[WolfTronix]

This post might be useful for those looking to upgrade lead acid based 144V systems...

Here is the configuration I came up with, while reusing the Nissan Leaf interconnects, safety covers, and BMS wiring.

This is one of four 150V 65Ah modules for my Solectria E10.

Total battery weight will be 656 pounds for a 150V 260Ah battery pack.
Under 700 pounds with all the module carrier plates, interconnects and BMS wiring.

I expect to get around 120 mile range when done (80-20 charge/discharge).

All the pics are here:
http://www.wolftronix.com/E10_LithiumUpgrade/index.htm

Video playlist is here:
https://www.youtube.com/playlist?list=PLQdu_G7xyFIQ-hNQHEqc3in4qWGy6dpfp

Enjoy,
Wolf

 

[Weisheimer]

That is one impressive feat there Wolf...

And you have temperature monitoring and battery monitoring by your upgrade of the original battery monitor system.

Nice work!

 

[WolfTronix]

Yup, I replaced the Nissan's microcontroller with my own:

Now I have full control over the BMS hardware:
Cell Voltages.
Cell Balancers.
Current Sense.
Total stack voltage.
12V Auxiliary battery voltage.
I/O for the heaters.
CAN bus.
Temp Sensors.
Isolation monitor.
High Voltage Interlock.
EEPROM.
Miscellaneous I/O (Ignition switch, Watch dog timer, etc...).

Trying to have the Truck back on the road before Spring.

 

[CKidder]

Building an adapter board and putting your own CPU on there as basically the only hardware modification is pretty neat. That's slick stuff. Congrats!

 

[Yabert]

Yup, I replaced the Nissan's microcontroller with my own

Wooo! Nice!
Good job.

 

[electro wrks]

Great work! Would it be possible to do something like this with Tesla BMS hardware? I haven't checked in a while. Maybe someone else has already done it.

 

[WolfTronix]

Great work! Would it be possible to do something like this with Tesla BMS hardware? I haven't checked in a while. Maybe someone else has already done it.

I would assume so...

This is typically what I do when I can't extract the firmware and decompile into something I am familiar with to make the changes I need...

In this case, I reverse engineered the hardware of the LEAF BMS, and once I new what pins on the micro controlled what.

I was able to make my own interposer board that routs the digital and analog pins over to the digital and analog pins on my micro.

The only reason I have to make changes, is because I am not using all 48 modules in the Leaf battery pack, (I am only using 40 of them).

This makes the microcontroller unhappy, and spends most of its time throwing fault codes and repeatably looking for the missing monitor/balance chips. And it disables balancing of the cells, since cells are literally missing.

Also, any unpowered monitor/balance chips breaks the serial LIN communications bus, and you get no cell voltages.

Anyway, yes, I would assume someone could do exactly the same thing with the Tesla BMS.

 

[WolfTronix]

Congratulations Wolf.
Any plan to sell your micro controller to the DIY community? How much $ would it be?

Sure if others want it...
How much do you thing others will pay for it?
I have about $6 in parts for the microcontroller and interposer board.
But lots of time and labor reverse engineering the hardware and software protocols.

Please post its specs and future plans for it.

Info is on the website for it, but is a bit out of date:
http://www.wolftronix.com/E10_LithiumUpgrade/index.html

The specs will be identical to the Nissan Leaf BMS, with the following exceptions:

You can have any number of even modules you want, for example:
48, 46, 44, 42, 40 ... 10, 8, 6, 4, 2 series modules.
(each monitor/balancer chip needs a 4 cells to power it, and still support a single cell shorted... when 2 cells short out the monitor/balancer chip stops talking, breaking all serial communications).

I will support the following balance modes:

Bottom balance (You will have to discharge the pack externally, remove the load, then let the balance shunts take each cell down to the same target voltage).

Top balance (charge to the first cell reaches target voltage, then let the balance shunts take each cell down to the lowest cell). This would happen at the end of every charge cycle.

Continuous balance (let the balance shunts take each cell down to the lowest cell voltage, all the time). This appears to be what Nissan does.

Support for more that one BMS unit on the same CAN bus:
By shifting the CAN address of each BMS unit, or using a different CAN protocol.

The thing I don't know how to recreate is the magic algorithm that Nissan uses to calculate battery degradation and state of charge... They would take in account the battery temperature, mileage, charge cycles, depth of discharge, voltage, current, etc...

I will probably just use amp-hours in and out.
Then set the battery pack capacity when the lowest cell reaches the discharge voltage knee, (since I would know how many amp-hours it took to get there). Kind of like when you do a battery gauge calibration on your cell phone. ;P

It would be really nice to be able to use as much Leaf Modules as the end-user wants. For instance in a NEV I would make 2 cars out of one Nissan Leaf battery pack.

Yup, see above... but you would need to acquire another Nissan Leaf BMS unit for the second vehicle.

How do you actually monitor (see) voltages and temperatures in real time? (I have not seen your YT videos), I assume you have a screen, "syslog" files for historic review?

How ever you want.
It will spit out CAN packets, you can use Leaf Spy, my Pack Sniffer, a laptop, a tablet, etc...

For my truck I will have a dedicated multi-function display for real time battery cell voltage monitoring.

I believe its time for you to test it in your EV. Have you confirmed everything works? (Balancing, temp sensors, etc)

It is in the works...
My goal is to have the truck up and running in the Spring...
But it might take longer to implement all the features I want.

A friend of mine in Florida used leaf cells for almost 3 years in a conversion, without BMS, just bottom-balancing them without issues, he was very careful on SOC, he never charged the cells to their full capacity. Your approach is the best.

