Need an automatically cycling test current source

[LATUBEGEEK]

I want to generate between 4 and 20 mA, to drive loads up to 500 ohms, and have it vary automatically (to test process instruments overnite, for ex). Want it to be simple. I do this manually with my Fluke process meter, but want something to cycle an instrument for an extended time period (hours) to be sure it is reliable. Cycle rate (4-20-4) around 10 seconds. Am thinking about a very slow oscillator, feeding a power transistor to handle the load. I don't mind having to adjust voltages to get the current I want ( a few adjustment pots). Will be checking the web for circuits along this thread. Thanks in advance for any input. I have a fair assortment of transistors and chips (555 timers and OP amps) to play with. Thanks!

 

[(*steve*)]

How do you want the current to vary? (ramp up and down, ramp up and reset then ramp up again, change in steps (possibly randomly)?)

How fast should the changes be?

e.g.

ramps up from 4 to 10mA in 10 seconds
then jumps to 15mA,
then ramps from 15 to 20mA in 30 seconds,
then drops back to 4mA in 1 second,
pauses for 3 seconds (at 4mA),
then starts again.​

 

[KrisBlueNZ]

Also, how accurate do the currents have to be, and how smooth do the changes need to be? In other words, would it be OK for the current to change in steps of, say, 0.1 mA, or does it have to move totally smoothly? And do you have experience with microcontrollers?