ADCIN & 16F870 - how to avoid non linear conversion?

[HenrikOlsson]

Hi,
OK, that's a great value because then each bit in the ADC result equals exactly 4mV but your numbers still doesn't add up....

If you input exactly 3V you should get a value of 750 +/- a couple of bits - not 678 as you're apparently getting.

Have you checked the VRef voltage with a scope? Perhaps it oscillates or have a lot of noise?

What if you replace the 10k pot with a 1k pot, and put a 10nF capacitor between the ADC input and GND. Or, what if you increas the sampling time? (I'll admit, I don't really have this sampling time and Tad stuff clear in my head at the moment, need to do a little reading)

Once you get a correct ADC result for the voltage you're feeding it you can increase the resolution by oversampling. If you take 40 samples, add them together and divide the result by 10 you have a value reading in mV (well, almost - there's 4mV "missing" at the top).

adval = 0
For i = 0 to 39
  ADCIN 0, adval
  Voltage = Voltage + adval
  PAUSEUS 100
NEXT

LCDOUT $FE, 1, DEC Voltage/10000, ".", DEC3 Voltage // 10000

/Henrik.

[CuriousOne]

Ok I've double checked wiring and so on. When getting 1023, I was connecting the upper end of pot to +5V, not to MAX output. Now, when I connected it to MAX output, it reads 987, instead of 1023, so I assume, there might be some more register settings required?

[CuriousOne]

Vin	Raw	Delta

0.1	10
0.5	88	78
1	213	125
1.5	335	122
2	460	125
2.5	580	120
3	700	120
3.5	807	107
4	927	120
4.1	960	33

These are raw ADC reads according to different input voltage.

[HenrikOlsson]

Hi,
First of all. With a VRef of 4.096 you can not input a voltage HIGHER than that, so the 4.1V (although very close) isn't a valid test. With that said something is seriously wrong - either with the code, the setup, or the hardware.

And again, have you actually looked at the reference voltage AND at the input voltage with a scope to verify that they are clean? Simply measuring them with a voltmeter may not show the error.

Instead of using the pot across the reference voltage, try using a low impedance source, ie a variable DC supply, to feed the ADC and put a 10nF capacitor from the PIC pin to GND to act as a noise filter. See if that makes it any better. Also, take a couple of readings and average them to further clean any noise.

If the above doesn't help then try another channel on the ADC or try another PIC. It's possible, but not likely, that something got damaged when you overloaded the input.

/Henrik.

[CuriousOne]

It is NOT 4.1V, it is 4.096v, I just wrote 4.1 to save space.

Yes, looked with scope and precision (3 digits after decimal point) multimeter.

Will try another channel and another pic, as well as another voltage supply.

[CuriousOne]

Set scope to 50mV resolution and AC input. Found an 120mv max noise pulses, with 3.35hz following frequency.

[HenrikOlsson]

Where?
On the power rail, at the output of the voltage reference, at the ADC input or all of the above?

If the noise is only present at the ADC input of the PIC then the problem is probably too high source impedance (that 10k pot you're using) and the 3.35Hz you're seeing is how often the ADC is sampling the input. Or is this with a low impedance source?

[CuriousOne]

Yes it was with 10k pot, measured at ADC input. no other tests so far.

Actually, I went with ADC route because I was not able to make BH1750 light sensor to work with picbasic, and since I badly need light measurement, I've decided to do a circuit with photoresistor and ADC, but it appears to have it's own bunch of problems.