Services (low cost) for measuring "off the shelf" calibratio

[Michael Shell]

I would like to buy my own "off the shelf" set of resistors, capacitors
and inductors (5-10 of each type) and send them out to be precision
measured at a "typical" room temperature (recorded), the LC parts at
1kHz, to an accuracy of 0.1% or better (1000 ppm). I understand that
long term component stability is an issue, especially with some types
of capacitors. I'd probably go with a "base" of 1% metal film resistors,
NPO ceramic, silver mica, and polypropylene caps, etc.

Value ranges are approximately:

R: 8 to 500k Ohms (might not need over 100k)
C: 10pF to 500uF (might add 1pF if can)
L: 10uH to 8H

The purpose is to have my own references solely for my own use to
field verify the accuracy some of my own test equipment without
paying the kind of money typically needed with high end calibration
standards. The company that does the measurement should have equipment
that is NIST traceable. I know HP/Agilent made/makes some models of
LCR meters that have 0.1% or better accuracy (e.g., HP 4274A, HP 4284A,
$1K-10K).

Has anyone had any experience with this kind of thing, and/or can
recommend a friendly lab that will perform such measurements at a
low cost (say, < $100)?


Thanks in advance,

Mike Shell

 

[Damon Hill]

You can buy .01% Vishay bulk metal foil resistors from any
of several vendors like Allied, Digi-Key, etc, or browse
through the resistors here:

http://www.surplussales.com/Resistors/Res-MetalFilm.html

Mostly you'll find various older 'y-foil' types; the newest
'Z-foil' technology is even more stable. Check also under
their wire-wounds and look at the Tc numbers.

Also some Vishays on Ebay; check to be certain of the tolerance.


--Damon

 

[whit3rd]

On Feb 2, 12:56 pm, Michael Shell <ne...@michaelshell.org> wrote:

I would like to buy my own "off the shelf" set of resistors, capacitors
and inductors (5-10 of each type) and send them out to be precision
measured at a "typical" room temperature (recorded), the LC parts at
1kHz, to an accuracy of 0.1% or better (1000 ppm).

The purpose is to have my own references solely for my own use to
field verify the accuracy some of my own test equipment without
paying the kind of money

The 'without paying' is the main problem. I'm gonna go out on a limb
here, and say ... it's insoluble.

The reasons you can't get very high accuracy in reference devices
are the reasons you don't need that much accuracy anyhow. Inductors
aren't 'pure' devices, they have resistance (and frequency-dependent
resistance, and frequency-dependent inductance). Capacitors are
likewise
not pure, there's a bit of inductance in the connect wires, and
dissipation
that models like resistance. Your wire connections aren't purely
conductive.
The resistance of a "reference" resistor doesn't stay accurate when
soldered.
It doesn't stay accurate when the leads are bent/strained (and if you
want to be
REALLY precise, there's harmonics generated due to self-heating when
you put that low-frequency AC through a resistor).

Where accuracy is really required, like frequency selectivity for
radio, there's
good standards available; the very-accurate resistors, for instance,
are useful
for bridge-type sensors. Why, oh, why, would one need to have a one
percent 500 uF capacitor, let alone a 0.01% one?

 

[Wild_Bill]

Sencore used to have a small box with terminals for about 4 inductors and
capacitors they referred to as a Field Calibrator.

The internal components weren't anything special, just a couple of regular
components.
The "calibrators" were for checking the earlier models of Z-Meters, the
inductance and capacitance measuring instruments.

The user would send the calibrator in to Sencore, and they would measure the
components with their own NIST-traceable equipment, and put a label on it
noting the date and the components' capacitance and inductance values.
That gave the user something to check the Z-Meter with.

If you constructed a box with components in it, they would probably give you
a price quote to measure, date and label it. I think the last price I saw
for the calibrator service was about $60US, but that part has been obsolete
for years.

For a precision resistance, you might just take a number of precision
resistors and put them in parallel to attain the resistance step you desire.
The individual resistors could be pre-checked with your existing equipment.

Paralleling 10 pieces of 100 ohm, 0.1% (or 0.01%) resistors would
essentially cancel out (substantially minimize) the value tolerances.
You would have to just trust the manufacturer's specifications for the parts
you use.

--
Cheers,
WB
..............


"Michael Shell" <news1@michaelshell.org> wrote in message
news:20100202155649.7e9522b5@bashir...

I would like to buy my own "off the shelf" set of resistors, capacitors
and inductors (5-10 of each type) and send them out to be precision
measured at a "typical" room temperature (recorded), the LC parts at
1kHz, to an accuracy of 0.1% or better (1000 ppm). I understand that
long term component stability is an issue, especially with some types
of capacitors. I'd probably go with a "base" of 1% metal film resistors,
NPO ceramic, silver mica, and polypropylene caps, etc.

Value ranges are approximately:

R: 8 to 500k Ohms (might not need over 100k)
C: 10pF to 500uF (might add 1pF if can)
L: 10uH to 8H

The purpose is to have my own references solely for my own use to
field verify the accuracy some of my own test equipment without
paying the kind of money typically needed with high end calibration
standards. The company that does the measurement should have equipment
that is NIST traceable. I know HP/Agilent made/makes some models of
LCR meters that have 0.1% or better accuracy (e.g., HP 4274A, HP 4284A,
$1K-10K).

Has anyone had any experience with this kind of thing, and/or can
recommend a friendly lab that will perform such measurements at a
low cost (say, < $100)?


Thanks in advance,

Mike Shell

 

[JosephKK]

On Tue, 2 Feb 2010 15:56:49 -0500, Michael Shell <news1@michaelshell.org> wrote:

I would like to buy my own "off the shelf" set of resistors, capacitors
and inductors (5-10 of each type) and send them out to be precision
measured at a "typical" room temperature (recorded), the LC parts at
1kHz, to an accuracy of 0.1% or better (1000 ppm). I understand that
long term component stability is an issue, especially with some types
of capacitors. I'd probably go with a "base" of 1% metal film resistors,
NPO ceramic, silver mica, and polypropylene caps, etc.

Value ranges are approximately:

R: 8 to 500k Ohms (might not need over 100k)
C: 10pF to 500uF (might add 1pF if can)
L: 10uH to 8H

The purpose is to have my own references solely for my own use to
field verify the accuracy some of my own test equipment without
paying the kind of money typically needed with high end calibration
standards. The company that does the measurement should have equipment
that is NIST traceable. I know HP/Agilent made/makes some models of
LCR meters that have 0.1% or better accuracy (e.g., HP 4274A, HP 4284A,
$1K-10K).

Has anyone had any experience with this kind of thing, and/or can
recommend a friendly lab that will perform such measurements at a
low cost (say, < $100)?


Thanks in advance,

Mike Shell


As has been noted you can just buy the standard resistors, 0.01% off the shelf.

For your reference capacitors you are going to learn quality capacitors above
2 uF are expensive, bulky and hard to find. I doubt you can find any at all
above 20 uF. Quality (high Q and high SRF) inductors larger than 200 uH are
going to be troublesome as well.
Finally getting them measured will depend on how accurate you want them measured
how many frequencies you want them measured at and how many you want measured.
Provide them for measurement mounted with 4-terminal connections.
Finally, keep the reference parts inside with well controlled conditions.
I measured ordinary and high quality parts for a few years running. Using HP LCR
meters and even more accurate GR bridges. Making the measurements was not my
normal lab duties, creating test specific test methods and test equipment was.