[ASK] Big and old C0G 100nF leaded capacitor

[BravoV]

Bought this on impulse recently at our local component store, its quite old and even the store keeper put on sale.

Even though I do NOT need it, its the C0G dielectric type, also reputable name Kemet and also quite big 100nF (imo) tickled my hoarding nerve. 

Aware that C0G is stable and usually for timing application, even though can't beat silver mica (isn't it ?), just wondering what kind of application need this kinda of relatively big capacitance of C0G cap ?

Any comments are also welcome.


The only printed label at the cap :

KEMET
C512C0G    -> Checked at Kemet, C512 is the case size , C0G = aka NPO
104J           -> 100 nF 5% tolerance
100V
9609           -> 1996, quite old isn't it ?

Attached below photo of the caps compared to TO-220 ic size.

[chris_leyson]

Hi BravoB, if they are marked 104J then that is 10 followed by 4 zeros, 100,000n or 100n. Temp characteristic is +/- 30ppm per degree C.

[MK14]

Beaten to it, above..

What makes you think it is 10nF ?

It appears to be a 100nF, 0.1uF cap ?

104J

Ignoring J,tolerance

10, then 4 (zeros) = 10,000 = 10 x 10,000 = 100,000 pF = 100nF

104J are used somewhat often (when making stuff) by me as decoupling capacitors, so it would be difficult to forget.

But I looked it up anyway (to avoid embarrassingly being wrong....)



http://www.csgnetwork.com/capcodeinfo.html

[BravoV]

Thanks guys, silly me, 1st post & title corrected.

[chris_leyson]

Hi BravoV, at 30ppm/C tempco they would be OK for a timebase.

[TimFox]

Those Kemet molded-box capacitors are excellent units.  Inside is probably a 1206 or similar size monolythic (chip) capacitor.

[tszaboo]

Like semi-accurate/stable RC analog filters.
They are handy for protoyping and such. You can bend the pins and solder it directly to a SO8 package.

[Audioguru]

I think it is a ceramic capacitor, not silver mica.

[T3sl4co1l]

Low loss, good for RF and power switching. The lead inductance is kind of unfortunate (not as good as an SMT chip!) but it can still be useful for low voltage/low impedance resonant circuits, high stability and so on.  Maybe a low voltage, LF to MF range, RFID transmitter?

I forget what dielectric absorption is like on those, probably not as good as classic polystyrene, but it's the best ceramic you can get, in any case.  I don't think it would be bad for S&H sorts of things.

Tim

[Cerebus]

I forget what dielectric absorption is like on those, probably not as good as classic polystyrene, but it's the best ceramic you can get, in any case.  I don't think it would be bad for S&H sorts of things.

Dielectric absorbance is fair to very good on C0G/NP0 caps BUT it can vary by a factor of around 100 between different brands/ranges of nominally identical capacitors (i.e. same value and voltage rating). Some are slightly better than polystyrene, some are just worse then polyester, none are as good as PTFE. Annoyingly it rarely makes it into the datasheets so you either have to get some and test it yourself or find a knowledgable application engineer at your chosen capacitor vendor.

[TimFox]

Also, NP0/C0G has a specified temperature co-efficient of 0 +/- 30 ppm/C, but silver mica has between +35 and +75 ppm/C.

[BravoV]

I'm aware of other dielectric type like PTFE or Silver mica, its just mostly they have quite small capacitance in pF range, while this one is at 100nF, and according to Kemet literature, it doesn't age, also has a linear tempco, cmiiw.

Btw, this particular cap aint cheap either, like at Mouser cost about $23 a pop  -> Kemet C512C104J1G5CA.

Actually what I have in mind with this huge cap is to turn it into sort of "good enough" enthusiast grade long term stable cap that once measured (& profiled with high end LCR meter), it will be a reference cap for years to come, is it suitable for this job ?

[tszaboo]

I'm aware of other dielectric type like PTFE or Silver mica, its just mostly they have quite small capacitance in pF range, while this one is at 100nF, and according to Kemet literature, it doesn't age, also has a linear tempco, cmiiw.

Btw, this particular cap aint cheap either, like at Mouser cost about $23 a pop  -> Kemet C512C104J1G5CA.

Actually what I have in mind with this huge cap is to turn it into sort of "good enough" enthusiast grade long term stable cap that once measured (& profiled with high end LCR meter), it will be a reference cap for years to come, is it suitable for this job ?

Whatta? Someone put the decimal point the wrong place.

[quarks]

Actually what I have in mind with this huge cap is to turn it into sort of "good enough" enthusiast grade long term stable cap that once measured (& profiled with high end LCR meter), it will be a reference cap for years to come, is it suitable for this job ?

If I had one of these, I would exactly do that and keep it for long time comparisons.

[TimFox]

Note that traditional General Radio standard capacitors, either mica or polystyrene, were in hermetically-sealed packages to avoid problems with humidity affecting the capacitance.  If you mount your NP0 device in a box, take measures to avoid humidity.

[BravoV]

Note that traditional General Radio standard capacitors, either mica or polystyrene, were in hermetically-sealed packages to avoid problems with humidity affecting the capacitance.  If you mount your NP0 device in a box, take measures to avoid humidity.

Thanks, I guess I have to come out with a way to sort of seal it, maybe in a small glass jar filled with silica gel, and with only two leads out while the whole unit will be sealed maybe using hot glue gun or epoxy.

[SeanB]

Do not use hot glue, it is not vapour resistant.  Glass jar with a feedthrough ( probably a waterproof BNC connector per lead, fitted with some slow cure epoxy during installation) connection, filled with some very dry silica gel ( deepest blue you get after baking in an oven for 2 hours at 140C) and then seales with the same epoxy.

[BravoV]

Do not use hot glue, it is not vapour resistant.  Glass jar with a feedthrough ( probably a waterproof BNC connector per lead, fitted with some slow cure epoxy during installation) connection, filled with some very dry silica gel ( deepest blue you get after baking in an oven for 2 hours at 140C) and then seales with the same epoxy.

Good idea SeanB, especially on the feedthrough BNCs. thanks. 

Time to find a suitable jar, with metal lid that wont corrode.