Optimizing Reading Rate; At Turn On; Your Signal Environment; Speeding The Reading Transfer - HP 3456A Operating And Service Manual

Operation
Your malC,mum reading rate with the 3456A is influenced
by a large number 01 lactors. not the least 01 which is the
s'gnal you are trying to measure. These lactors can be
d,v,ded into two catego,ies; signal related and voltmeter
'elated. Among signal related lactors are;
· desired accuracy (or resolution)
nature of the signal ldcv. acv. or ohms)
•
signal environment !line related and broadband noise)
·
Among voltmeter related factors are:
· method of measurement transfer JPacked Output.
Reading Storage, System Output!
· numb81 of convenience features selected (Math
operations. Autorangej
For all Operalions. an equally impOrlant consideration is
what you're trying to accomplished by measuring fast. Are
V O "
scanning a large number of points so that fastar
readongs mean a bette, picture of what is happening at
a s,ngle point in time �
trYIng to read last so Ihal you do nOI use up valuable
compute. lime wa,lIng for the measurement operation
to complete�
" y,ng to digitize wavelorms1
·
The 3456A can solve these application problems in many
different ways.
AT TURN ON
·
For most bench and system applications, the preprogram·
med seuling times assure accurate readings regardless of
the signal and signal environment. However, knowledge
about your signal andlor signal environment can let you
achieve up to 330 rdsls (at 60 Hz) with flOise rejection and
1 0 microvolt sensitivity.

YOUR SIGNAL ENVIRONMENT

Your signal is subject to line ,elated and broadband noise
which can interlere with the measurement. There are two
ways of rejecting noise on the 3456A; integration and in·
put fillering.
Integration is a process where the allect of line related
noise is averaged to zero over the period 01 an integer
n .... mber of Power Une Cycles
version, The basic integrator is an Op Amp with a lino
tegrato,) capacitor in its feedback loop. The signal is con·
nected to the input 01 the Op Amp for a period of the line
IIequency. This configuration theoretically provides infinite
noise rejection al integer multiples of the inverse of the in·
tegration period and single pote roll·off 120 dB of amplitude
attenuation fOI every decade increase in IIequency above
the knee frequencyl for broadband noise. Refer to the graph
lor more detail.
The Input tilter provides el(cellenl noise rejection. The price
you pay is an additional 650 ms settling lime which allows
the filter o .... tput to seule to a final value before the input is
measured. The fastest possible reading rale measuring
widely varying signals with line rejection is provided by 1
PLC inlegration time. But, if you lire looking at a slowly
varying signal or scanning similaf signals, you can get
faster readings by overriding Ihe preprogrammed settling
time selecting 0.01 or 0.1
filter. The trade·off is less accuracy and resolution as well
as more uncertainty about the filtered inpul.
3-12
Table

Optimizing Reading Rate.

3·4.
during the A to 0 con·
WLCI
integration time and the
PLC
SPEEDING THE MEASUREMENT
The thermal stability of the measurement environment is
impommt. By simply keeping the temperature of the
3456A at a fil(ed value, you can nearly double the reading
rate by turning Auto Zero off. The 3456A is slightly less ac·
curate but the faster reading rate may be worth it. In addi·
tion. any range. function. or lilter change that takes place is
automatically accompanied by an Auto Zero update which
removes any accumulated offsets. Of course. if the
measurement environment is quiet enough to omit NMR
then only the accuracy and resolution of your desired
measurement are the factors and you can achieve up to
Z 1 0 or 330 rds/s with . 1 and .01
Further, you can speed the measurement cycle by selecting
a fil(ed range instead of using the Autorange function. Even
al 1 integration time there is a 1 0 % reduction of the
PLC
reading rate with Autorange on. Of course. il your signal is
changing, Autorange is far faster than letting your software
range the 3456A over the HP·IB.
AC Volts requires some special attention. With the filter
011. you can measure signals of greater than 400 Hz fre·
quency al a rate of 1 2 rds/s with the preprogrammed settl·
ing time of 60 ms. 11 you are monitoring a slowly changing
ac signal or scanning similar signals, you can minimize the
settling time to achieve up to 330 rds/s. Keep in mind Ihat
the input could change drastically before your readings
would indicate a large change.
For signals of less than 400 Hz the input filter is connected
in series with the ac converter to slow the signal response
to the A to 0 converter. Again, you can override the
preprogrammed settling time of 800 ms but bewarel large
changes in the input signal level may let you read numerous
wrong readings while the ac converter and the input lilter
lire settling to a final value.
The 3456A can read resistance measurements as last liS
dc volts up to the 1 0 K ohm fange. but you can not use Ihe
input filter. Above the 1 0 K ohm range, additional settling
time is required fo, stable ,esistance measurements. 11 all
you want is an indication of the actual resistance above
1 0 K ohm, you can override the preprogrammed settling
times. However, just considering Ihe size of the resistance
you are trying to measure and any associated stray
capacitance in Ihe measuring circuit. you may want to add
settling time between reading for besl accuracy.
Offset Compensated Ohms is very slow ( 1 0 rdslsl by com·
parison with a standard ohms measurement. 8UI, for a
single input reading. il is the 'astest way to measure low
value resistances accurately in the presence of thermally
generated voltage ollsels.
8efore leaving the topic of Speeding the Measuremenl Cy·
cle, it is important to ask why you would use 100 or 1 0
PLC integration times. The answer is increased accuracy
and reduced internal noise. If you really need all the ac·
curacy and repeatability you can gel, the 100
place you want 10 be.

SPEEDING THE READING TRANSFER

You might make many accuracy compromises to attain an
acceptable reading rate and still nOI be able to read as fast
as you could. Why? Because you have nol optimally handl·
ed the transfer of data over the bus from the 3456A to
your computer,
let's lake a closer look at the reasons for fast reading rates:
Model 3456A
CYCLE
respectively (60 Hzl.
PLC
is the
PLC
)
J