Go Down

Topic: AnalogRead - readings are erratic (Read 9680 times) previous topic - next topic

cameronasmith

Thanks Guys,

So what do you recommend in this situation? Should I increase the gain of the INA125?

Thanks,

Cameron.

Grumpy_Mike

No the point was that it is all working as it should be and you should expect no better results. In fact you can not get any better results in terms of stability of reading, no matter what you do. It is how electronics works.

cameronasmith

So, how do people use these weights to measure 1 gram increments?

It came out of a unit that does this feature.

Is it that I just have to do some crazy programming?

dlloyd

Check the datasheet.

It says a single external resistor sets any gain from 4 to 10,000.

You'll need to set the gain so that you get full ADC code swing
for the measurement range you require.

That way when you take a reading, +/- 1 gram will give maximum response.

cameronasmith

Hi dlloyd,

When I try increasing the gain to reasonable readings I can only measure 44 grams, and I need to measure around 200 grams.

What would you suggest?

Thanks,

Cameron.

dlloyd

#20
Apr 08, 2014, 10:02 pm Last Edit: Apr 08, 2014, 10:05 pm by dlloyd Reason: 1
It would be great if you had posted a schematic, however:

If 200 grams is the maximum you'll ever need to measure,
and 44 grams gives code 1023,
then your gain is set too high.

In this case, you would need to lower the gain so that 200 grams gives code 1023.
Then you could measure any change of 0.2 grams with 0 to 200 gram range.

cameronasmith

Hi dlloyd,

I don't have the schematic in front of me right now, but when I get home i'll send it.

The problem is, that when I lower the the gain the readings of the grams is too close to the drift values from the Analog sensor to get a good reading.

Oh, and the readings are *something* like this:

600 to 860 when the gain is set high

200 to 860 when the gain is set low.

Thanks,

Cam

Grumpy_Mike

Quote
In this case, you would need to lower the gain so that 200 grams gives code 1023.
Then you could measure any change of 0.2 grams with 0 to 200 gram range.

That would only work if the response of the strain gauge were linear. Maybe you need a linearisation amplifier.

cameronasmith

Do you think if I get an ADC with a higher bits like 16 bit it might work better?

Also, what is linearisation amplifier?

Thanks

dlloyd

Quote
The problem is, that when I lower the the gain the readings of the grams is too close to the drift values from the Analog sensor to get a good reading.


Make sure you apply the maximum weight you'll ever need to measure prior to setting the gain.

This procedure sets the span.

cameronasmith

Hi dlloyd,

That's super interesting :

Quote

Make sure you apply the maximum weight you'll ever need to measure prior to setting the gain.
This procedure sets the span.


I've never heard that before? So depending on the weight I have on the scale will adjust my reading when I change the gain?


Grumpy_Mike


what is linearisation amplifier?

See attached, it applies to strain gauges as well:-

dlloyd

Also look at this from the datasheet:

Connecting VREFOUT (pin 4) to one of the four available
reference voltage pins (VREFBG, VREF2.5, VREF5, or VREF10)
provides an accurate voltage source for bridge applications.


The datasheet shows pins 4&16 shorted (VREF10).
I think you should try VREF5 (jumper pins 4&15) after reading this:

I have 2 load cell parts, one for 1kg and the other for 200kg. The 1kg load cell works good with pins 4&15 shorted, but the 200kg load cell gets the precision reference to drop a little.

dlloyd

Quote
I've never heard that before? So depending on the weight I have on the scale will adjust my reading when I change the gain?

That's what the stain gauge is doing - varying it's output resistance which varies the analog voltage.
You need to adjust the gain with maximum weight (max voltage) to set the full scale reading.
Otherwise,  you're adjusting full scale reading with 0 gram (minimum voltage).

cameronasmith

Quote
I think you should try VREF5 (jumper pins 4&15) after reading


Hi dlloyd,

Thank you for the suggestion. I have the pin connected to 15 because:

Pin 16 is for 10 volts

and Pin16 is for 5 volts.

Do you think if I get a ADC with higher bits I would get better readings? More gain and more results or do you think the drift would just match the increase?

Thanks,

Cameron.

Go Up