Proper Efficiency Gauge

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

johntaves

Well-known member
Joined
Nov 9, 2011
Messages
52
I posted "Please complain to Nissan to fix the climate control". In that thread, I argued that the white bubbles are useless and that Nissan should have provided a proper efficiency gauge. It seemed clear to me that most readers did not comprehend what a proper efficiency gauge would do and show, so I thought I would describe it here.

With the current white bubbles, you have no way of knowing how many bubbles will maximize efficiency when accelerating or decelerating. We also can't determine easily what speed is the most efficient and how costly other speeds, either faster, or slower are.

A proper efficiency gauge would measure how much of the electricity going to the motor is going into kinetic and potential energy. To do this, they need to measure the acceleration of the car. They can do that by reading the speed, but that would not handle hills. If they put an accelerometer into the car, it would handle hills fine. They would want to know the mass of the car and passengers. If they assumed some constant for the mass, and the mass was greater, then acceleration would appear very poor but deceleration would appear correspondingly better. I suspect that they could easily estimate the mass after one acceleration/deceleration cycle after any door opened/closed event and adjust the display.

This gauge would not need a reset feature like the gauge on the main screen, because it would not need to average energy use over some time. It would measure the lost energy every fraction of a second. It would not jump up and down as you accelerate and decelerate like all MPG gauges do and like the MPKw gauge does, because the lost energy does not jump up and down with the pedal.

I am guessing that there's little difference between accelerating hard vs soft, and thus the gauge would read a relatively similar value regardless of how hard you are pushing on the go pedal while you accelerate. As the speed increases the gauge would initially read a poor efficiency because going 1 mph is probably very inefficient, then peak as you pass through the optimal cruising speed and decline as you go faster.

The gauge could have 3 different readouts, or modes.

1) Percent efficiency - You simply maximize this percent at all times and you are doing the best you can possibly do. This could also be expressed in terms of Miles/Kw.

2) Projected range (round trip) - This would tell you how far you can go if you continue driving like this and return home. In this readout, it assumes the end point will be at the same altitude as the starting point. For example, if you have 60 miles of juice, and you drive up some big hill for 30 miles the current GOM might read 20 miles left. In contrast, this gauge would read 30 miles left, because this gauge would know you are up a big hill. The very cool thing about this gauge is that it will instantly tell you what speed to go at. If you accelerate up to 70mph, the gauge will drop immediately to tell you how far you can go at 70mph. If you relax and drop back down to 60mph, the gauge will rise.

3) Projected range (one way) - A basic implementation would in effect discard the built up potential energy, so the miles to empty would read 20 in the hill example above. A cooler implementation would use the Nav, your destination, and altitude information to know what to expect.

The engineers could do a bunch of cool things with the nav system. I am not certain, but I assume that the map data has road speeds, and the routing is optimized for time. This could route for best energy consumption. Knowing the amount of energy in the pack, and whether this is a round trip or not (electricity at the destination), the nav could tell you whether you need to limit your speed or not.
 
What you are proposing would be far less accurate using some of those metrics and the ones that are useful need many others added to the mix. Without going into great detail the car could use an accurate SOC meter but that would vary up and down today with the measuring devices they presently use. There are more costly ways to get a more accurate SOC but Nissan chose to do it the way they did and it is not bad compared to other EVs out there. There are ways to refine this but it is very costly in terms of "range remaining". I have a friend that developed very advanced software to do this for EVs and had a contract with a major auto maker to provide it but they are not implementing it because of the complexity and cost. This can be debated on and on but for the most part the system does a decent job in relative terms if you understand how it behaves. If one pays attention to the car it is easy to predict range with a bit of experience, after 5K mile son this car I have had no issue even with highly varied terrain. There are so many variables to dynamically measure it makes this a difficult tool to develop and he driver is always the wild card parameter.
 
EVDRIVER said:
What you are proposing would be far less accurate using some of those metrics and the ones that are useful need many others added to the mix.
Is there something stopping you from explaining why these assertions might be true?

An accelerometer does not seem expensive, since iPhones have them, and writing the software to do the math seems easy enough. I don't see how any accuracy will be lost by doing this. The accelerometer will certainly have an accumulated error, but even if you ignore that, this gauge should be much better than simply measuring the energy in/out of the motor/battery. It would be give you a steady readout, regardless of your pedal movements and hills, of how much energy it is costing you to go the speed you are going.

