I want to multiplex battery voltages of different battery cells using multiplexer ICs, in this case "Mux-24" is used. See the figure below for pins. I am wondering about how to use the pins GND and V-.
If I want to use an Arduino to power the mux, should I connect 5V to V+, and -5V to V-? Or should the arduino GND go to V-? Or is the GND used as reference for the signals S1-S4?
My plan was to use one multiplexer IC for each battery pole (positive and negative), and multiplexing them as floating voltages. Do you think that is possible?
niczor:
Thanks for the reply! The cells are not connected together. Each cell is around 5V, so I have to make sure the supply is greater than 5V.
Never heard of 5V cells before - something exotic?
What is the GND used for? Signal reference or the GND of the supply voltage?
Its used as the logic input reference, and its probably assumed to be mid-rail - as I said this
is designed for mixing desks. Don't expect the chip to work without symmetric supplies (although
it might).
If the cells are floating, no need for differential measurements, just ground all the negatives and multiplex
the positive ends. You'll need to use a voltage divider on the output if the cells are more than 5V ever.
74HCxxx series analog multiplexers work upto 6V supply, note.
You normally can't do a floating differential measurement, you need some way to ensure the
voltages dont stray outside the measurement system rails (protection diodes don't kick in to 0.3V or
so, so this can matter). Easiest way as I say is to ground all the cells and do single-ended measurements.
This also means your cells aren't picking up interference and noise.
I connected V- to GND and 5V to V+. It seems to be working, but the mux output is not as precise as I would like! Not sure if it's due to the chip itself or that I am not using symmetric supply. The output is off by around 0.1 V. I have ordered some MUX506 muxes. Maybe they will be better suited for this application.
Thanks for the advice on common ground of the cells!
MarkT:
You normally can't do a floating differential measurement, you need some way to ensure the
voltages dont stray outside the measurement system rails (protection diodes don't kick in to 0.3V or
so, so this can matter). Easiest way as I say is to ground all the cells and do single-ended measurements.
If you could explain a bit more on this, I would be grateful. I wanted to multiplex the two poles as floating voltages (one multiplexer for each pole), and then compare those two signals to get the battery voltage. You are saying this is not a good approach?
Floating is bad, you have two voltages both with lots of noise pickup (the same noise pickup, but
unless you measure the two voltages simultaneously you'll see the noise on your signal)
And you have the undefined voltage problem - if the battey terminal voltages are both within the
supply rails of your ADC it will clip the voltage that's outside it.
Basically if you let the battery float its an antenna, you don't want this. You could blow up your
ADC for one thing if there's lots of static around.
. The output is off by around 0.1 V. I have ordered some MUX506 muxes. Maybe they will be better suited for this application.
Did you measure that error with a multimeter or just with the Arduino? i.e. Maybe the multiplexers are OK and the Arduino reference voltage is off?
Or, maybe the multiplexers need a "load" (a 10K or 100K resistor)?
Ok! So the cells should be grounded to the supply voltage ground of the mux in order to avoid floating voltage?
Are you trying to measure batteries in series, or are these cells independent?
The Arduino always measures positive voltage relative to its ground. You'll notice that a multimeter doesn't have a "ground".* It's got plus & minus terminals.
If the batteries are in series, you'll be connecting the battery's minus terminal to the Arduino's ground. In order to avoid shorting-out the other batteries you have to "float" the Arduino ground relative to the battery ground. (When you measure the "bottom" battery the grounds will be connected and that's OK.)
There is an internal ground and that internal-ground might be connected to the minus terminal, but that internal ground is isolated (floating) relative to any external circuitry until you connect the probe(s). So, you can connect the minus terminal of the multimeter to the positive battery terminal and nothing bad will happen... If you reverse the multimeter connections, you'll simply get a negative voltage reading. (Don't try that with the Arduino because the Arduino can be damaged by negative voltages... That is, voltages that are negative relative to the Arduino ground.)
I take the outputs from my two muxes into a multimeter and get another voltage value than the input voltages to the muxes. Right now I use a power supply as voltage source. The positive pole goes into one mux, and the negative pole goes into another.
In the end, I will have several cells which are completely independent. All of this is used for impedance measurement of the cells. See the attached figure.
I am testing some multiplexers, but am not getting the results I expected. Mux used: MUX-24
I am using two multiplexers to multiplex the positive and negative poles of a voltage source, see attached figure. One mux takes the positive pole as signal input, and the other mux takes the negative pole.
The outputs of the muxes are connected to a multimeter. However, the resulting voltage is slightly different from the input voltage. For example, if I want to multiplex 2.5 V, I get something like 2.4 V from the multiplexer outputs.
For the multiplexer chip, GND is connected to Ardunio's GND, V- is connected to GND, and V+ is connected to Arduino's 5V. The power supply outputs, which are used as signal inputs to the muxes, are floating with respect to the Arduino.
See attached figure. Since the outputs are essentially open circuited (into the multimeter), I figured no current is drawn. Therefore the ON-resistance should not make a large difference? Or is the leakage current causing a voltage drop across this resistance?
The problems in this thread were a bit off-topic from that thread. That's why I created another one. Maybe I should not have
When I use a floating external source (power supply) as input to the muxes, I get 0.1 V offset in the mux output. When I use signals from the arduino into the mux, I get 0.01 V offset. So somehow it is a problem when I want to multiplex floating external signals?
You have been told about referencing the external cells to the logic ground. That is you should not have floating signals. I suspect that is the reason for the voltage discrepancy. But until I can see all your circuit I can’t be 100%.
Why will you not post the compleat circuit like you have been asked for?
Here is a more clear circuit diagram. As voltage source (the battery symbol) I am currently using a power supply, but in the end battery cells should be used. I measure 2.5 V into the multiplexer inputs, and 2.4 V on the output. The output is open circuited (into a multimeter), so there should not be significant voltage drop from current.
niczor:
My plan was to use one multiplexer IC for each battery pole (positive and negative)...
Why?
You say that the cells "are not connected together".
So can't you just connect the negative battery terminals together, and only mux the positive terminals.
Remember that battery voltage must be within the supply rails of the chip (MarkT, post#1).
Leo..
