I saw an article in "Nuts & Volts" about a capacitor test device based on the discharge time of a capacitor. It occurred to me that the pot command would work in the same manner, and be much easier to assemble.

With a 16F84A chip, I gave it a try, and it works, within limits, very accurately. I have an old Heathkit Capacitor substitution box, and my bread boarded effort is dead on from 100pf to 1uf.
Anything below 100 pf returns 0, even with a 10meg discharge resistor.

Above 2.2uf, I get a return, but it has no relationship to the cap's value. A 220uf cap might return 90uf, while a 47uf might return 148. I tried discharge resistors as small as 1 Ohm. Any return is useful, as open and short both return 0, but I would like to get a measurement over 2.2uf. And I would like to get a measurement below 100pf as well.

I tried adjusting the prescaler, but it had no effect on my results. I am using a 4mHz resonator. Perhaps a faster clock would help the low end?

For caps from 2.2uf - .001, I am using a 1k discharge resistor and a scale of 40. For 1000pf to 100pf, 20k and scale at 255.

I suspect that internal resistance of the high value electrolytic affects their discharge curve in a way that confuses the chip. I get better results with tantalum caps, but they are still off 50 - 100%. And, perhaps any charge on a very small cap is lost during the time the chip switches from charge to measure.

The author of the article claims better than 1% accuracy, 1pf to 999uf, but I wonder.

The program is attached. I will provide a schematic when I draw one if any are interested.

Any thoughts?