Old Email Conversations on Arduino board development

189 views
Skip to first unread message

RickyG

unread,
Feb 23, 2012, 3:13:56 AM2/23/12
to openBMS
Hi Henry,

I agree with your 12V setup proposal. I misunderstood your last email.

Also, I think it'll be a good idea to have a "self shut off everything
protection".
This is an article on Tesla owners bricking their $40,000 battery
because the left the car unplugged for a few weeks.
http://theunderstatement.com/post/18030062041/its-a-brick-tesla-motors-devastating-design

Ideally the BMS should be smart enough to know that when it's not
plugged in and the battery is almost depleted, it'll flick some sort
of master switch to turn off everything and minimizing the parasitic
drain. What do you think of this idea?

I haven't drawn up anything as I've been really busy. I know there
should be some development in the car club at UBC. Neil might know
more about it.

Also, the google group is setup.

Ricky
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
On 2012-02-22, at 11:28 PM, Henry Palonen wrote:

Hi,

Yes, it could be a good thing to add SSR's for enabling different
setups. (In my setup I can use also that CAN-PLC to drive anything
necessary, it already has an "relay-box" that has places for
automotive relays and automotive fuses.)

But anyway, I have many CLARE CPC1706Y SSR relays available so I could
draw a circuit with them so that BMS-shield becomes more universal.
They are TTL/CMOS compatible and work up to 4A/60V each, requiring
only about 5mA to drive them.

But for 12V input - what I ment that if car's circuit is wired so that
DC/DC doesn't come on when charging (as is in many cases) then we need
2 separate 12V sources going to the shield or Arduino 12V input. One
during charge and one during driving. Of course we could always power
up vehicle DC/DC while charging but I don't think it is a very good
idea because that also powers up everything else in the vehicle and I
don't want to charge while vehicle is in driving condition. Elithion
uses 2 separate 12V inputs and can tell if vehicle is connected to
wall by reading the status of these 2 separate inputs, you can see
connection diagram from [1]. That is the same way I would prefer to
input 12V to BMS-shield so that we would always know if the vehicle is
plugged in or not. It also allows quite easy separation of powering up
devices that are needed for driving from those devices that are needed
only while charging. In my motorcycle, this has allowed very easy
separation, lights and other motorcycle +12V items come on only after
it's powered on from key (eg. 300W DC/DC is powered on) and only BMS
+charger+PLC+EvUI-display are powered on during charge. So I can
charge without leaving key to the lock.

Later, we could of course have one single master-board that is
physically similar to the daughter board (so that same case can be
used) but has all of these integrated to the same shield (Arduino +
BMS-shield + CAN-shield). But for now, for me at least, it could be
quicker to move forward and begin testing and prototyping with simple
Arduino shields. Or have you guys already drawn up an some master
board having all these items integrated ?

And yes, a Google Group would be very fine. Will you guys setup up the
group or will I ?

[1] http://lithiumate.elithion.com/php/install_controller.php#Power

With very best regards,

Henkka
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Ricky Gu kirjoitti 22.2.2012 kello 18.50:

Hey Henry,

The diagram looks good. One thing I think it'll be good to add is
three 12v relay driver output for generic switching and compatibility
with different cars.


One relay can be used for cutting off charger that does not have
canbus.
One relay should be in series with the main contractor coil to open
the main pack when a cell is getting really low.

One relay used for overriding the brake Light during regeneration
braking.

As for the 12v. Arduino has its own dc dc built in you can take a look
to see if it'll supply enough current.


What do you say we start a Google Group to make our conversation like
this public? It'll be good for others who wants to join in or build
one them self.

Best regards


Ricky Gu

Sent from my BlackBerry® PlayBook™
www.blackberry.com



----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
From: "Henry Palonen" <henry....@gmail.com>

To: "Neil Roberts" <neil.r...@ubcecc.com>
CC: "Ricky Gu" <m...@rickygu.com>

Sent: February 22, 2012 3:52 AM
Subject: OpenBMS CAN-connection proposal

Hello once again,

I forked Ricky's Github and started to draw an simple BMS-shield for
Arduino. I also drawed a general proposal for connection diagram for
CAN-equipped basic vehicle. You can find diagram from [1]. Do you have
any suggestions based on that diagram ? I think that perhaps most
important issues are


a) controlling charger for not to overcharge (this can be done with
CAN-bus) and
b) controlling motor-controller for not to discharge too much (this
can be done with some resistance in parallel with motor controller
potentiometer input).


In addition, current sensing should be done during driving. During
charge the charger sends current with CAN-messages so no need to
measure that (at least in my setup). So I'm planning of doing just an
simple Arduino BMS master-shield with a connector for LEM HFTS current
sensor and some kind of "resistor-output" for motor controller. Also
input voltage for the Arduino should come from 2 places: the wall
adapter (during charge) and from the DC/DC controller (during drive).
I have drawn simple preliminary circuit for obtaining 12V from either
DC/DC or from cheap 12V wall-adapter through diodes. And perhaps those
should also be routed to Arduino I/O for detecting if we are powered
from wall-adapter or not so we can detect that we are not powered on /
driving while plugged in.


Slowly progressing...

Best regards,

Henkka

[1] https://raw.github.com/randomev/openBMS/master/Docs/System-general-connections.pdf



Henry Palonen

unread,
Feb 23, 2012, 4:31:32 AM2/23/12
to ope...@googlegroups.com

RickyG kirjoitti 23.2.2012 kello 10.13:

> Hi Henry,


>
> Also, I think it'll be a good idea to have a "self shut off everything
> protection".
> This is an article on Tesla owners bricking their $40,000 battery
> because the left the car unplugged for a few weeks.
> http://theunderstatement.com/post/18030062041/its-a-brick-tesla-motors-devastating-design

Ouch! That must hurt when owner discovers that his EV sports-car turned to "brick" :(

>
> Ideally the BMS should be smart enough to know that when it's not
> plugged in and the battery is almost depleted, it'll flick some sort
> of master switch to turn off everything and minimizing the parasitic
> drain. What do you think of this idea?

Yes, that is a very good idea to prevent that sort of thing from happening! New LTC6803-2's would take even less power from the cells than old LTC6802-2's. I don't yet know which IC's Neil & your local PCB-supplier choosed for the prototype boards ?

Even after all power removed, there will be a small amount of self-discharge in the cells. If discharged to 0V, successful recovery depends on cell-type and charge current also. I have had A123 cylindrical cells recovered to full power by slow initial charge rate even when they were very slowly discharged way below 1V.

To get all parasitic power removed, LTC-chips should also put to sleep-mode or similar. I don't know yet if they have such an mode and how do they recover from it if there is such an mode. Need to read more those datasheets. And EV electrical circuit should be designed with this "systemwide turn off" in mind so that there would be one point that can disconnect everything that eats battery power. After such an event, we could perhaps get things going again since we will have separate +12V input to system when eventually again charging. But after such an event, charge current should perhaps be kept low initially so that cells recover (if recovery is possible at all).

Best regards from Finland,

Henkka

Reply all
Reply to author
Forward
0 new messages