Getting rid of the accumulated error should be easy enough on flat ground by simply using the speedometer. The hill accumulated error can be eliminated if the nav has altitude info, and if not, by simply keeping track of your route and zeroing out the error when you return to the same location. Since most routes follow the same path there and back, this should be excellent for most routes.
 
johntaves said:
EVDRIVER said:
What you are proposing would be far less accurate using some of those metrics and the ones that are useful need many others added to the mix.
Is there something stopping you from explaining why these assertions might be true?

An accelerometer does not seem expensive, since iPhones have them, and writing the software to do the math seems easy enough. I don't see how any accuracy will be lost by doing this. The accelerometer will certainly have an accumulated error, but even if you ignore that, this gauge should be much better than simply measuring the energy in/out of the motor/battery. It would be give you a steady readout, regardless of your pedal movements and hills, of how much energy it is costing you to go the speed you are going.

Getting rid of the accumulated error should be easy enough on flat ground by simply using the speedometer. The hill accumulated error can be eliminated if the nav has altitude info, and if not, by simply keeping track of your route and zeroing out the error when you return to the same location. Since most routes follow the same path there and back, this should be excellent for most routes.


One can use GPS elevations to assist in range but how do you propose an accelerometer is going to help? Please be specific on implementation and calculations. GPS elevation can help but again it is very contingent on the drivers habits, speed at the grades, car weight load, etc. There are so many variables and the best estimate is an educated guess based on the elevation ahead and driver experience. This is a very long topic with much technical discussion that would take pages and pages to explain in detail, the bottom line is that the tools in the car need some interpretation and making an EV range meter "idiot proof" is a very expensive and data rich undertaking at this point in time.
 
Take total Battery Pack energy usage per some small unit of distance (from the RPM value) and that accounts for all power uses to give "instantaneous" efficiency. No other instrumentation is very meaningful.

Determining how much of the usage is due to:
Tire drag
Air drag
Headwind or Sidewind
Climbing
Passenger Load
acceleration
etc.
is interesting, but not very useful to the average driver.

What is the Goal of using your suggested "instrumentation"?

It might be interesting to know if (or how fast) one is changing altitude, but how would that help a driver with efficiency?
 
EVDRIVER said:
and making an EV range meter "idiot proof" is a very expensive and data rich undertaking at this point in time.
I think you might be misunderstanding the gauge I am suggesting. This is not about making a better EV range meter. This is a gauge that displays your efficiency, and the 3 modes are simply different ways of displaying the information. The last two modes, show the information in terms of range, but that does not mean that I am saying it will be a perfect or accurate range gauge. It should be more accurate than the range gauge we have now, simply because this could keep track of altitude and factor that in.

For example, I want to know the cost of driving at 70mph vs 60mph, and I want that information in terms of what that will do to my total range. Let's say the range meter is reading 50 miles when I am doing 60mph, I then accelerate up to 70mph and now it reads 40 miles, I then drop back down to 60 and it reads 50 again. This gauge would not dramatically jump around as I push on the pedal to accelerate or as I encounter hills. It would generally read a steady number in proportion to my speed. The total range number might be inaccurate, but the difference between the numbers I get when doing 70 and 60 would be very good.

Additionally, it would tell us the optimal amount of acceleration and deceleration to get the best efficiency, and more importantly it would tell us the relative cost of those choices. Right now we really have no clue whether stomping on the go pedal is dramatically more inefficient than accelerating gently. For example, I hit the go pedal hard so that there are say 8 white bubbles, and as I accelerate from say 5 to 50mph, the gauge reads 60 miles of range. I then hit the brakes relatively hard and as the car drops from 50 back to say 5 and again the gauge reads a reasonably steady 60 throughout that experiment. Now repeat with a gentle acceleration, say 3 white bubbles. Maybe the gauge reads 65 miles of range as you do this. You've now learned the cost of accelerating and decelerating hard. With the current gauges, people are fooled into believing that accelerating is very inefficient. You might be accelerating too gently to get the best range. We just don't know without this gauge.
 
johntaves said:
EVDRIVER said:
and making an EV range meter "idiot proof" is a very expensive and data rich undertaking at this point in time.
I think you might be misunderstanding the gauge I am suggesting. This is not about making a better EV range meter. This is a gauge that displays your efficiency, and the 3 modes are simply different ways of displaying the information. The last two modes, show the information in terms of range, but that does not mean that I am saying it will be a perfect or accurate range gauge. It should be more accurate than the range gauge we have now, simply because this could keep track of altitude and factor that in.

