Cell Simulator Circuit design

[Alok_Kumar]

Hi ,
For testing the BMS I want to design the 24-channel Li-ion battery cell simulator
In starting my thought was to use a potentiometer based resistive divider network but facing problem of voltage geting chage if i chage resistance of one channel.
my requirment is to get voltage range form 2.8V to 4V in each channel individualy i want to change.

can any one suggest or Help me to design the circuit to simulate voltage sensing to BMS.

 

[BradtheRad]

For sensing I imagine one or two led's across each channel. They contain their own built-in voltage regulation. They serve as visual indicators. With safety resistors at each led you can adjust them individually.

 

[Alok_Kumar]

yes in that way we can get but I don't want to change the voltage of each channel when I change the variable resister of each channel because all are connect in series.

For sensing I imagine one or two led's across each channel. They contain their own built-in voltage regulation. They serve as visual indicators. With safety resistors at each led you can adjust them individually.

 

[D.A.(Tony)Stewart]

To design anything you must have measurable specs. !

For a BMS , there are two states; charging and discharging.

1. For charging only OVP in each cell is needed while the string is regulated by CC/CV which must be specified.

2. For discharging, only UVP is needed to bypass loading on the weakest cell(s).

For simplicity let's consider specs for only 1.
Pls list all specs. The BMS will have a Pmax limit. for each cell.

Actual BMS designs are much simpler than BMS Simulator designs which are sold for several $k and have pages of specifications and are quite complex beyond the scope of this question.

Pls advise.

 

[Alok_Kumar]

To design anything you must have measurable specs. !

For a BMS , there are two states; charging and discharging.

1. For charging only OVP in each cell is needed while the string is regulated by CC/CV which must be specified.

2. For discharging, only UVP is needed to bypass loading on the weakest cell(s).

For simplicity let's consider specs for only 1.
Pls list all specs. The BMS will have a Pmax limit. for each cell.

Here i am not designing the BMS
i want to design Li-ion battery cell simulator for voltage sensing of BMS i don't want use actual battery cell for voltage sensing to BMS
in cell simulator i want to create UV(2.8) ,OV(3.9) in each channel

 

[FvM]

What are asking specifically? An universal cell simulator has galvanic isolation for each channel, e.g. by DC/DC converter, isolated digital interface for setpoint transmission. Full featured cell simulator has impedance emulation and bidirectional current capability and is able to simulate charge, discharge and cell balancing. For your purposes, a programmable low power voltage source may be sufficient.

 

[D.A.(Tony)Stewart]

You must outline the scope of your idea, purpose with functional requirements before any realization is possible. (not just 2.8 to 4.2 V)

This may help develop a plan.

Battery Management Systems (BMS)

Learn how to design and simulate battery management systems. Resources include videos, examples, and documentation covering motor modeling, control design, and automatic code generation.

www.mathworks.com

Do you have models for good and bad cells with memory effects, ESRs and C values?

 

[Alok_Kumar]

What are asking specifically? An universal cell simulator has galvanic isolation for each channel, e.g. by DC/DC converter, isolated digital interface for setpoint transmission. Full featured cell simulator has impedance emulation and bidirectional current capability and is able to simulate charge, discharge and cell balancing. For your purposes, a programmable low power voltage source may be sufficient.

I just want to simulate voltage for BMS and create UV and OV condition I don't want simulate charge, discharge and cell balancing
would you tell me how I create 24 channel for voltage sensing from programmable low power voltage source

--- Updated Jul 16, 2024 ---

You must outline the scope of your idea, purpose with functional requirements before any realization is possible. (not just 2.8 to 4.2 V)

This may help develop a plan.

Battery Management Systems (BMS)

Learn how to design and simulate battery management systems. Resources include videos, examples, and documentation covering motor modeling, control design, and automatic code generation.

www.mathworks.com

Do you have models for good and bad cells with memory effects, ESRs and C values

Here I am not doing battery modeling
I want to design hardware circuit which create voltage from 2.8V to 3.9V in each 24 channel to give sensing to BMS controller

 

[D.A.(Tony)Stewart]

For what BMS?
Are you trying to test a BMS charger only? under load too?
How much current?

It seems you are trying to simulate a string of battery voltages over this voltage range without any thought of the memory effects of voltage (R1C2) and the ESR * capacity (R2C2) and assuming your future tests of using a linear voltage source of unknown impedance spectrum is good enough. Any BMS is a control system and will have ideal responses with a resistive programmable voltage source but possibly unknown effects with a highly reactive cell voltage.

So your end purpose is still unclear and makes a big difference.

