Simp BMS

[Kevin Sharpe]

On a Tesla system I would say up to 32 slaves of 6 cells each should work if the communication is robust enough.

The EVTV controller supports 62 Tesla Modules (here)

[tom91]

On a Tesla system I would say up to 32 slaves of 6 cells each should work if the communication is robust enough.

The EVTV controller supports 62 Tesla Modules (here)

In theory it does and so does the SimpBMS, but I am not going to promise it will do 62 modules as the data travel time will become quite high. No one has yet proven it can do this high of a module count.

[Kevin Sharpe]

No one has yet proven it can do this high of a module count.

Are you sure that EVTV have not tested that? It's very clear on their product page.

[tom91]

No one has yet proven it can do this high of a module count.

Are you sure that EVTV have not tested that? It's very clear on their product page.

It is a number based on the message format. 63 ids available one is assigned to broadcast to all.

[mdrobnak]

No need for a custom firmware, I can roll it into existing firmware. Have done so in the past with the Outlander Slaves.

On a Tesla system I would say up to 32 slaves of 6 cells each should work if the communication is robust enough.

Oh, nice. Yeah my thought was BMS + CAN Man-in-the-middle to change the ID + N number of slaves.. ($30 for a complete set on Ebay! - https://www.ebay.com/itm/Chevy-Volt-Sto ... 2817016231)

That's good news.

[JaniK]

Does this work with phase2 e-golf too? With the 35.8kWh pack?

[tom91]

Does this work with phase2 e-golf too? With the 35.8kWh pack?

Yes it can read temperatures and voltages but not command any balancing.

[JaniK]

Thanks for fast reply.
So there would be no balancing happening at all or it will just run automatic on slaves and be uncontrolled?

[tom91]

Thanks for fast reply.
So there would be no balancing happening at all or it will just run automatic on slaves and be uncontrolled?

No balancing.

[JaniK]

Ok cool. Its still great to have the voltages and temps.

[Boxster EV]

Does SIMPBMS provide any outputs upon charge completion (IE going from 'charge' state to 'ready')? I'm trying to figure our the best way to manage things like the lock pin on charge socket and cutting the PP signal to disconnect AC.

[arber333]

Does SIMPBMS provide any outputs upon charge completion (IE going from 'charge' state to 'ready')? I'm trying to figure our the best way to manage things like the lock pin on charge socket and cutting the PP signal to disconnect AC.

See here EVSE signaling https://en.wikipedia.org/wiki/SAE_J1772#Signaling

So from PP-PE you want to cut the signal and EVSE will think it is disconnected, but i dont know if it will unlock cable? That would be good to test first.
I have a button that pulls additional 1K3 resistor to CP - PE loop and EVSE starts the power. If i remove connection i still need to manually unlock cable from EVSE. I did actually put a switch lock on my plug and i use it if my cable gets stuck.
What i do is i cut complete CP-PE line after diode and EVSE drops the lock. But it reports this event as an error!

[tom91]

Does SIMPBMS provide any outputs upon charge completion (IE going from 'charge' state to 'ready')? I'm trying to figure our the best way to manage things like the lock pin on charge socket and cutting the PP signal to disconnect AC.

There is an output, charge enable that is on when it is in charge mode. In ready mode no outputs are on.

12V output so you can drive a relay to do all sorts of things.

You can get so fancy to use some relay logic to do, car locked OR charging, lock actuator is locked.

Car unlocked AND not charging lock actuator unlocked.

[Boxster EV]

Does SIMPBMS provide any outputs upon charge completion (IE going from 'charge' state to 'ready')? I'm trying to figure our the best way to manage things like the lock pin on charge socket and cutting the PP signal to disconnect AC.

There is an output, charge enable that is on when it is in charge mode. In ready mode no outputs are on.

12V output so you can drive a relay to do all sorts of things.

You can get so fancy to use some relay logic to do, car locked OR charging, lock actuator is locked.

Car unlocked AND not charging lock actuator unlocked.

