How do you make a current limiting knob for power supplies?

[frank_p]

Hi,

I would like to build a dual tracking bench power supply with a current limiting pot and a "overload" led indicator.  I've seen many schematics on the web for a lab power supply but they do not have the current limiting characteristic.

I have one (in a big system) and find it useful.  I would like to build one so I could have a smaller appartus.  (Or at least have some insight on the topic.)

I did not see any topic on the forum about that using the SEARCH function.

Do anyone know a link where there is a schematic, some explanations or anything related to it.  Also if someone have knowledge about it or any reference, it would be great.

Thanks in advance.

F.P.

[GibsonGM]

A) fuse
B) circuit breaker
Most power supplies are just designed to provide up to a maximum current they can supply at a given voltage.  Over this value, the voltage sags, things get hot, and eventually can burn out.   I'm not sure I've ever seen one that "limits" current like you describe - if so, you'd have to find the schem, post it, and identify what components are doing the limiting. 

[frank_p]

Mike,

I think I have found some basic concept schematics here :



Ref:
http://electronics-diy.com/electronic_schematic.php?id=635

I am trying to post a photo of my device but photobucket is in problems I think.

I am looking for a variable current limiter (for flexibility) with indicator so a fuse, a breaker won't do it.

The circuits seems to be able to do it with R2 set as a pot.

You get the concept ?
I think I understand (without having calculating anything for now.) ...

Thanks

[frank_p]

An other example in a DC motor drive:

QUOTE:
Reference Voltage

The reference volta'ge for the motor speed control is obtained from a monolithic integrated circuit voltage regulator, as shown in figure 4. The integrated circuit contains a temperature compensated voltage reference of approximately 7 volts, an error amplifier, a series pass transistor, and a current limiter, as shown in figure 4A. The compensated reference is attenuated, to 5 volts by Resistors R51, Potentiometer R52, and R53, as shown in figure 4B. The attenuated reference is connected to the non-inverting input of the error amplifier. The error amplifier compares the 5 volt reference signal with the regulator output voltage. The amplified error signal controls the series pas s transistor to maintain 5 volts at the regulator's output. However, an abnormal current flow in external Resistor R54 will turn the current limiting transistor on and reduce the signal to the base of the series pass transistor. Reducing the base signal reduces the output voltage, thereby protecting the regulator against accidental short circuits or shorted logic components.





Ref:
http://www.joliettech.com/model_1681_instruction_manual.htm

[R.G.]

Good question. I have a couple of small HP bench supplies that have adjustable current limits as well as adjustable voltages. I use it all the time.

However.

It's not that easy to do a good variable current limit that you're sure will not be damaging to the power supply or the circuit. The only good way is to put a small value resistor (or Hall effect current sensor$$$) in series with the output to measure the current, then do some opamp circuits to extract that floating voltage across that resistor. That voltage is then compared to the voltage across a current limit pot, and if it's smaller, nothing happens; if it's larger, either the reference voltage or output transistor base drive, or both, are cut back.

There are quick and dirty circuits that may be good enough. Better than not having a current limit, I guess.
Here are some places to look:

http://www.powerdesigners.com/InfoWeb/design_center/Appnotes_Archive/an-19.pdf
http://www.physics.unlv.edu/~bill/PHYS483/current_lim.pdf
http://caladan.nanosoft.ca/c4/ccorner/4.php

[frank_p]

Well Thank you a lot R.G. .
I get the basic idea.  But gonna read these documents for sure.

The Power Supply I have is the modular TeKtronix PS 503 A; and as said, the limiter is pretty useful.

Thanks again,
I'll be back later on the topic.

F.H.P.

[frank_p]

Aaaaah ! 
Defensive Engineering !

Quote:

When I was a little girl, I got a pocket multimeter for xmas. It was nothing special, just a 3 digit autoranging DVM from Radio Shack. I used it for everything. And as it would happen, eventually, the button cells that powered it wore out.
...etc...

The moral of the story is that, no matter what you design, no matter who you think your target audience is, every time you leave out a protection circuit, you traumatize an 8-year-old.
-----

I still have the same DVM and it was my pride and joy when I was a kid 

[Gus]

I had a post written then my DSL stopped working.

You asked a good question and R.G. gave a good answer.  Dual tracking +- makes it a bit harder
I saved my post
Ohms law, current = voltage across the resistor / resistor value

The output impedance will be affected by the resistor so often the sensed voltage is amplified allowing a lower value resistor.

When the sensed current goes too high, the added circuits reduces the voltage, often by reducing the drive to a series pass element you could also have it light a LED if you use a adjustable comparator circuit.

The voltage control sense is "closed"  after the current limiting sense resistor.

Now what you need to note
voltage range wanted
max current
logic for the current limiting, independent limiting on both sides or logic to reduce the current on both side if one side starts to draw too much?

[George Giblet]

I've kept tabs on power supplies with *variable* current for more than 20 years and I've designed a few of my own.

You can condense the main aspects down to:

- To a large degree designing a variable current power supply requires more effort than adapting fixed current limit ideas.

- It is often hard to retrofit a variable control with a pot.  Adding a switch with a few fixed limits is easy.

- Generally need to add another opamp (which can be a transistor based circuit) to do the current sensing.  This should be obvious since after all you are adding another feedback loop.

- The current loop needs to override the voltage loop.  This can be done in two ways:   the current control loop can modify the reference voltage, effectively dropping the voltage to regulate the current.  Or, by making the circuit override the signals after the voltage control opamp effective making the taking the voltage control circuits out of action and then the current control circuit take over.

