more resistor talk...
Mark Amundson
mamundso at mr.net
Fri Apr 3 05:33:53 CEST 1998
Steven Maietta wrote:
> What are the criteria for choosing the wattage rating for resistors in a
> project? 1/8, 1/4, 1/2 etc.... Also, I would think that for applications
> like oscillators and other precision thingies of ours, a low tolerance
> would be best. Are the benefits of using 1% over 5% that much? I would
> love to know that I could imporve the long term performance of the things I
> build and fix by using the right components, and would definitely spend the
> extra money. Any input to this? ~Steve
Looks like its time for a resistor FAQ to go with the capacitor FAQ
generated this winter. My contributions:
As mentioned by R.G., a derating method is employed by circuit designers
to keep the resistors from operating near their maximum capability.
Myself,
I tend to compute a worst case RMS power dissipation and then choose a
resistor rated at least twice the computed value. Some military
handbooks
require a ten times computed dissipation to be used.
Another aspect to be noted is a temperature coefficient (tempco) of
drift,
typically given as parts per million (ppm) drift of total resistance for
each degree of temperature change. Obviously, if you dissipate heat from
the resistor it will drift a bit. It becomes a thermo-dynamics problem
in
that air flow or heat-sinking will mitigate resistance change. The ppm
number is used in place of percentages; e.g. +100ppm/C equals .01% of
total resistance value increase per degree celsius.
Resistance manufactured value tolerance is typically available in 20%,
10%,
5%, 1% and 0.1% percent ratings. In practice, the 20% and 10% are no
longer used as the manufacturing processes are good enough today not to
charge a premium price for 5% resistors. Metal film type resistors are
commonly available in 1% values at low costs. Below 1%, most resistors
are manufactured on a custom order basis. In surface mount parts thin
film
construction is used over lower cost thick film resistors for more
precision in manufactured tolerence.
Integrated circuits also have internal resistive elements fabricated in
the silicon. Since they can precisely control the geometry but not the
sheet resistivity of the fabrication process, on-chip resistors can be
held to tight resistor to resistor ratio tolerance but not to absolute
value tolerences. On chip tolerences of 10% or less are very difficult
to
obtain without post fabrication trimming and its additional expense.
My .02$ (20,000ppm of a $) worth.
Mark Amundson,
More information about the Synth-diy
mailing list