555 timing problems with ceramic capacitor

Hi
I am using an TLC555C chip from Texas instrument to form a monostable timing circuit (see attachment). When simulated this circuit gives a delay of approximately 17s, in real tests using a 10uF ceramic capacitor the delay is approximately 12s and 18s with a 10uF aluminium electrolytic capacitor. Have anyone else experience that the timing results gets wrong using a ceramic capacitor?

Ceramic and Al electrolytics have wide tolerances, high leakage and high temperature coefficients. They cannot be relied on to provide accurate timings. The collector-emitter leakage of Q8 may also be an additional source of inaccuracy. Moreover, although I haven't had time to analyze your circuit, it does not comply with the standard monostable circuit for a 555.

Here is a monostable circuit using 555 IC
**broken link removed**

time period, T = 1.1 × R1 × C1

Why 1.1? The capacitor charges to 2/3 = 67% so it is a bit longer than the time constant (R1 × C1) which is the time taken to charge to 63%.

Choose C1 first (there are relatively few values available).
Choose R1 to give the time period you need. R1 should be in the range 1k to 1M, so use a fixed resistor of at least 1k in series if R1 is variable.
Beware that electrolytic capacitor values are not accurate, errors of at least 20% are common.
Beware that electrolytic capacitors leak charge which substantially increases the time period (18s in your case) if you are using a high value resistor - use the formula as only a very rough guide!
For example the Timer Project should have a maximum time period of 266s (about 4½ minutes), but many electrolytic capacitors extend this to about 10 minutes!
See more : 555 and 556 Timer Circuits

Both the capacitor leakage and the BJT cut-off current may be a problem, in particular at high temperature.
You may try to test each component in this respect simply heating or freezing it (with a spray) to see how things change.
Anyway an aluminium electrolytic is not expecially suited here...

Pjdd said:

Ceramic and Al electrolytics have wide tolerances, high leakage and high temperature coefficients. They cannot be relied on to provide accurate timings. The collector-emitter leakage of Q8 may also be an additional source of inaccuracy. Moreover, although I haven't had time to analyze your circuit, it does not comply with the standard monostable circuit for a 555.

Click to expand...

I have tested without Q8 just to isolate the timing circuit, and I have the same results (12s and 18s).

Why trigger input is connected like that?

jdh_1984 said:

I have tested without Q8 just to isolate the timing circuit, and I have the same results (12s and 18s).

Click to expand...

As I said before, the biggest contributor to the timing inaccuracy will be the capacitor's tolerance and leakage. Al electrolytics may have tolerances as wide as -20 to +80% and the leakage may be several microamperes for a 10uF cap. High value ceramic caps also have large tolerances although the leakage will generally be lower than with Al electrolytics. Even the timing resistor R28 may introduce some inaccuracy unless you'e using a 1% type or a selected resistor.

I found what can be the cause of the problem. It seems that a 10uF+/- 10% 6.3V and a 10uF+/-10% 10V ceramic capacitor don't have the same characteristics, so when I change from a 6.3V to 10V the circuit the time delay increase from 12s to 18s (I have tested different capacitors of both type and get the same results every time).