[sokoloff]

We have found some experimental evidence that may be related to the
Smart car battery "bricking" issue.

Bill, a forum member, reached out reporting a phantom power drain in
his 2015 451ED. It was sometimes losing 10% charge overnight and he
was trying to chase it down. Both his high voltage and low voltage
(12V) batteries were original to the car.

After some investigation, it appears that the car was frequently
charging the LV battery from the HV battery and this was likely the
source of his phantom power loss. If left to continue unchecked, this
is also a possible source of the smart's "bricking" problem.

The generic Group 47 battery ($105 at Costco) is about an inch longer
than the old original battery, but it slid into place just fine. The
only problem is that the original hold down clamp would not go into
position. A little hacksaw work on it allowed it to do its job again.
There is a sticky near the top of the forum giving installation instructions.

More details in the next post about the data collected along the way,
which you can skip if you're not interested, but the takeaways for us
are:

If you notice an unexplained power drain on your HV battery, you need
to find out why.

If your battery is five years old or older, you should keep an eye on
it.

If you have a multi-meter or a battery test unit, you should check
the 12v battery every once in a while, more often if either of the
first two issues are yours.

Good luck. Hopefully this experience will help a few owners from owning
a "bricked" Smart car.

Len, Bill, and Jim

 

[sokoloff]

More details on the data collected are in this post. Those not interested can
skip to the next post.

After taking some measurements on his own, he reached out for help in
interpreting what he was seeing. As best we could, we've
reconstructed (and shortened) the findings below.

Bill originally tested the LV battery with the OBD2 test unit at
10.9V, which is very low. (Post 1). He then plugged the car in,
charged the HV battery, and tested the LV battery right after that and
got 14.02V on his multi-meter.

He let the car sit for 24 hours (no charging cable attached and no
driving) and re-checked with the multi-meter. It read 14.3V which is
typical of a 12V battery being charged and impossible for a 12V
battery to be resting at that voltage that long. Something didn't make
sense, but it seems clear the LV battery was being charged by the car
while it was resting. Additionally, Bill noticed a slight humming
sound when he was getting ready to remove the old battery.

This was from the DC-to-DC convertor that the
car uses to power the 12V systems and charge the LV battery.

Was the original phantom power drain caused by a weak 12v battery
causing the HV battery to charge it often, thus reducing the charge of
the HV battery by 10% or so over night? A good 12V battery shouldn't
need a supplementary charge often and a typical owner would see no
significant power drain of the HV battery.

Bill has one of our battery test units and noticed that even when he
wanted to test the 12v battery with it, it was reading over 14V.
Reading more of the report data, for all of these tests, the HV
contactor indicated that it was "ON," even though it should have been
"OFF" to give an accurate reading of the 12v battery at rest. This
contactor being "ON" was another clue that might explain the overnight
phantom power loss and the combination of a weak 12v battery and
phantom power loss suggests this might be the car trying to charge a
weak 12V battery.

Bill decided to get a new 12V battery. He bought one (Group 47) from
Costco for $105 and installed it. See photo below. The old battery
read 13.45V on the bench immediately after it was removed, again
indicating that the HV battery had been doing some charging. After
resting for 24 hours out of the car, the old battery settled on 12.49V
as expected.

The new battery read 12.5V as it was going in.

Subsequent tests of the LV battery with the car shut down gave
readings in the expected 12.4-ish range, there have been no further
phantom power losses, and the HV contactor with the car at rest now
reports that it's "OFF", all of which combine to suggest that Bill's
phantom power loss problem was likely related to the LV battery being
charged.

Len, Bill and Jim

 

[sokoloff]

How can an end user test their low-voltage battery?

There are a number of ways, all of which are equally fine, but the
important thing is to test the voltage with the car off, not plugged
in, and without the key inserted. Press the lock/unlock button to power up.
There is a sticky near the top of the forum giving more details about this test.
You're trying to test the 12V (LV)
battery while it's resting and in steady-state, not while it's being
charged or shortly after that.

Any reading over 13V should convince you that the car is charging the
battery (or was charging it moments ago and the surface charge hasn't
bled off yet).

Using a multi-meter, set to Volts, connected across any powered 12V
source and ground is the old-school method. I looked and couldn't find any convenient place to check the battery
voltage. Damn near everything is plastic. The fuse panel is not all
that convenient to open and use. Some guys have run a wires from the
battery to the "hood" panel, but opening and using that is no easier
than going in through the floor carpeting to the actual battery.

Many OBD2 diagnostic devices ("code readers") can also report the
voltage of the car battery.

Lastly, the smart-specific test unit can also report it as well as
reporting a lot of details about the state of your high-voltage
battery. We published a new version of the code, v1.0.8b, that provides
plain-text messages about the voltages seen. Note that older versions
of the tool are just as capable to do this diagnosis (Bill used
v1.0.7); v1.0.8b just makes it a little more convenient.

If you have one of our test units, here is a link to update it to the latest version:

Better Battery Status Messages in v1.0.8b

Interpreting the reading. If the reading is under 12.4 Volts, consider
the battery as slightly discharged. That might be because the car has
sat for a long-time un-driven, but if it's a few years old, it's more
likely because the battery is somewhat weak. If the reading is 13.0 V
or higher, the battery is being charged [or was very recently being
charged, or your meter is faulty]. This is the concerning situation
and the one you need to look at more closely.

Len, Bill and Jim

 

[Vincent Rapide]

I plan to disconnect the negative post on the 12v battery when I'm away on trips having concluded that this will save the HV battery from any inadvertent discharge.
Ideally leave the HV at about 50 to 75%. When re-connecting do so at exactly 12 noon. 8^)

VR. PNW USA
--

 

[Yinzer]

I plan to disconnect the negative post on the 12v battery when I'm away on trips having concluded that this will save the HV battery from any inadvertent discharge.
Ideally leave the HV at about 50 to 75%. When re-connecting do so at exactly 12 noon. 8^)

VR. PNW USA
--

No! don't so that!

There is a case in this forum of a Canadian owner in this forum removing his 12V battery and taking it indoors for the winter. When spring came, he found that the HV battery had somehow discharged itself completely, near-zero volts flat (he actually removed the battery, opened it up and measured the cell voltages). Very quickly, this led to cases of others who saw their HV batteries self-destroy when the 12V battery went flat. This led to a worknig hypotheses here that the 451 ED has some kind of bug that caused the BMS to self-destroy the HV battery if the 12 volt syatem is de-energized.

The perferred thing to do is to connect a maintenance trickle charger to the 12V battery if the car will be parked for an extended period of time.

 

[Vincent Rapide]

Thanks for that! But he should have been checking the batt every week or so as I will do. But this calls for finding the HV disconnect switch or installing one. This switch could disconnect the main feed to the BMS as well so there will be no chance of any discharge. Li-Ion cells disconnected take like forever to discharge.

VR
--