Yep, I'm thinking analogue and planning to introduce a few relays to manage what I want to switch on or off.

However, I don't quite understand how an output is achieved... Pin 6 of connector J5 is 'charge enable', fed from 12v supply to tell simpbms to go into a charge state. (I achieve this via the 240vac relay coil on my setup). My understanding is that the BMS moves the state from charge to ready once the pack is charged (IE pin 6 will still have 12 fed to it).

Therefore how can an output be achieved?

I'm probably being thick here and need a light bulb moment.

[tom91]

Yep, I'm thinking analogue and planning to introduce a few relays to manage what I want to switch on or off.

However, I don't quite understand how an output is achieved... Pin 6 of connector J5 is 'charge enable', fed from 12v supply to tell simpbms to go into a charge state. (I achieve this via the 240vac relay coil on my setup). My understanding is that the BMS moves the state from charge to ready once the pack is charged (IE pin 6 will still have 12 fed to it).

Therefore how can an output be achieved?

I'm probably being thick here and need a light bulb moment.

J5 is input

J4 has the outputs on it, these are matched to BMS states. Pin 13 is the 12V charge enable, so if the BMS is in the charge state this output is on.

https://github.com/tomdebree/SimpBMS/bl ... 0V0.21.pdf

[Boxster EV]

Yep, I'm thinking analogue and planning to introduce a few relays to manage what I want to switch on or off.

However, I don't quite understand how an output is achieved... Pin 6 of connector J5 is 'charge enable', fed from 12v supply to tell simpbms to go into a charge state. (I achieve this via the 240vac relay coil on my setup). My understanding is that the BMS moves the state from charge to ready once the pack is charged (IE pin 6 will still have 12 fed to it).

Therefore how can an output be achieved?

I'm probably being thick here and need a light bulb moment.

J5 is input

J4 has the outputs on it, these are matched to BMS states. Pin 13 is the 12V charge enable, so if the BMS is in the charge state this output is on.

https://github.com/tomdebree/SimpBMS/bl ... 0V0.21.pdf

And there's the light bulb moment.

[rwaudio]

Is it possible to use the Volt BMS slaves with the LG Chem / Chrysler Pacifica batteries? They are 16 series cells, none in parallel. That would certainly cut total acquisition costs on those if so...

You can wire any slave to any battery, just as long as the chemistry and cell count matches.

always some exceptions on this rule, you can wire some slaves with less cells if you tie the unused ones to the highest cell (tested on Outlander and Tesla Slaves.

Hey tom91,
I'm playing with a bunch of Tesla version one model S BMS boards (I use to break down modules to get raw 18650's) I see you mention being able to use the boards at lower than full cell count. I'm going to be using 10x 10S LG Chevy modules and would like to use two boards per module with cell 6 on both not connected. I've tried tying 5&6 together etc however the board always registers a fault. Do you know if there's any way to retain the hardware fault line with 5 cells per module? I'm just doing this on a Due for proof of concept.

[tom91]

Is it possible to use the Volt BMS slaves with the LG Chem / Chrysler Pacifica batteries? They are 16 series cells, none in parallel. That would certainly cut total acquisition costs on those if so...

You can wire any slave to any battery, just as long as the chemistry and cell count matches.

always some exceptions on this rule, you can wire some slaves with less cells if you tie the unused ones to the highest cell (tested on Outlander and Tesla Slaves.

Hey tom91,
I'm playing with a bunch of Tesla version one model S BMS boards (I use to break down modules to get raw 18650's) I see you mention being able to use the boards at lower than full cell count. I'm going to be using 10x 10S LG Chevy modules and would like to use two boards per module with cell 6 on both not connected. I've tried tying 5&6 together etc however the board always registers a fault. Do you know if there's any way to retain the hardware fault line with 5 cells per module? I'm just doing this on a Due for proof of concept.

No hardware fault will happen thats in the firmware fo the slave, however data will still flow correctly, you need to tie 5 and 6 together to get correct data.

I believe you can even tie 4 5 6 togethor.