For example, I want to know the cost of driving at 70mph vs 60mph, and I want that information in terms of what that will do to my total range. Let's say the range meter is reading 50 miles when I am doing 60mph, I then accelerate up to 70mph and now it reads 40 miles, I then drop back down to 60 and it reads 50 again. This gauge would not dramatically jump around as I push on the pedal to accelerate or as I encounter hills. It would generally read a steady number in proportion to my speed. The total range number might be inaccurate, but the difference between the numbers I get when doing 70 and 60 would be very good.

Additionally, it would tell us the optimal amount of acceleration and deceleration to get the best efficiency, and more importantly it would tell us the relative cost of those choices. Right now we really have no clue whether stomping on the go pedal is dramatically more inefficient than accelerating gently. For example, I hit the go pedal hard so that there are say 8 white bubbles, and as I accelerate from say 5 to 50mph, the gauge reads 60 miles of range. I then hit the brakes relatively hard and as the car drops from 50 back to say 5 and again the gauge reads a reasonably steady 60 throughout that experiment. Now repeat with a gentle acceleration, say 3 white bubbles. Maybe the gauge reads 65 miles of range as you do this. You've now learned the cost of accelerating and decelerating hard. With the current gauges, people are fooled into believing that accelerating is very inefficient. You might be accelerating too gently to get the best range. We just don't know without this gauge.

The easiest way is a real-time kw meter. You can see exactly how much power you are using in real time. Once you watch that you can tell the difference in real time. This has worked well in EVs for many years and is simple with immediate feedback. This can be done in the LEAF in addition to the generic bubbles. I would bet this would satisfy your need 100% but Nissan felt that was too detailed for most people. This info shows what is really happening under all conditions, I prefer to have this on the entire car load so one can turn things on and off and see the immediate result as one number in detail. You can also switch this to kw per mile, etc.
 
garygid said:
Take total Battery Pack energy usage per some small unit of distance (from the RPM value) and that accounts for all power uses to give "instantaneous" efficiency. No other instrumentation is very meaningful.

Determining how much of the usage is due to:
Tire drag
Air drag
Headwind or Sidewind
Climbing
Passenger Load
acceleration
etc.
is interesting, but not very useful to the average driver.

What is the Goal of using your suggested "instrumentation"?

It might be interesting to know if (or how fast) one is changing altitude, but how would that help a driver with efficiency?
Acceleration, passenger load, and climbing are not energy sinks. Acceleration is the conversion of chemical potential energy into kinetic energy and is not a loss. Climbing is the conversion to potential energy, and passenger load simply affects the ratio of speed to kinetic energy and height to potential energy. Passenger load will affect tire drag, but not dramatically.

Imagine that there are no friction losses. With the current gauges we will see roughly what we see today. When you accelerate, it will read that energy is going out with some finite range, and when you decelerate you will see energy coming back in and infinite range. It won't be obvious that all the energy is coming back in. With the gauge I am suggesting, it will read 100% efficiency and infinite range all the time. Now, clearly neither gauge is needed in that situation, but the point is that the conventional gauge that we have, and that you described above, is a misleading liar.
 
EVDRIVER said:
The easiest way is a real-time kw meter. You can see exactly how much power you are using in real time. Once you watch that you can tell the difference in real time. This has worked well in EVs for many years and is simple with immediate feedback. This can be done in the LEAF in addition to the generic bubbles. I would bet this would satisfy your need 100% but Nissan felt that was too detailed for most people. This info shows what is really happening under all conditions, I prefer to have this on the entire car load so one can turn things on and off and see the immediate result as one number in detail. You can also switch this to kw per mile, etc.
I am still not convinced you understand what I am suggesting. I don't have a problem displaying the information in units of Kw/miles or miles/kw. Kw alone is most certainly not useful for efficiency. Efficiency is the measure of how much energy is lost to go some distance (or whatever work you want that energy to do), so if you don't put miles in the numerator or denominator you don't have efficiency information.

