Difference between 2A104J Capacitor and plain 104 Ceramic ones on LM2917

[DIY Explorer]

Hi,

I have been trying to do a frequency switch circuit, and I succeeded, with the circuit described below.

Now, I am using 2A104J capacitors (big green ones) which as I said work fine, but if I replace those (C1,C2) with generic ceramic 104 ones it just doesn't work.

Any idea of why is that?

 

[Anon_LG]

Different types of capacitors have different properties, this section: http://en.wikipedia.org/wiki/Types_of_capacitor#Capacitor_features_comparisons
shows there are lots of differences, it might be to do with how they handle different types of circuits, it says that mylar film capacitors are suited to applications involving frequency. It does mention a few frequency related disadvantages to do with class 2 ceramic capacitors.

 

[DIY Explorer]

Thanks for your replay Lavaguava... this is a very low frequency project (< 300 Hz)

Is it that 2A104J are not really 100nf?

I would tend to think that in TI's spec, it would call out somehow to use mylar film (see frequency switch on page 9)

http://www.ti.com/lit/ds/symlink/lm2917-n.pdf

Something else interesting to me is that C3 works fine with ceramic or mylar film.

Luckily enough I tried mylar film first

 

[Anon_LG]

No, its definitely 100nf. How exactly does the circuit behave when you use a standard ceramic capacitor? Maybe its that the ceramic capacitor has higher losses than the mylar film type. I am sure one of the experts will arrive soon and answer your question and more

 

[DIY Explorer]

This is a frequency switch. When it gets certain frequency as input it switches something on.

Basically Q1 captures a light frequency emitted by an LED (not in the circuit); when that frequency is present at Q1, the IC (LM2917 FVC) triggers pin 5 to ground, and LED2 turns off.

If I interrupt that external LED to reach Q1, then LED2 turns on.

When I replace mylar film capacitors with ceramic ones, LED2 never comes off. It does change intensity of light on presence/absence of Fin though.

 

[KrisBlueNZ]

Type II ceramic capacitors aren't very suitable where stable capacitance is required. Their capacitance varies significantly with applied voltage, and of course with temperature. Z-type dielectrics are the worst. Also they are common targets for cheap low-quality manufacturers, because they're so widely used. These low-quality MLCCs can have performance problems such as significant leakage. Type I ceramics (NP0/C0G dielectrics) perform well, but aren't space-efficient for large values and it's unlikely your 100 nF ceramics are that type.

You could try X7R ceramics from a reputable manufacturer such as AVX. Otherwise, for C1 at least, you should probably use film capacitors.

Also, your phototransistor circuit is relying on similar characteristics of the base-emitter junctions of Q2 and Q3. This can't even be guaranteed for two transistors of the same type, let alone different types! You can avoid this by deleting Q2 and R2, and driving C3 from Q3's collector.

For more information on MLCCs, Google MLCC Kollman.