[rwaudio]

You can wire any slave to any battery, just as long as the chemistry and cell count matches.

always some exceptions on this rule, you can wire some slaves with less cells if you tie the unused ones to the highest cell (tested on Outlander and Tesla Slaves.

Hey tom91,
I'm playing with a bunch of Tesla version one model S BMS boards (I use to break down modules to get raw 18650's) I see you mention being able to use the boards at lower than full cell count. I'm going to be using 10x 10S LG Chevy modules and would like to use two boards per module with cell 6 on both not connected. I've tried tying 5&6 together etc however the board always registers a fault. Do you know if there's any way to retain the hardware fault line with 5 cells per module? I'm just doing this on a Due for proof of concept.

No hardware fault will happen thats in the firmware fo the slave, however data will still flow correctly, you need to tie 5 and 6 together to get correct data.

I believe you can even tie 4 5 6 togethor.

Thanks! Since the fault line doesn't seem to activate on temperature anyway I guess I'll ignore it.

[arber333]

Hi TomDB

I got SimpBMS this weekend and i went on and took apart original Volt BMS wiring, Up till now i read manual and made diagrams on how i would connect it in my Pug. I had had no problem with understanding of the manual. But i am left with a single wire which i think is a Fault wire.
I have made a number of tests on Volt BECM which led me to belive this wire sends quick +5V transient in case if i unplug one connector or one wire from BMS. This may be a latch release for BECM report during charging. But i never saw any difference in BECM output, Did you consider to use this wire or signal in your BMS?

I am talking about the blue wire in position pin 13 of BECM connector. Could we use it somehow? Or do i discard it?

TNX

[tom91]

Arber, I do not use any fault wires Volt or Tesla. The way things get done is verifying the transmitted data and if the data is transmitted and if there are any issues not allow anything to happen. I do not like assuming things are fine when data becomes spotty or drops out.

Thank you for the feedback on the manual.

[arber333]

Arber, I do not use any fault wires Volt or Tesla. The way things get done is verifying the transmitted data and if the data is transmitted and if there are any issues not allow anything to happen. I do not like assuming things are fine when data becomes spotty or drops out.

Thank you for the feedback on the manual.

OK so i will just leave the wire inside the loom for future use maybe...

[joromy]

I want to have the kWh sent out on the CAN bus.

Have put in "kWh = amp times volt" and added it to CAN send.
And are receiving i correctly on my CAN display.

But I don't understand the theory behind the calculation, what is 0.27777777777778.
It may be wrong way to calculate remaining kWh of the battery anyway?

Code: Select all

    SERIALCONSOLE.print("% SOC ");
    SERIALCONSOLE.print(ampsecond * 0.27777777777778, 2);
    SERIALCONSOLE.println ("mAh  ");
    SERIALCONSOLE.print(kWh);
    SERIALCONSOLE.println ("kWh");
  }
  kWh = (ampsecond * 0.00027777777778 * bms.getPackVoltage()/ 1000);

[tom91]

I want to have the kWh sent out on the CAN bus.

Have put in "kWh = amp times volt" and added it to CAN send.
And are receiving i correctly on my CAN display.

But I don't understand the theory behind the calculation, what is 0.27777777777778.
It may be wrong way to calculate remaining kWh of the battery anyway?

Code: Select all

    SERIALCONSOLE.print("% SOC ");
    SERIALCONSOLE.print(ampsecond * 0.27777777777778, 2);
    SERIALCONSOLE.println ("mAh  ");
    SERIALCONSOLE.print(kWh);
    SERIALCONSOLE.println ("kWh");
  }
  kWh = (ampsecond * 0.00027777777778 * bms.getPackVoltage()/ 1000);

AmpSecond to Amp Hour

Biggest issue is that Ampere Capacity (Ah or AS) is not linked to voltage, voltage varies over SOC so calculating kWh is not as easy as you think.

[Bryson]

Maybe the better way is SOC * Ah setting * #cells in series * voltage at 50% SOC (interpolated from whatever you chose your SOC settings to be)