More importantly, the car already has a mile/kw gauge, but they don't calculate this properly. They don't factor in the acceleration, which means that if I want to get some miles/kw number for some scenario off that gauge, I have to ensure to read the gauge at the same altitude and speed that I reset it at. In other words, I have to ensure that during the sampling period I have ensured that there was no kinetic or potential energy change.

If you are telling me that other EVs have this gauge, but the Leaf does not, then maybe you do understand. If you are telling me that the Leaf already has the gauge, then you don't understand what I am saying.
 
The F1.08 SOC-Meter displays "instantaneous" total Power from (and to) the Battery Pack.

Preview:
The next version of the SOC-Meter firmware should show RPM (probably in rpm up to 9999). Also, if I can get a good time-base to use, HOPEFULLY it will show some integrated quantities: Miles (from SOC On, in hundredths), and hopefully "instantaneous" miles/kWh (possibly showing hundredths of miles) ... and possibly watt-hours per mile.

In addition (thanks to TickTock) it may show "Braking Force" as Total, Mechanical, and Regen. The Mechanical and Regen might also be shown as a % of the total.

I had hoped to find Battery Temperature, but have not yet identified it.
 
EVDRIVER said:
The easiest way is a real-time kw meter. You can see exactly how much power you are using in real time. Once you watch that you can tell the difference in real time. This has worked well in EVs for many years and is simple with immediate feedback. This can be done in the LEAF in addition to the generic bubbles. I would bet this would satisfy your need 100% but Nissan felt that was too detailed for most people. This info shows what is really happening under all conditions, I prefer to have this on the entire car load so one can turn things on and off and see the immediate result as one number in detail. You can also switch this to kw per mile, etc.
It is presented graphically, but the LEAF does have three kW meters: one for the drivetrain, one for CC and one for accessories, all on the same display screen. I find it interesting to watch but haven't decided how useful it is with regard to climbing and descending hills or general driving efficiency. The CC meter sure is helpful though.

This doesn't address the OP's concerns though.
 
Oh, you are suggesting that a "Level, Constant-Speed Efficiency" Gauge would be useful?

Taking instantaneous energy usage and subtracting out any used for acceleration or climbing?
 
garygid said:
The F1.08 SOC-Meter displays "instantaneous" total Power from (and to) the Battery Pack.
What is this? If this is a 3rd party meter of some sort, why not put an accelerometer in it and do what I am suggesting? You'll probably need speed to help with accumulated error.
 
EVDRIVER said:
Please be specific on implementation and calculations.
Roughly speaking:

driving efficiency = D / (Kw - A)

D = Distance
Kw = energy going to the motor.
A = energy measured from the accelerometer.

This will work just fine if the energy is going into the battery from the motor (regen). The Kw will be negative and so will A.

A will have both potential energy and kinetic energy. If the computer compares A to the speed of the car, the computer can tell you the altitude change, and thus how much of that energy is kinetic and how much is potential.

Current gauges on the Leaf simply have S/Kw, or even more lame, just Kw, and thus the driver has to either ensure that A = 0 during the sampling period, or they have to average over a long time so that A is insignificant.
 
garygid said:
Oh, you are suggesting that a "Level, Constant-Speed Efficiency" Gauge would be useful?

Taking instantaneous energy usage and subtracting out any used for acceleration or climbing?
Yes.

I am not sure we want to call it a "level, contant-speed efficiency gauge". I'll bet that would confuse many.

I am sure that if this gauge was the only energy gauge with respect to motion (not counting radio, lights, etc), nobody would be asking for the gauges Nissan gave us in the Leaf today.
 
If you are climbing a slope at a constant speed, the accelerometer reads zero.

Integrating three axis very carefully gives three velocities, but it requires some rather high precision circuitry and computing, right?

A very slight drift in the integration after going up and down hills a number of times might leave the system thinking it was steadily climbing, or decending.

