Erratic analog inputs .

[timseven]

Hi, everybody. I am using the 16F88 pic for my first time and I want to read a voltage generated by four infrared leds as sensors into inputs (RA0-RA3). I have read somewhere that these analogs pins are limited to 240 mvolts and I don’t know if it’s true. I setup the hardware and ran a test program and after solving some initial problems I have the circuit working but not as I expected it to work. When I grounded (1k resistor) all four inputs I should have nothing as readings and I got something (about 45 to 5 mvolts) very erratic. Here is the test program for you to evaluate. Please, help me find where I am wrong.Thanks.


OSCCON = $68 'Sets the internal oscillator to 4 Mhz
ANSEL = %00001111 'set RA0-RA3 as analog inputs

DEFINE adc_sampling 50 'Set sampling time IN micro sec.
TRISB = 0 'TURN PORT B as outputs
TRISA = 255 'TURN PORT A as inputs

ti var BYTE
ti = 50 '50 MILLISECONDS DELAY

A VAR WORD
B VAR WORD
C VAR WORD
D VAR WORD

READCDS:
PAUSE 1000 ' One second delay

ADCIN PORTA.0, a ' Reads portA pin 0 into A variable.
PAUSE 50 '50 Milliseconds delay
ADCIN PORTA.1, b ' Reads portA pin 1 into B variable.
PAUSE 50 '50 Milliseconds delay
ADCIN PORTA.2, c ' Reads portA pin 2 into C variable.
PAUSE 50 '50 Milliseconds delay
ADCIN PORTA.3, d ' Reads portA pin 3 into D variable.

'* * * * * * * * * Display Serially the sensor readings * * * * * * * *
PAUSE 100
Serout PORTB.1, 0, [254,1] 'Clear display
Serout PORTB.1, 0, [254,2] 'Send cursor home

PAUSE 20 'Pause 20 mseconds delay
Serout PORTB.1, 0, ["A= " , #(A)] 'Print A on mvolts
PAUSE 20
Serout PORTB.1, 0, [" B= " , #(B)] 'Print B on mvolts
PAUSE 20
SEROUT PORTB.1, 0, [$FE,$C0] 'Jump to second line
PAUSE 20
Serout PORTB.1, 0, ["C= " , #(C)] 'Print C on mvolts
PAUSE 210
Serout PORTB.1, 0, [" D= " , #(D)] 'Print D on mvolts



GOTO READCDS '********** Provisional loop Sensors test ***********

END

[Charles Linquis]

A couple of things:

You need to set up your chip to have analog inputs, justification
and Vref by setting the bits in ADCON1

The A/D input range is basically 0 to Vref (which will give you counts
from 0 to 1023).

Your diodes will probably not produce enough output voltage to
be usable without first being amplified by an op-amp.

[timseven]

Thanks for your answer. I'll add the ADCON1 = 128. I think that's what I need.
The diodes puts out about 1 VDC max. at full exposure. I think it's enough to be amplified no? I'll do the changes and let you know. Thanks.

[timseven]

I added the ADCON1 right justified or bit 7 (1) with the four combination for bits 5-4. I got the same erratic result. Something that I want to point out is that the four channels show the same voltage no matter if they are grounded or not. Something here is odd. Any idea what should I do next? Thanks

[mackrackit]

When you measure the max voltage from the LEDs with a meter what do you have.

You will need an op amp like Charles said or move some things around and use the two VREFs. But that still might be a little tight with out an amp.

[Acetronics2]

Hi,

At first glance ...

Your Vref must be rock-stable ( Pic Decoupling ... What about ???) ... now result is +/- 1 digit ... = +/- 5 mv !!! ... a little HardWare scheme would be useful !!!

I'd recommend some external Vref ... if mV looked for !

Alain