is actually being used in real time
and not being sampled at discrete voltage or current points. Here
using something like PWM to control the current or voltage might
introduce artifacts in the operation of the devices. This is not to
say that one might still be able to use PWM and regulate the output.

What it all boils down to is that I need a simple programmable voltage
and, separately, current source that while operating at it's
programmed state acts closely, as close as any basic power source, to
an ideal source.

 

[[email protected]]

Could you also explain to me the reasons, for my own enlightenment,
why "arc loads" are different from any standard load?

 

[MooseFET]

Hi,

Is anyone here familiar with any circuit's for voltage and/or current
sources programmatically controllable? Voltage ranges max or fixed of
around 500-750VDC with a power rating from 5W to 200W. Regulation and
repeatability of about 1% would be needed.

My initial approach was to use motorized potentiometer with a fixed
voltage source with feedback as this seems relatively straight
forward. This is not the approach I want to take for various reasons.

I've looked at some of the basic current source circuits but all tend
to be low power and low current. I believe it is possible to design a
larger version by using higher capacity components and/or paralleling
or serializing several smaller circuits.

Any suggestions?

Do you need one or are you going into production?

A programmable power supply can be bought.

 

[neon]

You first have to understand a current source inpedance and voltage source inpedance. define that for me I will help. Finaly there is no absolute current sources or voltage sources.

 

[[email protected]]

Do you need one or are you going into production?

A programmable power supply can be bought.

yes, but I need a circuit so I can use it in my for my own
application. It most likely won't go into production but that isn't
the point.

I could use a triac to cut down the power but I'd need separate filter
caps for each "sub-supply". It would also have very low regulation and
high inrush currents when in low power mode.

AFAIK there are many ways to do "it" but I can't seem to find a way to
do it decently. Maybe some switched mode power design would work?

 

[Jon Kirwan]

A similar application which a variable source is needed is in
computing IV characteristic curves.

You want a general purpose semiconductor pin driver for test
equipment?

Jon

 

[Fred_Bartoli]

You want a general purpose semiconductor pin driver for test
equipment?

Hey, that's an idea: a pin drivers for power toobs...

 

[Jon Kirwan]

Hey, that's an idea: a pin drivers for power toobs...

4CX1000A? hehe. (I think I still have one or two in a box,
somewhere.)

Jon

 

[[email protected]]

You want a general purpose semiconductor pin driver for test
equipment?

Jon

No, it was only to compare the requirements for something that is
easily conceivable without me having to explain all the details of
what I'm trying to do. i.e., if I had a somewhat simple circuit that
solved the problem I mentioned about the IV characteristics then it
would also solve my real problem except for the caveat I mentioned.

All I need is a voltage controlled voltage source and a voltage
controlled current source for the voltage range.

e.g., http://www.daycounter.com/Circuits/Voltage-Servo/Voltage-Servo.phtml

Those circuits may or may not work for my application. I suppose I
have to do some experimentation. Unfortunately I haven't seen many
mosfets with a very large DS breakdown(about 1k).

I was hoping for a bit more circuits to look at besides using a fet in
linear mode. I guess I should try it out though as it does seem the
simplest I have found. Ultimately I need to use feedback so I can set
the voltage digitally. Maybe I can do that with a simple
microprocessor feedback loop but maybe there is some package that can
take care of all that for me? i.e., buffer's the fet input for drive
and has an input for selecting output voltage to be proportional to
supply voltage?

 

[Jon Kirwan]

No, it was only to compare the requirements for something that is
easily conceivable without me having to explain all the details of
what I'm trying to do. i.e., if I had a somewhat simple circuit that
solved the problem I mentioned about the IV characteristics then it
would also solve my real problem except for the caveat I mentioned.

All I need is a voltage controlled voltage source and a voltage
controlled current source for the voltage range.

e.g., http://www.daycounter.com/Circuits/Voltage-Servo/Voltage-Servo.phtml

Those circuits may or may not work for my application. I suppose I
have to do some experimentation. Unfortunately I haven't seen many
mosfets with a very large DS breakdown(about 1k).

I was hoping for a bit more circuits to look at besides using a fet in
linear mode. I guess I should try it out though as it does seem the
simplest I have found. Ultimately I need to use feedback so I can set
the voltage digitally. Maybe I can do that with a simple
microprocessor feedback loop but maybe there is some package that can
take care of all that for me? i.e., buffer's the fet input for drive
and has an input for selecting output voltage to be proportional to
supply voltage?

Winfield Hill and Paul Horowitz are authors of a book called "The Art
of Electronics." If you look in the 2nd edition (I think that's 1989,
or so), there is a nice section on "Special-Purpose Power Supply
Circuits," starting on page 368 in chapter 6. First thing talked
about is high voltage regulators and in figure 6.48 they talk about
regulating the ground return as one example in the ballpark of the
voltage you are looking at. Another is optoisolation, examples shown
in figure 6.50. Some of the basic ideas are in there, but a complete
circuit for your needs probably isn't.

Jon

 

[JeffM]

It most likely won't go into production but that isn't the point.

It sounds like you are making the classic mistake
(read: time=money) of building (a one-of) something you can buy.

If this is a short term need, it's even crazier.
Buying a programmable power supply (perhaps used)
then reselling it when your need has passed
would likely be essentially zero cost.

 

[[email protected]]

Winfield Hill and Paul Horowitz are authors of a book called "The Art
of Electronics."  If you look in the 2nd edition (I think that's 1989,
or so), there is a nice section on "Special-Purpose Power Supply
Circuits," starting on page 368 in chapter 6.  First thing talked
about is high voltage regulators and in figure 6.48 they talk about
regulating the ground return as one example in the ballpark of the
voltage you are looking at.  Another is optoisolation, examples shown
in figure 6.50.  Some of the basic ideas are in there, but a complete
circuit for your needs probably isn't.

Jon

I'll take a look at it. I have designed a simple feedback loop using
opamps that compares the voltage scaled down with a reference voltage
and attempts to match it by using feedback and a mosfet. I'll have to
play around with it some more and look at more examples to see if it's
a good idea. The difficult part is controlling the high side mosfet of
several hundred volts from a low side driver. I'll have to dust off
AOE and read it. Wasn't a new edition suppose to come out shortly?

 

[[email protected]]

It sounds like you are making the classic mistake
(read: time=money) of building (a one-of) something you can buy.

Let me decide that. It's my time and money so I should be the one who
makes that choice.

If this is a short term need, it's even crazier.
Buying a programmable power supply  (perhaps used)
then reselling it when your need has passed
would likely be essentially zero cost.

Trust me that a programmable supply, which I assume you mean some
thing like a large bulky bench supply and not some IC, is out of the
question. It totally defeats the purpose of what I'm trying to do.
This is a side project with part learning and part utility. If it
performs extremely well then it could be marketed so I don't want to
limit it's capabilities just to save a few bucks and hours in the
short term.

Now if there are small modules/IC's that do exactly that then it
definitely would be something to look into. In fact this would be
preferred rather than implementing it discretely assuming it isn't too
costly.

 

[neon]

If you want to learn about voltage-current sources the easy thing is look up lm317 regulators. Basicaly is a 1.2v regulator that can be designed to do both.

 

[JosephKK]

Let me decide that. It's my time and money so I should be the one who
makes that choice.


Trust me that a programmable supply, which I assume you mean some
thing like a large bulky bench supply and not some IC, is out of the
question. It totally defeats the purpose of what I'm trying to do.
This is a side project with part learning and part utility. If it
performs extremely well then it could be marketed so I don't want to
limit it's capabilities just to save a few bucks and hours in the
short term.

Now if there are small modules/IC's that do exactly that then it
definitely would be something to look into. In fact this would be
preferred rather than implementing it discretely assuming it isn't too
costly.

I think if it were me i would start with a 24 to 36 V feed that into a
two transistor inverter then to a transformer followed by a voltage
multiplier; finally a skimming regulator and current limiter. Best
efficiency, all reasonable parts, add fast shutdown for load faults
and all that is good.

 

[Winfield Hill]

I'll take a look at it. I have designed a simple feedback loop using
opamps that compares the voltage scaled down with a reference voltage
and attempts to match it by using feedback and a mosfet. I'll have to
play around with it some more and look at more examples to see if it's
a good idea. The difficult part is controlling the high side mosfet of
several hundred volts from a low side driver.  I'll have to dust off
AOE and read it. Wasn't a new edition suppose to come out shortly?

Actually, we do have several examples of exactly what Jon is
talking about. A voltage-controlled voltage-source, VCVS, is
simple enough: Use an amplifier. Voltage-controlled current
sources, or VCCS, as spice calls them, or transconductance
amps, come in several forms. The Howland circuit (AoE 2nd
pg 182) is often mentioned in textbooks. It's not one of my
favorites, but just yesterday I came across one in Keithley's
642 electrometer, where it does a great job of adding an input
offset voltage to a discrete MOSFET follower opamp stage.

I'm more fond of circuits like AoE fig 4.11, page 181, where an
intrinsically high-impedance device configuration (BJT collector
or MOSFET drain) is used for the output. These are easy to
setup in a bipolar form, if needed.

You can find many different forms of these circuits discussed
in detail on s.e.d in the Google archives, if you look.
http://groups.google.com/advanced_search

I'm especially fond of a circuit that I developed in collaboration
with Tony Williams, an old s.e.d. regular, R.I.P., such as here,
http://groups.google.com/group/sci.electronics.design/msg/50c27d2be1a63489

You can download my RIS-496 programmable CS circuit, see,
ftp://ftp.rowland.org/pub/hill/RIS-496-1.pdf
This circuit has been discussed several times on s.e.d, like
http://groups.google.com/group/sci.electronics.design/
browse_frm/thread/3ea0f2abab62703a/37b18ceb992e5b69

I recently upgraded the circuit to use power MOSFETs instead
of Darlington transistors (they have serious second-breakdown
problems, severely limiting their high-voltage power capability).
The new circuit is called an AMP-671, and its output features
a pair of opamp-controlled MOSFET current sources. These
circuits can have a voltage-control stage added, so they're a
programmed current source until a voltage limit is reached.

Win