--- Updated Jul 16, 2024 ---

You indicated;
" design hardware circuit which create voltage from 2.8V to 3.9V in each 24 channel to give sensing to BMS controller "

If I understand, you want to simulate a string of programmable battery voltages to test a BMS but ignore the battery characteristics.

Is that correct?

If so, then before we talk about hardware details, what specs? budget?

e.g. Cell String Simulator (specs)

BATTERY SIMULATOR 1200 | Bloomy

12 Isolated 500 mA Channels in 1 RUThe Battery Simulator 1200 provides twelve programmable cell channels that sink and source current for testing battery management systems (BMS) and battery-sensitive electronics. The cell channels are isolated to enable series connections for simulating battery...

www.bloomy.com

--- Updated Jul 16, 2024 ---

For requirements; you may consider, user interface, resolution accuracy, ESR , protection, isolation limits resolution, tolerances, noise limit, power limit.

Current mirrors are useful for programming remote voltage references with fixed R's.
Opto isolated voltage mirrors are used in SMPS feedback.
Voltage or current DACs for digital references.

It seems you are hunting for simple hardware architecture ideas before specs.

 

[D.A.(Tony)Stewart]

Are you wasting time with trial & error ideas when you should define your MUST-HAVE specs vs NICE-TO-HAVE specs.

Do you need 24 isolated voltage sources or non-isolated ones?
Do you need 4 digits of accuracy or 3? Then 10% pots are out of the question.
What input options are you limited to? Analog or digital or software controlled?
How fast do settings need to be changed?
What is your budget for time and $?
Will that be just for your education of profit ? Why not BUY vs MAKE?

Engineering Specs determine how to evaluate any design and make tradeoffs early.

 

[FvM]

An precision spec is surely required to choose topologies and components. Also dynamic demand (bandwidth, slew rate) is of interest.

Presumed you are designing 24 channels wired as a fixed series string, you might come away with a non-isolated design because the maximal output span is only 96 V. Feasibility depends on precision and accuracy requirements.

 

[KlausST]

Hi,

If I understand correctly you want to EMULATE individual cell voltages of cwlls connected in series.

If so I recommend to use a single voltage reference, feeding all the pots.
The pot output should feed individual OPAMP circuits that ADD the pot voltage to the output of the previous OPAMP circuit.
The first OPAMP obvioulsy has to reference to GND.

Circuits 1...n:

* Circuit 1: U_out(1) = U_pot(1) + GND = U_pot(1) + 0V = U_pot(1)
* Circuit n: U_out = U_pot + U_out(n-1)

Mind: the OPAMPs need a power supply that is higher than the maximum expectable output voltge

Klaus

 

[Alok_Kumar]

For what BMS?
Are you trying to test a BMS charger only? under load too?
How much current?

It seems you are trying to simulate a string of battery voltages over this voltage range without any thought of the memory effects of voltage (R1C2) and the ESR * capacity (R2C2) and assuming your future tests of using a linear voltage source of unknown impedance spectrum is good enough. Any BMS is a control system and will have ideal responses with a resistive programmable voltage source but possibly unknown effects with a highly reactive cell voltage.

So your end purpose is still unclear and makes a big difference.

Are you wasting time with trial & error ideas when you should define your MUST-HAVE specs vs NICE-TO-HAVE specs.

Do you need 24 isolated voltage sources or non-isolated ones?
Do you need 4 digits of accuracy or 3? Then 10% pots are out of the question.
What input options are you limited to? Analog or digital or software controlled?
How fast do settings need to be changed?
What is your budget for time and $?
Will that be just for your education of profit ? Why not BUY vs MAKE?

Engineering Specs determine how to evaluate any design and make tradeoffs early.

I cant't afford boomy cell simulator its a costly one and also offer only 12 channel in one simulator
I just want to test 24V channel BMS with 3 digit accuracy .
as of now only analog control if enough for testing .
I just want to create string of individual adjustable voltage with any one either isolated or non-isolated

--- Updated Jul 17, 2024 ---

Hi,

If I understand correctly you want to EMULATE individual cell voltages of cwlls connected in series.

If so I recommend to use a single voltage reference, feeding all the pots.
The pot output should feed individual OPAMP circuits that ADD the pot voltage to the output of the previous OPAMP circuit.
The first OPAMP obvioulsy has to reference to GND.

Circuits 1...n:

* Circuit 1: U_out(1) = U_pot(1) + GND = U_pot(1) + 0V = U_pot(1)
* Circuit n: U_out = U_pot + U_out(n-1)

Mind: the OPAMPs need a power supply that is higher than the maximum expectable output voltge

Klaus

would you explain more about this circuit how it is connected any model you have?

