Tuning an EV

[PixelpeepS3]

in my extremely limited understanding of all things spanner like:

Are all EV's detuned to improve range? - do the motors have anything in reserve?

Could you just increase the power which would decrease the range?

i was thinking about the ludicrous mode on the Tesla and wondered if that theory could be used on other models for a 'quick burst' ?

I get that there are limitations on what the motors/drive train can produce but similarly to mapping ICE cars, is there much wiggle room?

[PixelpeepS3]

Max_Torque said:

In simple terms, the power of an electric motor is limited (fundamentally) by it's supply voltage and it's thermal impedance.

Because modern power electronics can control the motor current to very very close tollerances, most EVs will be running very close to the absolute maximum short term (ie not thermally limited) performance available at their chosen supply voltage. Long term performance (thermally limited) will simply be the maximum heat flux that the motor can experience before going over temperature, and that is broadly set by the mechanical architecture of the machine. (too hot, and the electrical insulation on the stator fails (bad!) and the Permanent Magnets lose their magnetism (bad!). Within a modern eMachine are various sensors that measure critical temperatures within the machine, and in conjunction with other temperatures that are modelled, the power electronics will automatically de-rate to prevent that sort of failure.

Practically speaking, there are Four main "choke" points in the system, any or all of which can be the limiting factor:

1) The battery - normally, a battery system is calibrated to be very safe to maximise it's life. As more real world useage data is gathered, OEMs are starting to push the limits a bit more as in the real world, battery desegregation has been much less than originally feared. Hence, if you could get access to the battery management control software, you could almost certainly pull more power than std from the battery (limiting factors could also be the fusing, current measurement devices and even things like connector capability). A complete change in string architecture (more series cells) could furnish a higher nominal voltage, supplying more power capability, but the knock on effect could be enormous (OnBoardCharger might not be rated for the extra votltage, nor the DC;DC that drives the LV battery (12v systems etc).

2) The Inverter - There is really no practical difference between short and long term capability for an inverter. Because the active silicon switching area is so small (just a couple of square cm of semiconductor are switching several hundred kW!) it effectively has no thermal inertia. Over current events for as little as 100uS (1/10,000th of a second!) can result in failure. Because power silicon is expensive, typically the inverter current limit is the systems primary design point and ultimate limiting factor. In conjunction, things like phase current sensors are sized for the max current capability, so they would almost certainly have to be modified or replaced with larger range devices if the current is increased. Best option here would be to replace at least the power section of the inverter with a higher rated one. Currently, most power silicon has been developed for the industrial sector, and hence comes in two voltage ratings (600v and 1200v) to suit the normal AC line supply peaks. Most EV's currently use 600V silicon, but all systems will leave at least a 20% headroom from peak battery voltage to allow for dynamic effects. Hence most EV sit around 400 odd volts nominally. This is climbing as OEMs gain experience and want to get more power (and efficiency) from their systems

3) The eMachine itself. The winding architecture, the max rotational speed capability, and the insulation rating are all specified for a nominal supply voltage. Although increases in supply voltage may not cause immediate catastrophic failure, the life of the eMachine will most likely be shortened. Thermally, the only real way to improve the long term perf is to force cool the machine, ie, fit a system that can feed a lower temperature coolant to the eMachine. The colder, the better! (although depending on the machines sealing / venting arrangements, avoiding a condensing atmosphere is advisable)


4) The control system - Here's the biggy. All the system components tend to be controlled, and crucially, monitored by some higher level supervisor. That controller will monitor the reported state of each part, and generally can turn everything off if it senses any significant off-nominal behavior. Ie, say you crudely shunt the phase current sensors in the inverter by 10% (ie, they will report 10% less current than is actually flowing in the motor phases). You (may) get an extra 10% machine torque, but chances are, when the battery managements reports 10% more current for any given load the supervisor will call foul and open the safety contactors! So, then you also need to shunt the battery current sensors by the same 10%. But now, you fuel gauge and battery SoC calculations are wrong, and who wants to drive an EV that stops dead unexpectedly when it hits the critical low battery voltage limit before the dash says you are out of juice! So, now you need to re-calibrate the SoC calcs in the supervisor too!

So, yes, EVs are eminently tuneable, but imo, it's such a PITA to do so, requiring multiple units to be physically modified and re-calibrated, why would you bother? And it's not like EVs are slow in the first place. It's also worth noting that you are very, very unlikely to be able to increase the system efficiency with a re-map (unlike for some ICEs, where there is a trade off between tailpipe pollutants and fuel consumption). Typically, most current systems are mapped to within 0.5% of absolute optimum, so on say a 100mile range EV, an remap might get you another half mile down the road...

If i said i was hoping you'd be one of the people to contribute to this thread when i originally posted, it wouldn't be a lie

Thanks Max - do you think ICE mapping had the same 'kind' of barriers when they first started exploring tuning modern engines?

I think the many systems on modern ice cars would be similar to the above, ie seeing odd readings from other 'departments' - raising rev limiter would be one as an example.

With regard to 'fast enough' ...

its NEVER fast enough

i wouldn't mind exploring upgrading the brakes and chassis as these seem to be the limiting factor to anything spirited