You may instead want a way to average the result. Otherwise, you end up with
the 'random' result being displayed for 1 sec, followed by another random
result, etc.
You can do this with a low pass filter, perhaps. The simplest is a resistor
from input to output, with a capacitor from output to ground. The values to
use are dependent on how much you want to smooth the output. If you want the
output to average out at 1Hz or less, then you need components so that
R * C = 0.16 = (1/(2*PI))
So, if you use a 1uF capacitor, you would need about 160k of resistance.
So, here is the circuit:
160k
sensor ---- RRR ---o---- output
|
--- 1uF
---
|
GND
Regards,
Bob Monsen
I'll also add my vote for the filter approach as a first step.
But if you ultimately decide you do need a sample/hold,
it can be done in a fairly straightforward manner without
going digital. You have a buffer amp that charges a
capacitor through a switch (4016, etc). The capacitor
voltage is monitored by another buffer that feeds your display
circuit. A timer closes the switch for a brief interval, once
per second, to charge up the cap to the present input
voltage.
The input impedance of the op-amp output buffer will be
very high, so it won't draw down the capacitor voltage
between samples. In general you need a low input bias
current op-amp for this, but I'll bet a TL082 or LF351
will be OK. You will need to experiment with the cap
if dielectric absorption or leakage gives you problems,
but I suspect a big mylar will be fine since this is not
really a critical issue for your application.
The timer is totally non-critical. I prefer simple CMOS
gate timers, but you could probably use the popular (though
I could never understand why) 555 series here if you don't
mind all the extra parts and the supply current spikes.
Best regards,
Bob Masta
DAQARTA v3.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, FREE Signal Generator
Science with your sound card!
