I need to deliver a variable current between 0-2 Amps to a small load (15 Ohms).... At first I was thinking about building a current amplifier..... but after investigating that route again, i think, it's too difficult to build.
Then I heard about adjustable voltage regulators. And looking through the datasheet of an LM117, I noticed it had a curcuit example it called: "Precision Current Limiter" where the output current is controlled by a potentiometer and can be varied from 0 to 1.65 Amps.
I'm not sure if this is a variable constant current source, but it kinda sounds like it to me, is anyone familiar with this circuit?
In the "Precision Current Limiter" circuit the current is controlled by a potentiometer, but I need to control the current with a variable voltage (between 0 - 5 V), is there such a thing as a voltage controlled resistor? So that I could use that in this circuit?
My load changes non-linearly with the voltage across it, that's when i'm trying to steer towards a steady current. My load is a Shape Memory Alloy wire.
Instead of using the LM117 "Precision Current Limiter", use the LM150 "Precision Current Limiter". It goes up to 3 amps. Yes it is a variable constant current source. You set whatever constant current you want, and as the load varies, it will maintain that current (provided the voltage supplying the constant current regulator is high enough).
Then use a meter to measure the current. Surely the load doesn't change with a fixed voltage applied to it? Essentially you will be building a constant current source anyways(ultimately its a voltage regulator anyways and you can find schematics to build constant current sources in many of the voltage regulators datasheets).
That is, even if your load is non linear as long as your load does vary with time then it should be good enough.
i.e., by Ohms law V = F(I). As long as F isn't to complicated you can solve for I in terms of V. If we assume F is injective on some bounded interval then we can find the inverse of F.
So how to do this easily in practice? If you know the bounds on your load then you can plug in a voltage and measure the current and then adjust V until you get the current you want. Essentially you yourself is the constant current source.
Say your load registers 1A at 10V but you want to supply 1.5A's. What to do? Try increasing V to 11V and see how much your current changes and which direction(although its most likely in the same direction as the change in current).
Doing this you can put any current you want through the load and as long as the load doesn't depend on t to much then it will be pretty accurate.
Essentially you can map out the V-I curve doing this and get the dynamic resistance of your load. I don't know about memory alloys but as long as they don't get to hot then it will work fine. It really depends on your application though. You'll just get a different V-I graph for different temperatures.
If you plan on doing a lot of work with it you will be better suited to get a constant current source though. It would be much easier to turn then dial to the current you need instead of measuring the current and voltage. (although it wouldn't be so bad with two multimeters, one measuring current and the other voltage.)
Ofcourse there are some issues with my method above as there is no current regulation. If the load changes with time it could cause some serious problems. A constant current source is just a voltage source that will change the voltage to that it can deliver a constant current. Essentially my quick and dirty method above is where you become the control feedback circuit. Not the best way but works in a pinch.
What the last poster says is true, and recommendation of the LM150/350 is good, but in context of the original question, there are problems with this approach. (1) The OP stated that he needed to supply a fixed resistance with 0-2 amps. The "Precision Current Limiter will not go to zero amps. It can get close, but never to exactly zero amps. (2) The current limit pot that controls the current must be able to handle at least 2 amps... not an easy task for a common potentiometer. I suggest that the OP consider the circuit titled "Adjustable Current Regulator" in the LM150 datasheet for full realization of his requirements. The additional regulator and negative voltage source in that circuit lets the output current go down to zero amps, while still giving good control up to full output current.
--
Dave M
MasonDG44 at comcast dot net (Just substitute the appropriate characters in the
address)
Some days you\'re the dog, some days the hydrant.
Very good observations, Dave. When you don't go through the numbers, it is easy to overlook some of the details.
Item (1): That may or may not hurt the Op, based on what he is using it for.
Item (2). That would probably be a problem.
The "Adjustable Current Regulator" would be okay, if he has a negative supply. At least according to the drawing, he would need a -5 to -10 volt supply, too (even though a very low current negative supply). Or, am I reading it wrong?
here's one way to do it using an LM350 +33.5V --[RL]----[LM350A]---[0.1]--- 0V | 5W | Vin (0-5) --[2K4]---+--[100R]--- -1.3V The silly looking 33.5V supply is needed to accomodate the 30+3+0.2V drop under full load and stat under the 35-1.25V maximum under no load. it works on paper with ideal parts but 1% drift in the part translates to a
"panfilero" wrote in message news: snipped-for-privacy@v45g2000cwv.googlegroups.com...
[limited crossposting to two]
I made two circuits that may do what you want, using just two transistors and some resistors and diodes. The first one uses bipolar transistors, and the second uses MOSFETs, which seem to be more linear. You may want to use devices that are better rated for the purpose than what I found in the LTspice library. In each case, the voltage input varies from Vbe or VgsOn to about 2 volts above that, to vary the output from zero to two amperes. I tried it with loads from 1 ohm to 15 ohms and it stays close to 2 amperes (or wherever you set the current). You need a raw supply of 35 to 40 volts to get the 30 volts into 15 ohms. I used a one ohm sense resistor, which drops 2 volts at full current, but you might get by with 0.5 ohms or less. Temperature stability will be an issue with this circuit, but it might do the job. Of course, an op-amp could be used with current feedback to achieve near perfect regulation. The LTspice files follow:
Hi, Joshua. You're saying you need to provide 0-2ADC to a 15 ohm load (30V compliance) which is controlled by a voltage source. The simplest way to do this is with a floating load (neither side of the load connected to circuit common). Given that, you can do the whole thing, including perfboard, with Radio Shack parts like this (view in fixed font or M$ Notepad):
(This cookbook circuit is courtesy of the estimable Mr. Winfield Hill at s.e.d.)
With a 30V supply, you'll be seeing maximum power dissipation at the transistor when I(out)=1A and V(load) = 15V. That means the other 15V will be dissipated by the darlington transistor, giving you a transistor power dissipation of 15 watts. Be sure to provide a really good heat sink for your transistor if you don't want to let the magic smoke out.
The op amp is another consideration. You'll need an op amp which can accept input voltages down to the negative supply rail to accept 0V input. You could do worse than use 1/2 of an LM358. But be careful not to have the supply voltage for the op amp exceed 32V.
Depending on just how you need to hook thiungs up, you could do this relatively easy, using an op-amp to measure voltage accross a small value resistor (one end ground other end your load) as compared to your control voltage. Use output of op-amp to drive a big fet to send juice into other end of your load.
depending on type of load take appropraite care so it can not oscilate.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.