[TriEmbed] measuring analog voltages near the battery voltage

[Alex Davis]

I’m testing out LEDs as a substitute for the expensive UV photodiode in my UVI business card. I’ve found a good one, in terms of visible light blindness and UV-A response. However, it outputs a bit more voltage than the internal 2.56 vref of the ATTINY85 above a UVI of 5. It works great with a 3.3v ref, but even a fresh coin cell is only 3.2v.

I can knock it down with a divided negative feedback op-amp arrangement, which isn’t expensive, but I’m trying to make this as cheap as possible.

I tried a 10K 1/2 voltage divider, but that was too much of a drag on the tiny current from the LED. I have a feeling going higher than 10K is not a good idea with the ATTINY ADC.

It is possible to have the ATTINY read it’s own voltage, and then use that with Vcc connected to AREF, but I know from practice that the bandgap measurement has pretty poor accuracy.

Looks like external voltage reference is no-go; they’re way too expensive.

So, op amp it is? 1.0K 1% resistor is cheap, and a 1.2K 1% is a few pennies more. That would give me -0.833 gain which would work within the 2.56V AREF.

Alex

|\ |  (¯  \/ |¯\  |V| |\ ¯|¯ |¯) | \/ | | | |¯\ (¯   /¯  /\ |V|
|-||_ (_  /\ |_/ @| | |-| |  | \ | /\ |^| | |_/ (_ . \_  \/ | |







-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.triembed.org/pipermail/triembed_triembed.org/attachments/20180127/8ee9a8ca/attachment.htm>