Signal Buffering - DIY Analog Hygrometer

[CoffinDodger]

Hello everyone. Recently got into FDM printing and found myself in need of a way to measure humidity in the workspace. I have a little humidity sensor for a printer I tore apart a while back (HSU-07P) and a cheapy Southwire analog meter. I figured I'd just combine the two. Of course I found the sensor doesn't supply enough current for the meter movement. At just 4uA it doesn't have a hope in moving the needle with a full scale current of 180 uA (measured). So i went to putting together a little unity gain buffer. In the lab I'm limited to LM358Ns, and TL0721Ps. I was going to go with the TL072s but found a note regarding the input voltage for unity gain to avoid less than 1V and to avoid less than 4V. The sensor's signal is 0-3V. I'm a little all over the place now with how to go forward.

- Do I ignore the datasheet and do it anyway?
- Do I use a BJT/FET as a pre-buff to the buff?
- Do I do-away with the opamp entirely and go with a FET? (isn't a TL072 just a fancy pack of FETs anyway?)

The hope was to leave the output to the meter as 0-3V to make mapping the values on the scale a bit easier. I may just have to abandon that entirely. Any advice for how to proceed is welcome.

EDIT: Is the common mode range only a concern for oscillating signals and not steady state signals?

[mlamoore]

The common mode range definitely matters at DC as well as AC.

Looking at the TI TL072 datasheet, the supply voltage is rated for either 10-30V or 4.5-40V depending on version, the input common mode range is V- + 1.5V to V+ (V- + 2.5V to V+ - 1.5V for best CMRR), and the output voltage range is V- + 1.03V to V+ - 0.965V under the worst case datasheet conditions.

This is definitely a part that was meant for dual power supplies. If you power it with V+=+5V and V-=-5V it should do just fine for your 0-3V signal with plenty of safety margin.

It has a worst case input offset voltage of +/-5mV which will add a small error to your measurements.

If you only want to use a single 5V supply, not dual supplies, then you will need to look for a different op-amp with an input voltage range that includes the negative rail. The LM358 running on 5V has an input common mode range of 0-3.3V and can provide output down to 20mV (with the load resistor pulling down instead of up), but it has a higher input offset voltage of up to 9mV.

Of course, there's tons of better options amps available if you order them online, but the two you have available are cheap and common.

[Kleinstein]

The TL072 is not really happy with the low supply and it also takes more power. So the LM358 is the better and easier choice.
The sensor may even get away without any buffer. Discrete JFETs have very large scattering the DC operating point - so not a good choice. It would be more a BJT that could work if one does not like OP-amps.

For the relatively low power (some 1 mA for the sensor and maybe 0.5 mA for the amplifier) a low power linear regulator is likely the better choice than a switched mode converter.