An example of the first case is this one,

http://www.electronics-lab.com/projects/power/003/

There's an example of the second case on the web but I can't find it. If you look at the LM723 you can see the the current transistor kills the drive of the output transistor, and the voltage control opamp before that cannot do anything about.  The voltage control is more often done by killing the output between the voltage opamp and the next stage.

- If you want to add voltage and current LEDs then that requires more effort again.  The second method I mentioned usually works better here but the example I gave shows you can get LED to with the first method.

- When you design feedback control loops you need to take a lot of care making sure the feedback loops are stable (in the sense they do not oscillate) under all conditions of voltage, current and loads.  This can be harder than you might think and adding current control loops makes this job harder.  OF course, something ignorance of this can save you a lot of work if it happens to work first go.

- The circuit like RG gave the link for do work on a two control loop principle.  The problem is the circuits that use adjustable regulators can be very touchy.  I've seen a few go up in smoke.  The problem is often caused at power-up.  At power-up the control opamps sometimes have not powered up properly and the current limit is not active for a short space of time and this can blow the circuit or the power supply itself.

http://www.powerdesigners.com/InfoWeb/design_center/Appnotes_Archive/an-19.pdf

[frank_p]

Okay, I think I am hooked on this one.
This is all is missing for a second small lab that I am putting together.
I've repaired my old oscilloscope yesterday so I have everything to make an other experimentation corner.
Your posts had been printed Gus, George and R.G. and there is already some underline in the text.
There are a few things I don't understand but I'll get it all together for having a clear(er) picture.

Thanks Gus, George and R.G.

[frank_p]

P.S. : This is the one that I have now
With also a modified electric train throttle. 

As you can see Gus (and Mike), this one have:
one knob/led for each sides (two bottom grey knobs) for Current limiting;
0 - 20V @ 1A on each sides;
Pull-knob at the top for dual tracking;
No displays for V or A;
Plus a fixed 5V@1A.



Simple but glad to have that.

[Gus]

George's post has more detail than mine and good links.  I left out a lot of stuff because what sounds like it might be a simple question opens up all kinds of issues.  I have not designed any power supplies I have built a number of them more or less from app notes BUT I have repaired a number of lamda linear and other brands Linear and smaller switching ones over the years.

I an looking for some of my manuals I have  old motos and RCA one with power supply circuits.

The stability stuff is important as posted by George.  Look at the voltage and if you have a current clamp current output of a power supply when you use a step current draw.

[petemoore]

  wrong title.

[frank_p]

  wrong title.

Would be: How to make a power supply with adjustable current limiting features ( then ? ).
Thanks for correcting my bad english.

The stability stuff is important as posted by George.  Look at the voltage and if you have a current clamp current output of a power supply when you use a step current draw.

Gus, I don't understand completely the second sentence (I think it's just the structure of the sentence).
Yes, stability may not be a small issue with these feedback loops.  You can put or design some PID but it can still be tricky to adjust it correctly before the damn thing start to get out of control and blow something.   

[Papa_lazerous]

I think mostly you have covered allot of groundon this post, but I have a Penfold book somewhere on power supplies and he covers current limiting to great extent, I shall dig it out and try and put a few examples up for you,

[frank_p]

I think mostly you have covered allot of groundon this post, but I have a Penfold book somewhere on power supplies and he covers current limiting to great extent, I shall dig it out and try and put a few examples up for you,

Thanks.
Anyway, I think this topic is full of vitamins for everyone since there is not a ton of topic about this subject on the forum. Current limiters are useful when debugging.  And everyone is debugging something on this forum.  Even if you don't build or buy one, some may learn that it can be used.

[George Giblet]

Here's a typical "professional" power supply with variable voltage and current:

http://www.eserviceinfo.com/downloadsm/35282/Mastech_3003,%203020.html

The part of the circuit on the right detects the output voltage and selects transformer taps with a relay.  This is very common and the idea is to reduce power dissipation in the power transistors. 

There's few "funny" things in the circuit that might make it hard to understand but they are common tricks.  One is the strange connection of the power to the opamps - this simplifies the current sensing.   The opamps need to operate over their common mode voltage range and this connection allows this to occur.

On the left you can see the voltage and current sensing opamps and the LEDs which eventually drive the output transistors.

It is possible to design stripped-down and simpler looking circuits which have the same functions.  However you need to choose the right opamps.  You can also trade things off get rid of all those transformer taps.  If you want to reduce power dissipation then you have to use the relay switching (or a noisy switchmode pre-regulator).  When you design lower current supplies say 1A then can trade out the relay switching because the full power dissipation isn't that bad as to require enormous heatsinks.

If you are selling power supplies for a business then the custom transformer taps aren't a real issue but at home it's a pain.

[frank_p]

Ouch !   
I have been at the library today, and read a lot to realise that I will have to read more...
I will have to pass over that schematic many times to understand all that.
But still, while I look at it now, it seems to be less abstract than what popped in my eyes when I openend the PDF.
Tomorrow will be an other learning day.
Taking the schematics module by module will be beneficial.
Thanks George.

I already had the first link that you posted on this topic.  I just forgot about it.  The description of the schematic really helped me.

I've also looked for hall effect detectors.  The simple "plate" one, inserted in a ferrite core is a viable way to sense current ?
They are not that expensive at Mouser. 

[George Giblet]

You might be able to find the HP ( Hewlett-Packard) DC Power Supply Handbook online.  It's fairly old and covers quite a few topics (some a little obsolete, like triacs and scr's).   What it does do is outline some circuit structures in an uncluttered manner.  It was a standard reference for a long time but IMHO it's not *that* great and there's better commentry on websites and in applications notes.  If you can't find it PM me and I'll have a look through my stuff - I know I have it.

[frank_p]

Yes, here:

http://www.slack.com/pdf/HP-AN90B.pdf

Thanks George.