Not really. I was just trying to clarify that the statement “range test is how efficiency is defined” (which is what you said!)...is not the case, at all. In fact,
the charging phase is not needed to determine the range. In this picture, you can derive the (unweighted) UDDS range according to the formula UDDSu = (SUM A)/(SUM B) * C. This involves
no charging information. To get the weighted value you have the change the weights according to the cold start phase weight, etc. (note that first cycle consumption is very high presumably due to lack of regen, but it has less weight, so the UDDSw number ends up higher than UDDSu!) - so you have to know the weights that are mandated, which can be looked up. Furthermore, there is no way to determine the EPA efficiency from these numbers alone (as you said later).
I don't think I was trying to prove anything at all, just trying to make sure we were on the same page...
View attachment 496731
Not exactly, but close (the range is
not the total range driven until the battery is discharged - the 420 miles
in the picture above). They do fully discharge the battery with the SS cycles. But anyway I think you understand, it's just hard to describe precisely in words what is done. It's an extrapolated range, fundamentally (as you can clearly see above). That's what I'm saying, and like I said, I think you understand.
I'm sticking with my 75%-baked theory
that they have just 18 power MOSFETs in the inverter of the 990, vs. 24 in the 980. So 75% as much current capacity. Those MOSFETs can be expensive (and you can probably get rid of some of the gate pre-drivers as well)! Here is an example - obviously Tesla would get them cheaper than this, but they are not cheap! It makes sense to remove MOSFETs if you don't need them, even if it has a very small negative impact on efficiency (they still have better range than the Model 3 originally had, due to motor drive efficiency improvements):
SCT3017ALHRC11 ROHM Semiconductor | Mouser
Anyway, it's all speculation until someone actually tears down a 990. It would save Tesla a lot of money to remove some SiC MOSFETs that aren't needed since that motor never generates as much torque or power as the 980. Since it's basically the same design it wouldn't be extra cost really. It's possible they depopulate the same board that is used in the other inverter but I don't think that would be required - building a different board is easy. Everything else in the drive unit could be the same.