Hector

Nissan has a BMS unit for a reason.

Thanks,
Wolf

 

[Karter2]

It would be really good if you could resize your photos such that they do not distort the page format to micro print when this thread is opened.
Thanks in advance.

 

[MathisLaurant]

Just out of curiosity. This battery has 3 connections. I know one is the positive terminal and the other is negative. But what is the center? Is that the BMS terminal?

 

[WolfTronix]

Just out of curiosity. This battery has 3 connections. I know one is the positive terminal and the other is negative. But what is the center? Is that the BMS terminal?

The Nissan Leaf modules actually have 4 cells in them...
Two are in parallel, and those two groups are in series, (2s2p)

The center post is connected to both cell groups, (center tap), this lets the BMS measure and balance every two cell group in the module.

 

[MathisLaurant]

The Nissan Leaf modules actually have 4 cells in them...
Two are in parallel, and those two groups are in series, (2s2p)

The center post is connected to both cell groups, (center tap), this lets the BMS measure and balance every two cell group in the module.

Ok, I've been looking online for a Data sheet on these Cells or a diagram how Nissan wires the BMS to for these cells but cant find anything. I'll keep looking thanks

 

[WolfTronix]

From the Service Manual

 

[WolfTronix]

From the Service Manual

 

[WolfTronix]

Leaf module battery specs:
http://www.electricvehiclewiki.com/Battery_specs

 

[MathisLaurant]

Leaf module battery specs:
http://www.electricvehiclewiki.com/Battery_specs

is there a Charge and discharge C rating for these Cells?
I contacted the supplier of the cell at http://www.eco-aesc-lb.com/en/product/liion_ev/ for that information, but they told me that is proprietary information haha. but maybe there is a way to calculated based on the charging time for lvl 1, 2, and 3 chargers?

 

[WolfTronix]

is there a Charge and discharge C rating for these Cells?

The pack is fused at 225A... so that is probably the max continuous discharge current.

According to Wikipedia:
https://en.wikipedia.org/wiki/Nissan_Leaf

The motor is 80kW... so 80,000 / 360V = 222.2A
And 44kW charge on CHAdeMO = 44000 / 360V = 122.2A

 

[D a n n y^]

Sure if others want it...

How much do you thing others will pay for it?

I have about $6 in parts for the microcontroller and interposer board.

But lots of time and labor reverse engineering the hardware and software protocols.







Info is on the website for it, but is a bit out of date:

http://www.wolftronix.com/E10_LithiumUpgrade/index.html



The specs will be identical to the Nissan Leaf BMS, with the following exceptions:



You can have any number of even modules you want, for example:

48, 46, 44, 42, 40 ... 10, 8, 6, 4, 2 series modules.

(each monitor/balancer chip needs a 4 cells to power it, and still support a single cell shorted... when 2 cells short out the monitor/balancer chip stops talking, breaking all serial communications).



I will support the following balance modes:



Bottom balance (You will have to discharge the pack externally, remove the load, then let the balance shunts take each cell down to the same target voltage).



Top balance (charge to the first cell reaches target voltage, then let the balance shunts take each cell down to the lowest cell). This would happen at the end of every charge cycle.



Continuous balance (let the balance shunts take each cell down to the lowest cell voltage, all the time). This appears to be what Nissan does.



Support for more that one BMS unit on the same CAN bus:

By shifting the CAN address of each BMS unit, or using a different CAN protocol.



The thing I don't know how to recreate is the magic algorithm that Nissan uses to calculate battery degradation and state of charge... They would take in account the battery temperature, mileage, charge cycles, depth of discharge, voltage, current, etc...



I will probably just use amp-hours in and out.

Then set the battery pack capacity when the lowest cell reaches the discharge voltage knee, (since I would know how many amp-hours it took to get there). Kind of like when you do a battery gauge calibration on your cell phone. ;P







Yup, see above... but you would need to acquire another Nissan Leaf BMS unit for the second vehicle.







How ever you want.

It will spit out CAN packets, you can use Leaf Spy, my Pack Sniffer, a laptop, a tablet, etc...



For my truck I will have a dedicated multi-function display for real time battery cell voltage monitoring.







It is in the works...

My goal is to have the truck up and running in the Spring...

But it might take longer to implement all the features I want.







Nissan has a BMS unit for a reason.



Thanks,

Wolf

I'll be watching for the "add to cart" button for this set up on your website! Make sure and post some videos when you get the truck going. You've got my attention.


Sent from my iPad using Tapatalk

 

[WolfTronix]

I'll be watching for the "add to cart" button for this set up on your website! Make sure and post some videos when you get the truck going. You've got my attention.


Sent from my iPad using Tapatalk

I have my microcontroller running the BMS hardware now.

Here is the latest video:
https://youtu.be/YilgYwfq_fI

I need to open up my other Leaf battery pack and swap in this BMS to finish the code to read out the cell voltages and command the balance shunts... but I have been sick... hopefully I can get to it this weekend.

Thanks,
Wolf

 

[Caps18]

The pack is fused at 225A... so that is probably the max continuous discharge current.

According to Wikipedia:
https://en.wikipedia.org/wiki/Nissan_Leaf

The motor is 80kW... so 80,000 / 360V = 222.2A
And 44kW charge on CHAdeMO = 44000 / 360V = 122.2A

That is what I had thought too, but it appears that there are time-current issues when dealing with fuses. I think I figured out that I could get 650A for 60 seconds (which would be 100% on my motor controller)