Hey gald - that looks like a blast, thanks for posting the video, I really like the overlay too.
I'm telling you man, dial up the amps and it will be better than you expected. I'm pretty sure you said you had cooling on the batteries, so the batts can handle it, the controller can handle it, the motor can handle it; assuming your drivetrain can handle it, it's all good. Since I don't think your motor is cooled, I'd check it after a few runs at each new power level, to get a sense of the thermal load.
Oh, wait, did you say you have not run-in the brushes? Maybe wait on that advice I just gave... What does the patina look like on your commutator, can you post a few pictures? The brushes need to be worn down/shaped to maximize their contact area with the commutator, and a good patina on the commutator helps transfer watts better. Don't remove the brushes though, it shortens their life. Lift the rear wheels, and run it in first gear on a directly-connected 12v battery for a few days - you'll have to swap out batteries a few times - and you will be good to go. Blow at the carbon dust afterwards, so you don't risk a flashover.
So 149v/42cells = 3.54v/ cell, which is equivalent to 340.57v on the 96-cell pack tested for the EREV discharge curve, which correlates to taking 35ah from a fully discharged pack. Since your pack wasn't fully charged, 165v /42 cells = 3.93v/cell, which is equivalent to 377v on the same discharge curve, where about 10ah would have been pulled.
I would speculate that you pulled 35ah-10ah=25ah from the pack out of a potential 42ah usable in the pack (or 47ah if you discharge past the "knee" in the discharge curve, as many do but I do not). This assessment is pretty crude because there is a lot of extrapolation going on against a random discharge curve under load, but that is the best I can do with the data you have so far, hope it helps.
Also keep in mind that AH is not a good gauge of power consumption, because it does not consider the volt side of the equation (volt*amps=watts), so you are getting far more watts at the start of the discharge when volts are high, far less when the pack is depleted and voltage is low. At the same number of watts needed to maintain a given speed, AH will go by much faster at the end of the discharge, so when you have used 1/2 the ah you have actually used up 2/3 of the pack's capacity. Watt hours (WH) are a more effective method of measuring discharge. I use a CycleAnalyst V3 for my "gas gauge"/coulomb counter, so I always have a sense of how many WH I've taken from the pack, and how many I have left.
I'd be really interested to see the sag under a given discharge - like at a 100a load, how much does the voltage drop from resting voltage - but I don't think you can capture that yet.
Hillhater said:
Even the discharge curve you linked to showed there was 6Ahr available above 4 volts ?...(13% of capacity)
So on the discharge curve on page 2 of the link I supplied earlier ( http://avt.inl.gov/pdf/EREV/batteryvolt3929.pdf ), the curve starts at 4.11v/cell in a 96 cell pack, which fits the charge setting I recommended. Not sure where the 4.0v figure came into the discussion, but the if the Curtis cutoff is at 170v, that translates to 4.047v, which is 388.47v on the pack's discharge curve. By my eye, that is 2 or 3ah from fully charged; suboptimal, but between a fixed 170v cutoff and difficult-to-alter pack, the point is moot.
Even if it were possible, removing 4 cells from the pack as you recommended would decrease 9.5% of the pack's capacity, I'd prefer leaving 5% on the table by not charging to the peak rating. In addition, while peak charging is useful for occasional top balancing it ages the cell, so usually I don't fully charge my cells to extend the life of my packs.
Hillhater said:
JD, no argument here, ... But , most all lipo capacity maps show there is useful capacity above 4.0v ...
Again, I would caution against making assertions about "LiPo" - a123/thunder sky/CALB LiFePO4 are also "LiPo", but their peak charge is 3.65v, so I assure you that their discharge curves (not capacity maps) do not show any useful capacity above 4.0v.
-JD