unknown ceramic capacitor value

[Paul_L]

I seriously would not trust the values of these caps. I picked up a nice assortment of cheep resistors online. I wasn't getting the expected results in designs. I measured over a hundred of them and found them to be way off some as much as 35% they all were 10% (silver band).

 

[throbscottle]

Physical size may be your best indicator here. If those squares in the photo are 5mm then if it's not 1pF, then you are looking at low 10's of pF.
Also as a last resort you could pick the outer coating off one and measure the thickness of dielectric and diameter of the plates. You could work it out from that (but don't ask me how!).

 

[gary350]

Unless you are Superman or Time Warp Man

But seriously, 1pf is not easy to measure. That's almost like trying to measure a resistance of 0.001 Ohms. Yeah, you can do it, but it takes special care.

My $100 meter doesnt even go that low, but it can measure 10pf so i measure the 10pf and get a reading, then parallel it with 2pf and get another reading, then use the formula for two caps in parallel:
Cp=C1+C2
where Cp is the total capacitance and C1 and C2 are the two actual capacitors.

C1 is the known capacitance which i take a reading first, and say i get 11pf, then C1=11pf, so then i parallel with unknown then get:
Cp=11pf+unknown_pf
and say this second reading is 13pf, then we have:
13pf=11pf+x_pf
and solving for x_pf we have:
x_pf=13pf-11pf
and we all know that 13-11 is equal to 2, so the unknown cap must be 2pf.

But what if the unknown was higher. Then i would have seen maybe:
20pf=11pf+x_pf
so here the x_pf (the unknown cap value) must be 9pf.

Note that i had to take a reading first with the known capacitor.

With a frequency oscillator like the 555 the frequency will be close to following the capacitor value in proportion to that capacitor value. So say we connect a 100pf cap with some resistance (possibly variable) and we see a frequency of 1000.0 Hz. If we connect a 10pf in parallel with that we have:
Cp=100+10 (in picofarads)=110

and 100/110=0.90909

so if we had a frequency of 1000Hz to begin with, the new frequency would be 1000*0.90909 which would be 909.09Hz which might show up as 909.1Hz.

This means if we measure a frequency of about 909Hz then the unknown cap is about 10 percent of the known cap. So we have the formula:
C_unknown=(F1/F2-1)*C_known

where F1 is the frequency with only the known capacitor connected, and F1 is the new frequency with both caps connected.

If the new frequency happened to come out to 500Hz, then we would have:
C_unknown=((1000/500)-1)*100=100

so the unknown cap would have to have been 100pf also.

Here is a list of factors for 'new' frequencies of 900Hz down to 100Hz:
900Hz, C/9
800Hz, C/4
700Hz, (3*C)/7
600Hz, (2*C)/3
500Hz, C
400Hz, (3*C)/2
300Hz, (7*C)/3
200Hz, 4*C
100Hz, 9*C

where C is the known capacitor here and the original frequency was 1000Hz.
If the original frequency was 10kHz instead, then just multiply all those frequencies by 10.

So for example if we measure 800Hz with a known cap of C=100pf, then the unknown cap must be C/4 which comes out to 25pf.
But if we measured 200Hz, then the unknown cap must be 4*C=400pf, but we usually dont have to measure larger caps this way.

It works best if your don't parallel the unknown cap with the known cap.

When I was in high school 1968 almost 50 years ago there were few reasonable price devices that could test a capacitor so I invented this timer circuit with a neon light to test my caps.

First you use a known cap to get the neon flashing at a certain speed like 2 seconds. Then you remove the known cap from the circuit then take a pile of unknown caps 1 by 1 you touch the cap wires to the 2 brass test posts and hold it there. If the light flashes every 2 seconds you know the test cap is the same value as the known cap.

I use to strip parts from old TV sometimes certain caps had no value just manufacture code numbers. You can test a pile of caps in just a few minutes and pick out a certain value then mark the container so you know what you have.

I built a pretty good cap tester with a selector switch to several known caps. Once I got the variable resistor set to flash the neon at the speed I wanted I rotated the selector switch to the 2 brass test posts. Then I could test unknown caps at a pretty fast speed. Then rotate the switch and test a different value cap. I could test all the caps from a TV in just few minutes the tester also told me if the cap was bad. If the neon does not flash the cap is bad.

Your right it wont test a 1 pf cap unless you solder a million of them in parallel.