ICM7555 in Astable Mode - actual frequency not same as calculated frequency.

[Rod___]

Hello,

My first post here. I need some help with an ICM7555 Timer in Astable mode.

I'm working my way through "Learning the Art of Electronics - 3rd Ed" in order to keep my brain from going to mush in retirement I'm now on the chapter that introduces the 555 timer, I have no problem with the theory but I'm getting a strange result when I build the circuit on a breadboard. - I'm sure it is something I'm missing but I can't see it.

The circuit is the timer in astable mode, Fig 8L.8 page 341 for those with a copy of Learning the AOL (or Fig L10.5 P236 in the 2nd Ed. - Silver cover)

When I work out the frequency of the o/p I get 480 Hz. (f=1.44/(Ra + 2Rb)C, Ra and Rb are 10k, C is 0.1uF)

When I use LT Spice I get 480 Hz., same as the math.

On the breadboard the circuit runs at 846 Hz. I checked this on a modern scope (R&S HMO 3034), on an old Tektronix 2225, and even on a DMM. They all say 846Hz. approx.

I'm sure I'm using the correct component values. I've changed the ICM7555 for an LM555 with the same result.

The duty cycle is correct, and Vdd is at +15V as per the AOE book.

Anyone got any ideas why the frequency is so different in reality, I've run out of ideas?

Happy to post Spice and scope screenshots if needed.

Many thanks for any help.

Rod

[floobydust]

Your timing capacitor may not be what you expect. I would make sure to use a good film cap.
Ceramic caps have a "voltage coefficicient" and sloppy tolerance which together give you less than expected capacitance.
https://www.edn.com/design/analog/4402049/Temperature-and-voltage-variation-of-ceramic-capacitors--or-why-your-4-7--F-capacitor-becomes-a-0-33--F-capacitor

[Rod___]

Hi Floobydust,

Fantastic, you are spot on. 

I read the article, it appears, to my limited knowledge, that it's risky to use ceramic caps for any time-critical applications. When measured on a capacitance meter I see a value of 100nF but if I leave it in the circuit the frequency gradually creeps up from 840Hz to 960Hz. Clearly they are not reliable or stable.

I had a 10nF film cap, worked out the frequency (4.8kHz) and the scope shows 4.7kHz. Also the frequency is stable over time and much more stable when changing Vdd.

Many thanks for your response, I'm now more knowledgable about ceramic caps and their use in timing circuits.

Rod

[Zero999]

Yes ceramic capacitors have a poor temperature coefficient.

Here's another experiment you could try. Add a much larger capacitor in series, say 10µF, with your 100nF ceramic capacitor and put a large value resistor in parallel with the ceramic, about 1M. Apply power and monitor the frequency. It will start off high, near 840Hz, then fall down, closer to the calculated frequency. When the power is applied, the voltage across the ceramic capacitor will be quite high, since it will charge very quickly, compared to the 10µF capacitor and its capacitance will fall, as the 10µF capacitor charges, the voltage across the 100nF capacitor will fall, causing its capacitance to rise.

It's also possible to build a voltage controlled oscillator by varying the bias voltage on the ceramic capacitor.

[Rod___]

Thanks Hero999, I'll give that a whirl.

Still lots to learn

Rod

[TimFox]

The negative comments on ceramic capacitors are relevant to X7R, Z5U and other dielectrics that should be used only for bypass caps and similar applications.
NP0, aka C0G, capacitors are stable and linear, but become expensive, large, or unavailable in high capacitance values.