• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Current Limiting - Without Voltage Drop and/or Power Waste (Disipation)

Status
Not open for further replies.

RichG

New Member
Hi there. I'm a mechanical engineer, so I'm struggling a bit with this particular problem...

The problem: I want to test some electrodes, but when I run the system the electrodes are effectively shorting the circuit (which is to be expected). The power-supply trips out until the load is removed - I assume because it's trying to draw too much current?

Previous Activity: I've built and tested some basic voltage and current limiting circuits, which work to an extent... The transistors successfully limit the current to 3A each (near the maximum capability of the transistor). However, due to this shorting effect, the voltage-drop is from 12V down to 3V, and the heat / power-dissipation is significant.

The aim: While the above works, my main concern is the inefficiency of this power supply method (as it happens 3V is adequate, voltage isn't critical to this) - it's wasting a lot of power, and creating a lot of heat. SO is it possible to create a power supply circuit that 1) doesn't drop the voltage when shorted (it simply continues to limit current to 3A regardless), 2) doesn't require excessive cooling to maintain such operation??

Thoughts: New current Limiting circuit, using different Transistors, or better regulator control??
Lower the current-limiting threshold of the Transistors (just use more transistors set lower)??
Using a transformer that only supplies a specific current, regardless of draw?? does this exist?

Any ideas, thoughts or suggestions would be most appreciated. Useful learning references also good :)

Thanks in advance. Rich
 

dknguyen

Well-Known Member
Most Helpful Member
You have to pulse voltage between full on/off at a certain duty cycle through an inductor. THe inductor smooths out the spikes of on-off current that would normally occur to form produce a current with the average you want (but with ripple as well). It's how switching DC-DC converters voltage regulators work more efficiently than linear regulators that burn off extra votlage as heat
 
Last edited:

dougy83

Well-Known Member
I suggest you get a >=3A SMPS buck converter IC and build up an SMPS. As you want to regulate the current, add a shunt resistance in series with the load and amplify it using an opamp to provide a feedback signal to the controller IC. Any of the National Semiconductor simple switchers with an IOut of >= 3000mA (3A) should be fine (see http://www.national.com/cat/index.cgi?i=i//206).
 

crutschow

Well-Known Member
Most Helpful Member
If you don't need the high current to continuously flow when shorted, you could use fold-back limiting which significantly reduces the current under short circuit conditions, or hiccup limiting when allows only short pulses of high current when shorted. Google "fold-back limit" or "hiccup limit" for more info.
 

dougy83

Well-Known Member
I think he's after constant current output to drive some immersed electrodes (just guessing).
 

Diver300

Well-Known Member
Most Helpful Member
RichG;833479 SO is it possible to create a power supply circuit that 1) doesn't drop the voltage when shorted (it simply continues to limit current to 3A regardless) said:
Any power supply will drop when shorted. That is almost the definition of "shorted".

The questions that need to be answered are:-

How much current do you need to test the load?
How much voltage do you need on-load and off-load?
What power supply are you running from.

If you are testing a low resistance load, it is normal to transform the voltage down a lot, which increases the current.

http://www.jmwlimited.co.uk/Datasheets/Fluke_6000_Series_PAT_Testers_Datasheet.pdf takes 60 W from a 230 V supply (1/4 A) while supplying 25 A for a current test.
 

RichG

New Member
Thanks for all your help, most appreciated.

The DC-DC converter method (full I/O pulses through Inductor), makes sense, and sounds feasible. I've been doing a little research into this, and I have to admit it's a little out of my comfort zone. Looking at the 'Buck regulator' suggestion - using an LM2596; will the circuit attached work to limit current as I require? Seems very simple, the methodology makes sense, just need to know if I'm on the right track..?

Re the 'series shunt resistance' suggestion, just checking, but you mean a current shunt resistor right? (sorry, still learning).

Also, as a short term solution, does anyone know where I cud buy a complete ready to use 3A limiting regulator device to achieve this goal? That way I can continue testing while I get things together to make my own.

Thanks again. Rich
 

Attachments

Diver300

Well-Known Member
Most Helpful Member
A buck regulator would be the best way of generating 3A at a low voltage from a higher DC voltage like 24 V. You need to adjust the feedback. Buck regulators need feedback and they are usually arranged to give a constant voltage output.
 

dougy83

Well-Known Member
Looking at the 'Buck regulator' suggestion - using an LM2596; will the circuit attached work to limit current as I require?
The current will be limited (but not regulated) as the converter has a built in current limit of 4.5A peak. I don't know what you require exactly, it may be good enough for you as is.

Re the 'series shunt resistance' suggestion, just checking, but you mean a current shunt resistor right? (sorry, still learning).
The shunt resistance allows the current to be measured (as in the attached schem) and therefore regulated

CONST CURR.gif
 

mbarazeen

Member
the safe and best way is to use a transformer and get a high current low voltage output by mdifying its winding. if you select the maximum terminal voltage required for your application, it can be the out put of the transformer. use a current limitter too.

then the voltage drop you may need to waste as heat will be very less.
 

RichG

New Member
Latest circuit I'm trying looks like this (see attached). I think it should work (but like I said, not exactly my area, but worth a try - lots and lots to learn still!).

