(data here
http://www.teslamotors.com/blog/most-peculiar-test-drive 

and original review here
http://www.nytimes.com/2013/02/10/automobiles/stalled-on-the-ev-highway.html?_r=1&pagewanted=all&;
 



So I'm not saying the NY times guy wasn't an class A-one jerk.  It's pretty 
obvious he wasn't being careful
and actually seemed to want to wreck the car.   (I had roomates (actually Intel 
engineers) do similar 
things to an old restored vehicle of mine.  Some people (even educated) turn 
into infantile sociopaths when you let them
operate some piece of machinery that they aren't held accountable for, And  
unfortunately you usually learn
it too late after they've done considerable damage.

However, I'm not willing to let Tesla off the hook, just because they had a 
jerk reviewer.  They've also
got a problem that they should address.

If you look at Musk's data, you see a peculiar vertical line at 400 miles in 
showing the range decreased
from around 80 miles remaining to around 20 miles remaining overnight.. So it 
lost 60 miles of estimated
range.   (roughly 25%) but if you look at the S.O.C. it only lost about 7%.   
Furthermore if you look at
the last charge at the members-only coffee shop, you'll see that he charged 
about 5% of his charge back
(roughly).   Furthermore the car went  around 80 miles.  Virtually every line 
in the S.O.C .vs. miles shows
very nearly the same slope.   So I'm not buying this is a cold weather thing as 
far as driving is concerned.

Point being that the car lost about 7% of it's S.O.C. overnight in roughly 12 
hours.   That's a drain of
7%*85kwh = 6kwh/12 hours = 500 watts.    That's pretty big.  

Does the Tesla REALLY use that when parked?  That just seems ridiculous.   I 
could see several possibilities
(a)  It does use this much power (uggh) ...(perhaps the logging equipment 
stayed on?...DC-DC converter inefficient)
(b)  It has a very bad problem with self discharge
(c)  The amp shunts sucks and they can't measure low currents with accuracy.   
Kind of a problem in any electric car.  You
have to be able to measure nearly 1000 amps (at peak power) but then measure 
0.01 amps over long periods of time to measure
the S.O.C.   
(d)   BMS guessing game?.
(e)  all of the above?

Suppose it was (d)   One problem with most BMS is they use the power directly 
from the batteries.   That means the power they use
doesn't go through the main shunt and therefore it's a current draw that might 
not get measured.   That means you have to estimate it
and that means you could be really far off.    If the BMS is in full shunt 
mode, it might actually use 500 watts...so maybe they
just guessed conservatively?... 

Now, if you look at the Estimated range .vs. SOC, you see they drastically 
underestimated the range after sitting overnight?  Why would they
do this unless they actually didn't have a good handle on what the actually had 
in the batteries.

Now, my lifepo4 batteries are pretty damn amazing.  I had a bunch in my Garage 
sitting about 6 months and 
I don't think they lost 10% of their charge OVER 6 MONTHS.   But a Telsa loses 
nearly that overnight?   

Furthermore if you look at his charts showing the S.O.C. (state of charge) they 
don't seem to jive with the
estimated range very well.   For example it shows the SOC at the overnight stop 
changing from about 37% to 30%. but
the estimated range went from 80 miles to 20 miles  (30% to  7.5% with a 265 
mile range car).   And again
as you can see, virtually every line is the same slope.   So why make the 
estimated range so conservative?

Now why would they do that, unless there is some funny business going on? 

I personally think the whole "estimated range" is a recipe for disaster.  
Trying to make it idiot-proof isn't going to work.   There are too many 
variables that can affect range (headwinds, traffic, speed, hills, 
temperature).   Idiots don't know that an "estimate" is an "estimate".   When 
you want to make something idiot proof, you need to think about what George 
Carlin said:  


      "Think of how stupid the average person is and then realize 50% of them 
are even stupider than that."

So here's a possible answer.  Set your range estimation software to the 
absolute worst case scenario reasonably possible.   So instead of  Estimated 
range, you say "estimated range" and
"guaranteed range".  Now, I personally think that might not be the worst thing 
in the world.    You exceed your guaranteed range and you are responsible.  
Meanwhile
you can use an estimated range for when you want to push the limits, knowing 
that you may very well have drive very very conservatively and you are on your 
own.

Now this wouldn't have fixed a sociopath like Broder.  He started on a 61 mile 
trip with 32 miles of range and then wrote a bad review.  No fixing that.

But IMHO there does seem something very wrong about the range estimation 
software on the Tesla and Tesla's own data seems to confirm it.

http://www.teslamotors.com/blog/most-peculiar-test-drive 



Anyway all of this is just my random conjecture and I really know nothing, but 
I thought I'd throw in my 2 cents.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
<http://lists.evdl.org/private.cgi/ev-evdl.org/attachments/20130218/f115b067/attachment.htm>
_______________________________________________
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
http://lists.evdl.org/listinfo.cgi/ev-evdl.org
For EV drag racing discussion, please use NEDRA 
(http://groups.yahoo.com/group/NEDRA)

Reply via email to