[ARC5] Carbon Comp Drift (was: GRC-9 Update III)

[David Stinson]

Ages ago, I was told to always look closely at any
carbon comp with

either a green or blue multiplier band, as these
drifted high more

often than others.  And that seemed true at the
time.

 

Strangely, in the last few years, when refurbing
sets, have begun

seeing far more with a yellow or orange multiplier
band out-of-bounds

than the green/blue bunch.  Most common resistors
to find bad

are 33K, 47K, 100K, 330K and 470K.    Maybe it's
just because those

are more common in sets than 1 Meg or 10 Meg, but
I don't think so.

Rarely see a carbon comp  in a set below 10K
that's out of spec.

Black, brown and red multiplier bands are 90% or
better OK.

"NOS" carbon comps that have been sitting in boxes

without climate control are a different story.
Just thinking

of ones I've seen actually in sets.

 

Have a theory about this:

Inside a carbon-comp resistor is a carbon-based
resistance

"filling" or "matrix" with conductive plugs on
each end. 

 One can visualize this matrix as a bunch of
resistors in series 

and parallel;   think of a series string of
resistors, then many

strings layered like an onion.  Taken
all-together, they

make one equivalent value.

 

As the matrix ages, the outer "onion layers"
fracture.

Fracturing one of the series resistors does not
open 

that string, because the bottom of the fracture is
still

in the matrix.  It makes that string longer than
the others,

making two resistors in series where there was one

and thus, increasing the overall matrix
resistance.

As aging continues and the number of these
fractures 

grow, the overall resistance of the part climbs
ever

higher.

 

The effect is less pronounced in lower-value
resistors,

because the value of each of the imagined
individual

resistors in the matrix is lower, so the value of

the effectively "extra" resistors added by each
fracture is  

also smaller , i.e. the same physical fracture
rate that would

move a 1-Meg resistor  up 500K  would move a 

1-Ohm resistor only 5 tenths of an Ohm.

 

Thoughts on drifty resistors?

73 Dave S.

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.qth.net/pipermail/arc5/attachments/20190301/130203d0/attachment-0001.html>