Dougy83 - Thanks very much for the diagram! Great help! Main things I've added to it are more capacitors for ripple filtering (I'm hoping the circuit is something like right). If you get a few mins spare at some point it would be great if you could have a quick glance over the circuit and let me know what you think? The circuit you sent mostly made sense - still need to workout and understand the Opamp bit (roughly understand but going to do further research in that area for future knowledge). Also, pretty sure but is the 10K rheostat for final adjustment to the feed-back, to get current limiting bang on the desired value?

mbarazeen - Like the idea, sounds interesting. Again this is a little out of my league. However am I right in thinking; incoming power to the transformer would create a certain level of flux within the core, and by adjusting the outgoing wire / winding, there would only be so much current output available? and by using thicker wire and fewer turns the available maximum current would not exceed the new winding or transformer capability? Or would a heavy load even at low voltage, 1) cause the flux to create a resistance 'effect' and thus reduce the power output? or 2) simply cause the incoming winding to draw and handle too much power and blow? - Does the outgoing load draw have any effect on the flux?

PulseCurrentLimiterCircuitA1.gif
 

dougy83

Well-Known Member
If you get a few mins spare at some point it would be great if you could have a quick glance over the circuit and let me know what you think?
Looks very nice, unfortunately there was an error in the garbage I posted. It should be more like this image, as the shunt needs to be in the diode's current loop also (I also changed the second opamp's reference point, which won't make much difference in the end).

PulseCurrentLimiterCircuitA1.gif

Note that the shunt should be ~22milliohms (not microohms). Value is not important; there's a potentiometer in one of the gain stages.

Make sure that the opamp is powered from the input voltage (<30VDC) and not the output, as the output voltage can go very low with the electrodes connected.

The circuit you sent mostly made sense - still need to workout and understand the Opamp bit (roughly understand but going to do further research in that area for future knowledge). Also, pretty sure but is the 10K rheostat for final adjustment to the feed-back, to get current limiting bang on the desired value?
Check the wikipedia opamp page. The first opamp circuit set up as a non-inverting amp, gain is 11 (i.e. 1 + 10k/1k). The second opamp has an adjustable gain of 1-11. So the total gain is 11-121, so for a shunt current of 3A we get 66mV at the 1st opamp input, 726mV at the output of the 1st, and 726mV - 8V at the output of the second (which is fed back into the switcher). The switcher will regulate the output such that the feedback voltage (voltage at output of 2nd opamp) is 1.2/3.3/5/12V depending on the LM2596 version you use (I'd recommend using the 5V one as it keeps the opamp voltage away from the rails).
 
Last edited:

RichG

New Member
Hello again, I'm not sure if this thread is still being watched, but quick update below....

I finally got round to making and testing the previously discussed circuit, using a BUCK converter etc. After a few further mods it worked a treat (for the original purpose). The main thing I had to change: increasing the 2nd opamp gain potentiometer, from upto 10k to upto 20k (I actually added a 10k resistor in series with the 10k pot). With only 10k the opamp output voltage was too low to influence the BUCK. ~20k allows the BUCK to limit current efficiently down to about 1.8amps. I've been running it at 2.5amps, cool & stable, and achieving nearly 90% efficiency! Fantastic! Again thanks to all for your help and ideas (special thanks to 'Dougy83', for you efforts, advice and time - most appreciated).

For my next project (more of a curiosity this one), I've decided to improve my material trials a little more, so I'm trying to work out how to control the Voltage as well! Using the above circuit, the voltage does drop quite a bit (as expected). When running at 2.5A, it's supplying around 3V. Is there any way to bring the voltage back up to 12V, or somewhere near, and actually regulate it? (whilst still limiting the current as before)? I've tried some different ideas, with minimal success. I think this needs a completely new strategy / circuit. In fact, if I could just regulate it at that 3V that would be an improvement! (voltage does vary quite a bit depending on the load, materials, etc). Any thoughts??

One interesting observation: I found by wiring 4 or 5 electrode units in series (don't ask why I tried this, I don't know lol), the voltage actually went up and well over 12v (to 19v in one case) - Any ideas why that would occur? Would like to understand that. I suspect it’s due to the Inductor energy, and the way it’s released?
 

dougy83

Well-Known Member
Welcome back,

Glad to be able to be some help.

With only 10k the opamp output voltage was too low to influence the BUCK.
it was designed for a 5V regulator outputting 3A, which it should've been capable of.

When running at 2.5A, it's supplying around 3V. Is there any way to bring the voltage back up to 12V, or somewhere near, and actually regulate it?
Not sure what you're getting at; 3V/2.5A = 1.2 ohms. If you put 12V on the same load, you'll get 10A through it - increasing the voltage on a load generally increases the current through it. If you just want to limit the voltage and current (rectangular regulation curve), you can diode-or the output of the current-amplifying opamps already connected with the output voltage and feed that into the regulator (see attached).

PulseCurrentLimi&#1.gif

One interesting observation: I found by wiring 4 or 5 electrode units in series (don't ask why I tried this, I don't know lol), the voltage actually went up and well over 12v (to 19v in one case) - Any ideas why that would occur? Would like to understand that. I suspect it’s due to the Inductor energy, and the way it’s released?
On startup, the output voltage could foreseeably exceed the input voltage as the output voltage (due to the initial duty cycle being 100% - when the cap voltage reaches the threshold, duty cycle is reduced but there's heaps of energy stored in the inductor, hence the overshoot.. the fact that the opamps are particularly slow types won't help this but they should be fine for steady-state regulation into constant loads [just not transient loads]), but that shouldn't continue (i.e. in a steady-state condition) with a load connected.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top