I an working on a project that requires several 3.7V batteries to be connected in series.
I would like to use a microcontroller for each battery to monitor and to make some other adjustments as well. Each microcontroller board will be powered from it's battery which will be connected in series with the other batteries.
I also would like to have an RS485 communication between all microcontroller boards and also with a computer.
My question is regarding the RS485 connection between the uC boards. Will they have any problem to communicate each other due to different ground or not ?
I'm not expert on RS485 but I don't think this will work. I believe you need a shared ground for RS485 comms to function but I could be wrong on this as the bus is differential.
One way might be to put all RS485 transceivers on the same voltage and then drive the I/O to the transceivers via Optos to disconnect the comms voltage from the source. Another way might be to use some form of current loop e.g. 4-20mA then the voltage is irrelevant.
With that circuit I think you will more than likely blow the micro, this could also end up in unequal discharge on your batteries
Can you advise what you trying to do ?
How about using a multiplexer or two and then strobing round your batteries ? Then you have a break in the circuit and less chance of creating a short across the battery through the micro.
What i am trying to do is a battery monitoring system. I want to monitor a few batteries which are not close each other. All batteries will be connected in series and each of them must have it's own microcontroller for the purpose of this project.
I really want to avoid multiplexing since i want to use one microcontroller for each battery.
I should use an isolated rs485 Ic or use optocouplers as Benj suggested. However the isolated RS485 ICs are too expensive and therefore i am trying to find another cost effective way to accomplish this circuit.
If the distances are not too great, and your UART data rate low, then you could probably use simple (darlington) opto-couplers between the Tx and Rx on the processor side and wire OR’d together onto a single communication signal wire on the common bus side, with it’s own ground and power supply rail.
I did a really quick test with some draft boards that i had from a previous project.
I used 1.5V batteries for my test and connected all in series. I connected the MCU boards as the attached picture shows and i powered all the boards from a single power supply 12V. Note that each board has build in voltage regulator for 5V.
Then i wire up the RS485 (daisy chain of all boards) and i connected as well a PC with a Terminal software to observe the data.
I could easily get all the information of each boards with out any problem. Do you thing that it will cause a problem in the feature this circuit configuration that i used?
It probably works with 1.5v batteries due to a few differences, such as the total voltage is less than your supply voltage,
And the RS485 ground differential has a maximum less than 8 volts.
Depends on the exact RS485 driver you are using, but max ground offset is usually in the -7 to +10 volt region, not the over 22 volts you have with the 3.7v batteries. That is when the boards are powered as the original circuit. With them powered in parallel all should be well.
I'll draw up circuit on Monday of the opto solution suggestion.
Btw, I've only been considering the comms side, as Steve001 said earlier you will need to check very carefully as to what you are doing at the battery monitoring side. With the new circuit the monitor inputs will need to cope with higher voltages.
Leigh
I would look at these to isolate the battery voltage and prevent a back-feed through the micro and shorting out your batteries.
Its easy done and makes a mess !
I am working this project in my mind and to be honest i can't see the reason for not connecting each microcontroller across each battery without any isolation circuitry. The ADC of the Micro will go at the positive of each battery and the ground pin of the Micro will go at the battery's negative pole.
I think if i would use only isolation on the communication between the microcontrollers i would be all right. Any thought about that?
I tried to quickly sketch the circuit that i have in my mind right now. Please see the attached PDF file with the circuit drawing.
Hi George,
Regarding the isolated RS485, just don't connect anything from the bus side to the processor side, see VDD!
I've not much experience with batteries, unlike Steve001, so you ignore his advice at your own peril
Leigh
To prevent signal on the RS485 bus from "reflecting", you should add termination resistors to each end of the bus.
Missing termination or unbalanced termination could make the bus act weird on higher data speeds or on emi disturbance.
RS485 is differential signal, so no common grounding is required between nodes.
I made a mistake when i was drawing the schematic.. I already replaced the pdf file with the correct version.