I have an Arduino Uno project successfully running which uses a DHT22, an Adafruit TL2561 and a SparkFun ML8511 sensor array to provide lots of information about a museum environment.
However the UV values are always low even in situations where the UV should be high. The count numbers for UV Intensity do move up but the range covered is small. Also, when in a low UV or no UV lighting situation the uv Intensity is often a negative calculated number of about -0.9.
I am wondering if the internal 3.3V which the sketch assumes to be 3.3 V may be a bit lower than that. If that is even a possibility,how do I measure the 3.3V reference?
Would I get reliable results if I measure from one of the 3.3V supply points on my data-logger shield which is attached to my Uno while all the sensors are also attached and actively drawing power? I assume that I would need to measure voltage under working load.
Or is there a way to poll the Uno itself to tell what the actual 3.3V reference is?
You should never assume that 3.3 volt from a regulator is exactly 3.3 volt. Especially if it provides current to various loads. The voltage might differ depending on the current draw. (though it doesn’t seem to be much if it is only those devices) You should measure the voltage under the various load conditions. Get yourself a descent voltage meter and start probing.
Without any code for us to view it is impossible to say anything about why you get negative numbers or why it doesn’t range al that much.
I was hesitant to post very long code until after looking at the 3.3 or not Voltage issue. I have used a multimeter to read the voltage when all is connected and when all is not. There is quite a difference when using Uno R3’s. I tried several Unos and with the DHT22, LS2561, ML8511, SD card/RTC and a backlit LCD display attached, both Unos showed varying voltages above the 3.3V reference. My original Uno fluctuated under load (possibly in sync with writing to SD card) from 3.81 to 3.72. The other Uno was more in the 3.54 V range under the above load. As bar boards they were delivering 3.3 to 3.4 Voltages.
For either board when I used the actual 3.3 reference Voltage (3.78V) the calculated UV intensity became more nearly “real world” for my current lighting situation.
I tried the DUE board and found that the 3.3 V reference voltage with the same sensor and datalogger and LCD display was much more stable and closer to the stated reference voltage with a constant value of 3.34V.
So for now it looks as if I must go with the Due for this even though the longer form factor is a potential problem. However, before I leave this problem I will strip back the sensor load to just the ML8511 with no datalogger or lcd and see if the reference voltage is so random or if it truly is a matter of full load perturbility.
Can you tell more specifically which devices are fed from the 3.3 volt regulator? (It’s not really a voltage reference, atleast not intended to be precise; it only burns away excess voltage) So then it is possible to locate the datasheets. I made that load presumption based on the sensors you named, for which the datasheets only mentioned a few miliamp each. Lcd and sdcard and what else powered by 3.3 volt might make the difference. I just don’t know how much that is.
I don’t even know out of the top of my head what the 3.3 volt regulator can deliver.
20 x 4 back lit lcd QCC2004A with i2C back pack UC-146 (unknown manufacturer, Chinese) possibly supplied by Electric Freaks but also possibly an Ebay purchase but from China
You may not be familiar with the ML8511 UV sensor which produces a variable output voltage depending upon the UV amount falling on it. The Sparkfun sketch uses the 3.3 V as a reference voltage in order to calculate the amount of UV. I connected that to the 3.3V supply used to power the DHT22 and TSL2561.
I don’t have a data sheet for the data logger and can’t tell whether it is a 5V or a 3.3V shield. My thought is that it is a 5 V shield. It can be viewed here: http://www.play-zone.ch/en/catalogsearc … %5B%5D=356
Well, you can always try to measure the sensors when you have disabled the backlight temporarily, or are not storing data on the sd card. So at times of reduced current load.
I find it odd though if that 3.3 regulator goes up to 3.8 volt or something. That does not make sense.