MODEL165
SECTION 2.
1. Tbermoelecrric Naise. Potentials generate*
by rhermal differences Bf rtle junction a= ,unctions
of Ed0 dissimilar metals are thermoelectric noises,
more commn1y called thermal mm. These potencia1s
may be significant when making millivolt or mic=o-
volt mea*uremencs. TO minimize ehermal nOiS* ---
which may appear as a drift --- caused by thermal
FXYO, use PUT* capper circuirry emi tswinaricms
thrau*iw*t the source an* in a11 connectians to
ehe 165. The KeicNey accessory Model 1483 Low
Thermal Connection Kit contains all necessa=y ma-
cerias for making very law-thermal copper connec-
tions for minimir*n* tharmal mm..
2. AC Power Line Noise. The presence of electric
fields generated by power lines o= other power
.9O"TceB can haYe an effect 0" inaaur,ent operarim.
Also ac vok?.ges presenr in Ehe BOUrCe which are
very large vith reapact. to the full-scale range
sensitivity of the Model 165 could drive the analog
amplifier into sa~uracio". producing an e==OneO"a
digiral display.
a). Shielding. Proper shielding of the sowxe
or cabling can minimize noise pickup when the
instrument is in tile presence of large ac fields
or when very seneirive me*e"rements are being
rude. ‘ornoiae shielded cable, such as Keith+,
SC-9 cable, should do a sufficient job of shield-
ing rhe inpur signal. Metal shields may be re-
quired LO be installed arD""d the BOUIICB. me
shields of the inpur cable and source should be
connected together co ground at ooe point only,
typically at the input of the Model 165. This
one-point-ground method ia a frtree'f configuracio",
which minimi=es ground L,ape in the measured cir-
cuiery. Ground loops are a secondary source of
interfering noise Which may also be considered in
larle"el meee"reme"ts.
OPERATION
b). AC Ke,ecrim. The Model 165 provides
artenuation of line frequency noise superimpose*
on a dc input signal. The ac re,ection Of the
165 is specified as follows:
NOWL MODEkE.lECTION RATIO (NMI(Io: Greater
than
60
dB above one digit for a voltage of
line frequency o= twice line frequency with
at leas= 10% of full-range dc applied.
COMmN MODEREJECTION RATlO (cxea): 120 dB
on the 10 mv, 100 mv. and 1" ranges; 80 da 0"
the 10" range; 70 da on the 100" range; and
60 dB an the 1000" range; far a dc, 50 HZ, o=
60 Hz voltage with at least 10% of full-range
dc applkd.
3. Magnetic-fFeld Noise. The presence of strong
magnetic fields can be a possible so"=ce of obJec-
eionable ac noise. The Model 165 has been suffi-
ciently sbialded from typical magnetic incerfarence;
however, additional shielding may be required at the
eo"rce or in the cabling to the 165. Magnetic flux
lines which cue a conductor --- like an input cable
--- can produce large ac noise, especially at power
line frequencies. me voltage induced due to nag-
neric flux is proportional to the area enclosed by
the Circuit *s well as the rate of change of mag-
netic flux. POC example, the motion of a ,-inch
*iemeeer loop in the earth'. magnetic field will
induce a signal of several tenths of a microvolt.
The ac rejecrion characteristics of ebe Model 165
vi.11 help minimize specific effects of magnetio
fields. Magnetic pickup may be furthe= minimized,
by arranging all so"=ce and input-cable wiring so
L&E the loop area enclosed 18 BB small as possible
(such as by rwisring input leads). "sing conetic
(magnetic) shielding in cables and around cirwia-j
may further help in seve=e cases.
b. Effective Shielding. Here are general shielding
rules fo consider far measurements in the three func-
tion categories of the 165. Be sure that shielding
is even needed before proceeding.
1. Yolemeeer maaurements. Consider shielding
input leads when sou=ce resistances a=e g=eaCe= than
100 kilc,hm o= when long input cables are used. Avoid
even slight movements of input cabling a= Ct\e oource
when making high s.au=ce resistance measurements.
2. C"rrenC Mea8"reme"C*, on the mA and ii.4 current
ranges, generally no ape&a1 shielding precautions
need ba taken. However, consider shielded input
leads for me~~"=ements an the lowese ranges.
3. Reeiscance *easuremenc*. Shielding Of the
input cabling and sou=‘ce may be necessary far meas-
"remente on the 10 megohm and 100 megohm ranges.
4 1073
.~.
P .
