Control current despite variable resistance

[waunaclock]

For a senior engineering class I need to design a device that, using a 5 volt source I need to deliver a 5mA current with a load that can vary from 500-100K ohms. Using transistors I know one can create a current limiting circuit, but once the resistance gets high enough (not sure how to calculate how high) then ohms law takes over.
I spoke to a professor who suggested I look at op amp voltage followers which from what I gather that isolates voltage from the load, but voltage is not my current concern.

A friend of mine suggested I use an inverter to make the current AC and connect a transformer, thus allowing me to generate enough voltage to handle the load's resistance and since the transistor circuit limits current to 5 volts that's not a concern.

Can someone direct me? I'm in over my head and time is short. thanks in advance.

 

[vk6kro]

Better check the problem.

You can work out the voltage across 100 000 ohms with 5 mA flowing in it by using V = I * R.

So, somehow, you would have to generate this voltage, at least, from 5 volts and then produce a current regulator at that voltage.

It seems unlikely that you would be given such an assignment. So, could you check that that is what was actually asked for?

 

[waunaclock]

Yes, it is a ridiculous task, but that is the task. My source cannot be greater than what you would pull from a dc converter or a few batteries, meaning in the 4-12volt range.

 

[Carl Pugh]

E=0.005 amp X 100,000 ohm= 500 Volt.
Convert 5 VDC to say 525 VDC.
then use a constant current source.
Google will give you information on power supplies and constant current sources.

 

[waunaclock]

E=0.005 amp X 100,000 ohm= 500 Volt.
Convert 5 VDC to say 525 VDC.
then use a constant current source.
Google will give you information on power supplies and constant current sources.

How would I step up voltage that much in a DC circuit? I can't find any components that go above about 50 volts

 

[jim hardy]

vk6 has a point.

5mA current with a load

does that mean all 5 ma go through the load ?

 

[vk6kro]

I had to make a Geiger counter a long time ago. This needed about 500 volts and it had to be portable.

I had an oscillator consisting of a couple of CMOS 74C00 NAND gates driving an audio speaker amplifier chip (a LM380) which drove the low voltage side of a 250 volt to 12 volt transformer.

The output was rectified and filtered and used in the instrument.

However a reduced feedback voltage was taken back to the 74C00, inverted and then applied to one of the inputs in the oscillator gates. If the output got too high, then the oscillator would stop.

This is a bit general, deliberately, as this is an assignment, so you will need to fill in some of the bits I haven't mentioned.

If you haven't done DC-DC converters in class, you could verify that your lecturer knows you are about to build a 500 volt power source. He might like to rethink that project.

I should mention that 500 volt power supplies are lethal devices and you should not build one unless you are familiar with the precautions that are needed.

 

[waunaclock]

vk6 has a point.



does that mean all 5 ma go through the load ?

Yes it does

 

[jim hardy]

Okay , thanks...

As VK6 hinted it seems to me you'll have to build a step up power supply and control its output to get desired current. It's no small task. Fortunately it's a common practice in switching supply design to measure voltage or current and reach backward across an insulating barrier with a control signal. In 1980's that was in expensive industrial supplies only but recently i encounterd one in daughter's Vizio flatscreen TV. So they're in home electronics now. Optoisolators are often used and I've seen articles on magnetic amplifiers designed for that application.

I'm no power supply designer . Myself I would start with application notes from IC manufacturers. I'll look tonight and see if i saved any .

Perhaps somebody else here has an easier approach. I'd ask professor if that was his goal, to get you into switchers.

Lastly i just had a thought... Since film cameras are a thing of the past, thrift shops are full of accessories. Look around for a flash unit - bigger the better. It will have inside a high voltage supply that runs on typically 4AA cells and makes around 400 to 600 volts for the Xenon tube. I do not know if it will make 5ma, but it's got the inductor and switching transistor inside. Worth a look to see what it will do.

 

[waunaclock]

Lastly i just had a thought... Since film cameras are a thing of the past, thrift shops are full of accessories. Look around for a flash unit - bigger the better. It will have inside a high voltage supply that runs on typically 4AA cells and makes around 400 to 600 volts for the Xenon tube. I do not know if it will make 5ma, but it's got the inductor and switching transistor inside. Worth a look to see what it will do.

This actually helped lead us to our current approach. CCFL inverters more or less do what we are looking for http://search.digikey.com/us/en/products/BXA-12553/289-1049-ND/284629
The question then is how to map it out in Multisim. Wikipedia has a diagram, http://en.wikipedia.org/wiki/File:CCFL_Inverter_Circuit1.gif, but having never dealt with inductance or transformers I do not know where to begin nor do I know how to work out the values, and our textbook helpfully leaves of inverters.

I of course am not looking to have the answers handed out right, but I would like someone to help me understand the wikipedia diagram or perhaps explain what each bit does and then I can work outthe exact numbers based of an example?