I am working on a magnetohydrodynamic boat project. The goal is to run a current of 5 amps through salt water. I bought an LM138 voltage regulator, configured it as a 5A current regulator by setting R1 at 0.24Ω. I have 24V on the input and the output is shorted directly to ground with a section of the output running through the salt water.

My problem is that the regulator is shorted, so its overheating. My next thought was then to add a high wattage resistor on the output as a load. But then if I am going to configure it that way, why have the regulator at all? My professor said to use an active circuit to avoid draining the battery. But it seems to me that whether I simply use a resistor between a battery and ground, an LM138 regulator, or a power mosfet circuit they all will be consuming 5A*24V = 120Watts, so what is the difference?

If anybody has some advice I would greatly appreciate it! Thanks!
JWK

 

I tend to think you have a misunderstanding, not a circuit problem. Your circuits look good, your descriptions make sense, your requirements are exactly as you think, unreasonable.

and yes, you have figure the overheating chip correctly.

 

Thanks for your response, I need to know the unreasonable parts of the circuit, please expand on that part if possible. I know there has to be a simple way to run 5A current through salt water. Is it the use of the LM138 thats unreasonable, or am I failing to understand that I am going to have a lot of heat dissipation issues no matter what method I use?

 

With such a high voltage(24 volts) it is very possible you will succeed in making lots of hydrogen and chlorine gas. Look up the data on electrolysis to see what the voltage is where atomic disassociation begins to occur. It is very small. Around 2 volts if I remember correctly. You may have to drastically increase the salinity of your salt water to pass that much current without creating more than a 2 volt drop through its resistance to avoid creating gases in your propulsion project. Depending on the size of your chamber, avoiding electrolysis entirely may not even be possible.

 

Electrolysis will occur, I wrote a couple of pages on it with calculations on the expected amounts of Chlorine that will be released at different currents. I have attached it if you are interested in looking at it. Feel free to let me know if I made any errors in calculations.

 

With a large heat sink and fan, would it be possible to keep the LM138 from overheating if I used 12V instead of 24V?

Is 12V enough potential difference to make a current flow across salt water?

 

I was thinking of the requirement to use an active circuit to avoid draining the battery. Using 5 amps from a battery will drain it, no matter what else you do, active, passive, or whatever.

In post 1, you said the output is shorted directly to ground. I thought you meant the salt water shorted the output. Not being any kind of chemist, I would try varying the voltage to see how much you need, then provide about 3 volts more than that to the 138 regulator. It's the excess voltage that is causing the heat problem.

 

Thanks for the input, I think the plan from here will be to do what you said and find the minimum voltage needed, account for the 3 volts drop across the LM138, and then try to cool the regulator with a fan and heat sink.

 

You could also consider using mechanical methods of regulation. No circuits involved other than electrodes, battery and salt water. By varying the salinity and adjusting the distance between electrodes, you can achieve a modest range of adjustment in your current flow.

 

Hello JWK,
I received your E-mail. I have been offline since Thursday afternoon.

I don't happen to have any of my reference materials with me. However, I suggest that what you need is more on the order of a switching supply rather than a linear regulator; as no matter how you configure a linear regulator, you're going to dissipate lots of power as heat.

Unfortunately, switching supplies are much more difficult to understand than linear supplies - and I'm afraid that I simply don't have the time to explain them at the moment.

But very basically, with linear regulation, you wind up with power dissipation in the regulator and in the load; with your scenario, you'll likely wind up with more power dissipation in the regulator than the load. With switching regulation, current flow through an inductor is switched on and off; the inductor has a relatively low resistance, instead of dissipating power as heat, the energy is stored and released from the inductor's core. Efficiency can be quite high over a broad range of output voltages and currents.

You might find some of this thread useful: http://forum.allaboutcircuits.com/showthread.php?t=54719
however, you really can't use PWM in your application, as even though your load is (basically) resistive, you are not trying to heat it up - you want a fairly constant current flowing through it. PWM without an inductor will give you either too much current or no current at all.

You might go on Linear Technology's site and have a look at a few regulators, along with downloading LTSpice; it's a good and free PSPICE tool.
A few candidates:
http://www.linear.com/product/LT1374
http://www.linear.com/product/LT3844
http://www.linear.com/product/LTC1624

I have to run.

 

Thanks for the idea, when I go to the lab at school tomorrow I'm going to do some tests with current through different salinity contents, as well as trying out a heat sink and fan that I cannibalized from an old cisco router. Do you think any thermal compound will do or should I order some top of the line stuff? If so, what thermal compound is the best for somebody with a low budget?

Thanks for the help!

 

SgtWookie, Kermit2, #2,

Thanks for the replies, your idea's were a great help. I'm really glad I found this website. I have been reading it for a while but this is my first thread posting.

 

http://www.romanblack.com/smps/a05.htm

This is a switching current regulator. Not exactly right for your purpose, but if you can understand it, you can modify it. If you can't understand it, study some more. I think this is what your teacher wants.