 

[KlausST]

Hi,

did you do an internet search for "OPAMP ADDER circuit" or "OPAMP SUMMING circuit"?

Klaus

 

[Alok_Kumar]

Hi,

did you do an internet search for "OPAMP ADDER circuit" or "OPAMP SUMMING circuit"?

Klaus

but from adder ckt how we get from 100v to individual variable voltage for each cell channel

 

[BradtheRad]

I just want to create string of individual adjustable voltage with any one either isolated or non-isolated

More than once at Edaboard, schematics from certain simulator programs have appeared, which contain several of the identical component. In a corner is drawn a single master component. Maybe a voltage source. All the identical components take their parameters from the master component. Change the master component's parameters, and you change all the other matching component parameters.

I didn't notice the name of the simulator which does this. Anyway that could be one way to alter all 24 cells by altering only one. If you want some cells to have a different voltage, give them a different name and create a second master cell in the corner with that same name which governs the second group of cells.

To simulate a cell, I resort to drawing a voltage source equal to the discharged voltage, and a series capacitor several hundred Farads. In parallel with the cap I put a resistor which represents the cell's tendency to drop some voltage as it settles down to room temperature and/or self-discharge. To signify a cell's internal impedance, I put a resistor in series with the whole caboodle. Either I watch the simulation run for a few minutes until it creates false readings, or I devise circuitry which detects the range of the cell's voltage.

 

[KlausST]

but from adder ckt how we get from 100v to individual variable voltage for each cell channel

I´v already answered:

Mind: the OPAMPs need a power supply that is higher than the maximum expectable output voltge

And it should be obvious that you need an OPAMP capable of handling this high voltage. But that´s why there are selection tools to choose the right OPAMP.

So if you ask me for a different approach:
* you still need a power supply higher than the maximum expectable output.
* then place all your [pot with series resistor] in a string and supply them with a current source from the high rail.
But the output will be weak (high impedance) .. thus I wouldn´t be surprised if you need some kind of buffer. (Again: OPAMP?)
(We don´t know the load_current_range and expected accuracy ...)

***
It seems you somehow don´t like my answers. I´m a professinal electronics designer ... and my answers just show how I would solve the task.


Klaus

 

[D.A.(Tony)Stewart]

I´v already answered:

Mind: the OPAMPs need a power supply that is higher than the maximum expectable output voltge

And it should be obvious that you need an OPAMP capable of handling this high voltage. But that´s why there are selection tools to choose the right OPAMP.

So if you ask me for a different approach:
* you still need a power supply higher than the maximum expectable output.
* then place all your [pot with series resistor] in a string and supply them with a current source from the high rail.
But the output will be weak (high impedance) .. thus I wouldn´t be surprised if you need some kind of buffer. (Again: OPAMP?)
(We don´t know the load_current_range and expected accuracy ...)

***
It seems you somehow don´t like my answers. I´m a professinal electronics designer ... and my answers just show how I would solve the task.


Klaus

Klaus he wants to simulate a string of 24 Li Cells and test various scenarios of mismatched voltages to see if a BMS does its job whatever that is ( no specs) except he demands 3 digits accuracy but doesn't say how he is going to measure it. To properly do this needs variable ESR and variable V and he is willing to neglect variable C. He has no functional user interface specs, no testable design specs, nada. Yet he wants a cheap battery string simulator. How will he know if any of the cells deviates from the 3 digit accuracy at any time? How much power must be absorbed in this bidirectional cell synth, he won't say. He has no battery model and no gives no answers to my questions on specs. Commercial 24 channel battery string simulators are more than $3k and are 3U 19" racks or more and do a lot more than what he wants. Yet he is unable to create a functional design spec. and does not understand he is supplying the power with no specs that must be aborbed in this battery string simulator. Each cell is basically 10kF and 50 mOhms = RC but real batteries have 3 series R||C values and the total ESR must absorb the load. Imagine 100W to be dissipated per Amp at 100V CV for the total string. No specs. Your idea might go poof. (lol)

 

[KlausST]

Klaus he wants to simulate a string of 24 Li Cells and test various scenarios of mismatched voltages to see if a BMS does its job whatever that is

I understand post#8 differently:

I just want to simulate voltage for BMS and create UV and OV condition I don't want simulate charge, discharge and cell balancing
would you tell me how I create 24 channel for voltage sensing from programmable low power voltage source

Klaus

 

[D.A.(Tony)Stewart]

OK that is trivial.
Just cascade 24 LM317's with 24 pots and resistors to tune mix max range
Then use a flying capacitor 24 channel multiplexer to create a signal that monitors all channels on one serial analog signal.

Feed the input from the next higher output, drive the string with a current limiter LM317.