tantalum cap leakage?

[AndrewT]

Hi all,
I want to build a preliminary timer using 555 to get about 20minutes delay. Eventually I will go digital using a microprocessor after I have proved the philosophy of my control system works.

To achieve 20 minutes I will need a large value resistor and an electrolytic to get the time constant. But leakage through the electrolytic will upset the timer. To complicate it there will be about 10 cascaded timers giving an adjustable total delay of upto about 3 to 4 hours.

I tried measuring the leakage of a 47uF 16V tantalum cap but I can't measure it. I tried to measure the voltage across the 1M0 resistor in series with the cap and the 10V supply.
The resistance (10M) of my DMM upsets the voltage readings indicating that the effective leakage was a lot less than 1uA.

Q. is tantalum this good at 10volts? It appears to be a lot better than the data books state.

 

[anatech]

Hi Andrew,
I've used tantalum caps for that. They are better than electrolytics. In my application, the interval was not critical, nor was it necessary to be repeatable. You could use a lower value cap and a cmos counter too.

-Chris

 

[dhaen]

Not what you asked, but:
You might consider the 74HC4060 / CD4060 14 stage ripple counter with built in oscillator. It requires only 4 external components and the cap can be in the order of 20n for your required time interval.
74HC4060

 

[poobah]

Here's is another...

http://www.onsemi.com/pub/Collateral/MC14536B-D.PDF

There are a few more in the same (OLD) HV CMOS family...

These won't be so touchy over temperature.

 

[anatech]

I used the IC that John "pointed" to for more critical applications. 4060 at the time.

-Chris

 

[poobah]

Yep...

And these things are built to be cascaded so you'll only need 2 instead of 10.

14 or 24 bit counter doubled... 2^28 & 24^48 = big number