Sure, submarines can tell where they are, but what can one do with a $5 part and 1:255 resolution?
 
garygid said:
If you are climbing a slope at a constant speed, the accelerometer reads zero.
Right, I guess I meant a level gauge. The car can't go up a hill without raising it's nose.
garygid said:
Integrating three axis very carefully gives three velocities, but it requires some rather high precision circuitry and computing, right?

A very slight drift in the integration after going up and down hills a number of times might leave the system thinking it was steadily climbing, or decending.

Sure, submarines can tell where they are, but what can one do with a $5 part and 1:255 resolution?
Even if the acceleration (both kinetic and potential) was off by 10% the gauge would still be a huge improvement over the junk we have now. The accumulated error would be hideous, but that does not really affect the benefits of the gauge except for display option #2 on the original posting. Your comments about drift are the same as I am saying above with respect to accumulated error, and to recap, there should be plenty of opportunity for the computer to zero out the accumulated error.

I don't know the accuracy/cost ratio for these parts. My guess is that they are not expensive and would provide plenty of accuracy. I doubt they would be off by any thing like the hideous 10% I suggested above.

To a large degree, the acceleration that can be measured from the speedometer would be the bulk of the benefit of this gauge. The level gauge/accelerometer to determine altitude would be icing on the cake.
 
This entire thread after much back and forth will end up with same conclusions as all the others on this topic. This will go round and round here and again in new threads. In less time it takes to post this I can look at my car instruments and estimate my range well enough to get to where I need to go regardless of the terrain. Since I don't drive my LEAF hundreds of miles from my home to distant unknown places it is quite easy to estimate my range with a decent level of accuracy with the tools in place now. It would be nice to know the actual SOC percent but beyond that this is more about people needing to adjust to EV driving and not expecting the car to tell the very last few miles under constantly changing conditions. This is about some basic experience (using the tools in the car) and knowing your car and driving habits a bit. At times people can't believe the range I get on some trips and I can't believe their range is so poor with all their coasting and other nonsense. It comes down to driving skill and observing the car consumption under different conditions and paying attention. The word hyper miler will never be in my vocabulary and I won't be worrying about pushing my car limits within reason. Sure I would like numbers over bubbles because of the better feedback but at this point I already know how the car behaves. The feedback for real time efficiency is quite clear on the car, what is lacking is more detail.

Nissan may not have mentioned this but there is a learning curve to driving an EV.
 
EVDRIVER said:
This entire thread after much back and forth will end up with same conclusions as all the others on this topic.
It seems that garygid and dgpcolorado have comprehended what I am suggesting. You have not provided any details to refute the concepts I have stated. Again, this does not create some perfect range gauge.

I searched the forum for something like this but failed to find it. If there is some thread that discusses this concept, please point it out. I did find some thread that was huge and discussed the most efficient way to drive. It started with some table that showed the efficiency (miles/kw) at different speeds. This gauge would deliver that information.
EVDRIVER said:
The feedback for real time efficiency is quite clear on the car, what is lacking is more detail.
This makes it clear that you do not quite get the concept that I am saying. An analogy might help. Imagine a ICE car that had a 5 speed manual. The car has a tach, but no speedometer. I am stating that the car should have a speedometer. You are stating that the information is within the tach and you are capable of getting the speed information out of it just fine. It is true that you can determine the speed from the tach. You need to know the gear ratios and the tire diameter, and then do some math. Or you could make a few mental notes about what gear and what RPM is right for the common speeds we drive. So, technically the information is there, but clearly any auto manufacturer that attempted to pass off a tach as a speedometer would be laughed out of town. The gauges that Nissan gave the Leaf to determine efficiency are just as idiotic.
EVDRIVER said:
Nissan may not have mentioned this but there is a learning curve to driving an EV.
No, there isn't. The efficiency is overwhelmingly determined by the speed, thus if the unlearned drives at the speed limit and so does the efficiency expert, the two will have minimal difference in range. Everyone's natural instinct to slow down to conserve energy is correct. If you want to split hairs and get even more range out if it, then you do need to do the crap described in the other thread where people talk about how to get the best range, but that is only necessary because Nissan did not provide a proper efficiency gauge.

In short, the only "learning curve" is entirely caused by the fact that they did not provide the proper gauge.
 
Back
Top