Hi there,
My prototype is about a battery energy storage system. I am seeking to achieve a state of charge balancing among the cells by controlling and monitoring each individual cell. The cells are connected together in series by integrating each individual cell with L-bridge while each pack of cells is integrated with H-bridge [please, see my papers]. The cells will be charging and discharging through the grid. As a result, a multi-level converter (DC-AC and AC-DC) is achieved. My papers: 1) https://ieeexplore.ieee.org/abstract/document/8572728 2) https://ieeexplore.ieee.org/abstract/document/8917245 The best way I have tried (I am not sure if this the easiest and the optimal way): Using the Op-Amp chip as a differential amplifier. I have followed this article [https://circuitdigest.com/microcontroller-projects/multi-cell-voltage-monitoring-for-lithium-battery-pack-in-electric-vehicles]. I have tested many types of Op-Amp chips such as LM833P, LM2902N,..., but I can't read higher than 55 V. So, I tried to follow the circuit of INA117P chip (5 resistors) with LM833P [see Fig. 1]. As a result, I can read the input voltage up to 90 V but the output voltage value resolution is very low when the input voltage is higher than 90 V or lower than 3.5 V. The analog-digital converter (ADC) chip is connected with the OpAmps outputs to read the cells' voltages. The failed ways I have tried: 1) Using ADC in a direct way. To read the cells' voltages, an ADC chip needs an analog signal and ground of each individual cell. Thus, A) when connecting the positive (+) and the negative (-) side of each cell to its related ADC input pin and the ground pin, respectively, a short circuit will happen across the cells [see Fig.2a]. B) when using the same previous way but with only one ground (the ground pins of the microcontroller and ADC chip are connected to the negative side (-) of the last cell), the voltage applied on the ADC chip will be equal to the total cells' voltages (from the target cell until the last cell)... ADC chip can't read higher than 5 V (the reference voltage)...etc 2) Voltage divider method. Voltage divider way can't be used where we can't know how many cells will be connected in series with the target cell since each cell is integrated with L-bridge (each cell could be ON or OFF). So, this way will be so complicated.
Any help would be highly appreciated.
Hi Ashraf Bani Ahmad, my first idea is: Just dedicate one microcontroller (ATtiny25 or 45 or 85, for instance) with an internal ADC and an internal reference voltage to each cell. The internal reference voltage is usually about 1.1 V, so in order to measure the cell voltage you need a voltage divider = 2 Rs. Supply for the microcontroller is taken directly from the cell (there are many microcontrollers with a supply voltage range of 1.8 V to 5.5 V). Connect all microcontrollers plus a master by an SPI chain with galvanically isolated connections.
You could use 2 optocouplers per microcontroller (inputs SCK, DI; output DO connected to an optocoupler of the next microcontroller) but probably they would be more expensive and would take more space than the microcontroller itself.
Therefore, if the electrical potentials of the cells don't change rapidly resp. if rapid changes are known beforehand and do not occur too often, I would try to realize the SPI connections by capacitors (with the necessary Rs and diodes on the side of the receiver). One drawback is that you have to encode your data in such a way that the numbers of "0"s and of "1"s are equal in each value because you cannot transmit the DC component of a signal via a capacitor. Just use a look-up table to blow up each ADC result to a 16 bit value with 8 zeros.
Another drawback is that the SPI chain is broken if one cell (or one microcontroller) fails. A remedy would be a star shaped network but then the master would need n + 1 pins for n cells (1 x SCK for all, n x individual DIs).
So, per cell you would have 1 microcontroller, 2 Cs, about 6 Rs, (and 4 diodes, if the integrated diodes at the inputs of the microcontroller aren't sufficient).
Dear Joerg Fricke
Thanks so much for your detailed reply.
I will look through your first idea.
I would consider using analog multiplexer based on optomosfets like
Toshiba TLP179D
Thank you for answering my question Nikolay Pavlov. I solved the problem by using INA117P.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity