Hi,
I need to measure signal that changes between -100mV to 100mV,
what I think about is to add dc offset with op amp similar to this :
This might do what you want. I suspect it may be easier to DIY or find an evaluation board from Analog Devices or similar.
Depending on your requirements you may be able to achieve the same thing with just two 10k resistors forming a potential divider between the Arduino's power rails and a capacitor. The DC offset is then Vcc/2.
There is an example in the following post DTMF decoder library - #75 by 6v6gt
For how to design an op amp for DC offset see this application note from TI
Is it a slowly changing signal?
yes, very slowly, I need something shelf product , no time for new design
Good luck with that quest, but I can't help thinking you might spend longer looking for something that exactly matches your requirement than the op amp and handful of resistors.
In that case use an external ADC like this
Use it in differential mode and connect a stable positive voltage (voltage divider and capacitor to ground) to the other input
A similar question, but if the OP objects, we can move this to a new thread.
I have a need to take a current shunt signal (-10mv(charging) -> +100mv(discharging)) and read it with a Nano, or other 10-12 bit 0-5V ADC. What's the simplest way to do this? It will be an adaptation (kludge) of an existing single-ended 0-5V input my friend has already committed to PCB, so I'm looking to basically shift and reduce the signal range. No negative supply or other complexity allowed, basically I'd like to piggyback a simple daughter board and have four connections - in, out, and of course 5V and GND. I2C, SPI, and other digital-domain addons have been excluded from consideration. Ideas, suggestions? My analog is rusty, but hints will do, as I can protoboard it until we get it right.
...working through the above-linked TI document shortly, so this may be a wasted post.
Yes, so clearly I need a hint. That guidance clearly leads to a direct connection from my shunt to the + input of the opamp, which of course means, when charging, we apply an input signal outside the supply range; I'm not sure that's a universal no-no, but it's not a normal thing to do with unipolar-supply configurations.
So, suggestions, folks? Am I really headed for some form of iso-amp, or is there another solution?
Thanks in advance!
Never mind. Found my "Horowitz and Hill". Taking a read. Thanks!
I'll be interested to see how you accomplish this without a negative supply. It is possible but it's far simpler if you add a -5V inverter.
I thought I had a post about an arduino-based charge pump voltage inverter that I built but I can't find it now.
That's what I was thinking, but TBQH I haven't done any serious analog design in 25+ years, and in those days we used AD210 isolation amplifiers, which were super, but super$$; I had hoped for innovation in the meantime. We'll see.
Thanks for your input!
Connect your signal to the negative side of a AAA battery and take the positive side to your adc input. Measure the battery volts to determine the offset voltage. Hardly any current taken so should last ages.
Not likely. Ultimately, this has to go into a commercial design, so we need to work towards viable solutions that don't involve batteries, to avoid the customer laughing too hard. But thanks. I suspect we're headed towards a small charge pump, if there's no OTS solution that pleases my friend.
Another option that could bear some investigation would be making this a parasite on something like a MAX232(or modern equivalent), which has a built-in charge pump arrangement for it's own needs. They already have some RS232 comms going on, so it may be feas. TBD.
I was replying to the OP actually who seemed to want a breadboard implying something temporary.
ahh. sorry.
If there's already one in the design, that should work as long as you filter it. ICL7662 charge pumps are cheap, but they were never that great to begin with.
I did some experimenting with a capacitive charge pump circuit driven from an arduino pin (that's the one I thought I had written a blog post about) and IIRC, I was able to get about -3V at 5mA which might be enough depending on the op amp you are planning to use.
Using the CPU pins for toggling a charge pump isn't an option, unfortunately, but I'll discuss the charge pump idea with him. The RS232 charge pump idea may work out, as they're not too worried about exactness for the current readout, so a little noise may not be an issue.
Thanks.
