Small modular reactor company Gen4 Energy will not pursue DOE SMR Licensing funding
Neste Oil sells first batch of NExBTL renewable diesel to the US market

Fallbrook Technologies Inc. developing variable speed supercharger drive using its CVP technology

Performance of supercharged Mustang with and without NuVinci CVP; variable ratio vs. 1:1. Source: Fallbrook. Click to enlarge.

Fallbrook Technologies Inc. is developing a variable speed supercharger utilizing its NuVinci continuously variable planetary (CVP) technology. (Earlier post.) Fallbrook has targeted and is soliciting select automotive OEMs for such a variable speed automotive supercharger.

Fallbrook says it has been working closely with a tier one automotive equipment supplier on the development of the device. Test results from that supplier have demonstrated potential fuel-saving, engine down-sizing and/or down-speeding opportunities without adversely affecting performance and drivability.

The NuVinci CVP uses a set of rotating and tilting balls positioned between the input and output components of a transmission that tilt to vary the speed of the transmission. Tilting the balls changes their contact diameters and varies the speed ratio.

Performance gains result from boost optimization over a wider power band, particularly at low engine speeds. NuVinci prototypes designed for use in an OEM application have also passed automotive class durability testing by the tier one supplier.

Schematic of variable speed supercharger drive. Click to enlarge.   Prototype variable speed supercharger drive. Click to enlarge.

Fallbrook believes, based on testing and independent analysis, that vehicle manufacturers can utilize smaller, more efficient engines with no loss in performance or drivability, thanks to the capability to tailor supercharger boost to driver demand offered by a NuVinci-enabled supercharger.

By controlling supercharger speed independent of engine speed, the NuVinci CVP enables ingestion of only the airflow required by the engine with little to no bypassing, thereby minimizing bypass losses and their associated NVH issues.

In light of the successful test results, Fallbrook and the tier one manufacturer are currently in discussions with potential OEM customers for the NuVinci supercharger drive. Fallbrook believes the drive can be packaged easily, as the current prototype is designed to mate with an existing supercharger line of products.

Fallbrook initially demonstrated its development of a variable speed supercharger drive by designing and building a prototype system coupling a NuVinci CVP with an aftermarket supercharger.

The demonstration car is a 2008 Mustang Bullitt, equipped with a ProCharger supercharger, and a NuVinci DeltaSeries continuously variable speed drive. With assistance in tuning by Lingenfelter Performance Engineering, it demonstrates considerable performance increases at lower engine speeds, when the variable speed drive is activated. The Bullitt prototype has logged more than 3,000 demonstration miles, and remains operational today for regular demonstrations.

Simulated performance of downsized 2.0L I4 with Fallbrook supercharger drive. Click to enlarge.

Using an SUV equipped with a 3.6L V6 engine as a baseline, Fallbrook has also simulated the use of a downsized 2.0L I4 engine which has been supercharged in production. In the graph at left, engine torque is depicted on the Y axis, and engine speed on the X axis. The blue dashed curve represents torque from the standard 3.6L V6 engine.

A normally aspirated 2.0L I4 engine would result in the lower curve, producing about 175 N·m peak. By supercharging this engine (grey curve), more than 330 N·m can be produced, approaching the 350 N·m capacity of the larger engine. However, this peak output only comes at the top end of the engine speed range; the vehicle remains underpowered throughout most of the range.

The green area is the chart represents the increased low-end torque generated by a downsized gasoline engine, as compared with the same engine without the supercharger. The red line indicates that the smaller NuVinci supercharger-equipped engine performs on a par with a larger engine.

The NuVinci supercharger drive is part of the NuVinci DeltaSeries line of accessory drive solutions. The NuVinci DeltaSeries line eliminates the compromise of fixed ratio accessory drives by de-coupling accessory RPM from engine RPM. Other DeltaSeries drives in development include applications for HD vehicle cooling fans, high output alternators, AC compressors, and engine crank-mount units, which control the speed of the entire accessory beltline.



@Roger Pham
I have not done any calculation of my own on an electric drive but I more or less anticipated that the losses would be greater than for a conventional system with a fluid coupling or the NuVinci system. People have tried various types of rubber couplings instead of a fluid coupling in conventional systems with mixed success. However, such couplings will also have losses. I would expect the losses in a 2-stage gear (at full load) to be somewhat lower than what you cite, presumably down to 2% in each stage. This would favor the NuVinci solution even more for a compound system.


Peter XX, I don't doubt all the results of the MIT report, which, as I read it, finds ICE several times more polluting than EVs.


Well, read it again (page 116). The 2030 EV uses more energy than the 2030 HEV. You should not compare a 2030 EV with an ICE of today. You should also note that the 2030 diesel ICE is better than the 2030 BEV and the 2030 gasoline ICE come fairly close. I bet both diesel and gasoline cars will all be hybrids (not necessarily electric hybrids) by 2030. So, what is the rationale for promoting BEVs?

Electrical transmission loss will be considerable. The very high frequency AC current from the turbocompounder generator will have to be rectified at about 10% loss. Loss in the generator will be another 10%.

When going from one AC machine to another (such as a PM or induction generator to an induction motor), a cycloconverter is an efficient way to move power.  Cycloconverter efficiencies can exceed 99%.

The 2030 EV uses more energy than the 2030 HEV.
Again, Peter XX repeats this assertion without showing where it came from (the graph on page 116 doesn't point back to the calculations or data sources; the battery performance figures in section 3.2 appear to have already been surpassed by progress, the studies cited as sources for Figure 14 go back to 2000, and there is no section on electric generation).

By 2030, the power mix of the world is likely to have changed a great deal.  If the USA went on a Manhattan project-level commitment to decarbonizing, thorium-based reactors could probably have replaced coal by then.  Even without nuclear, a system based on BEVs would allow penetration of intermittent generation (wind and solar) to upwards of 50%.  That MIT study is clearly not a source on which to rely.

That Peter XX is asking us to take a non-source as gospel says a lot about him.  He certainly doesn't accept such shoddy work from anyone else.

Roger Pham

At such a vastly different frequencies between the alternator and motor, I don't think the Cycloconverter in your reference would work.

Here is the calculation for rectifier's efficiency:
The rectifier efficiency can be easily calculated from the datasheet and input voltage.
For example, lets say Vf (forward voltage) of the diodes is 1V @ 4A RMS. If we consider an RMS in voltage of 100V, the efficiency becomes.
((100V*4A - 1V*4A) / 100V*4A)*100 => 99%.
Now consider a similar scenario, 4A RMS @ 10V RMS in.
((10V*4A - 1V*4A) / 10V*4A)*100 => 90%.
As you can see, the efficiency of a rectifier varies with loading. It depends on the voltage of the electrical system of your vehicle. For 12V system, don't expect efficiency much above 90%.
Losses in the alternator and motor for the sizes applicable in heavy-duty trucks' turbocompounder (10-15 kW range) is estimated at ~10% apiece. For larger synchronous PM motors, losses would be less, at ~5%. Invertor losses at 5-10% is typical. Grid-tied inverters for home solar PV panels may have losses as much as 15%!

Future ICE with 50% engine efficiency at typical usage coupled with a HEV drive train can have over 100 mpg efficiency. It can run on H2 on daily commuting basis with range of 150-200 miles needing to carry only 1.5-2kg of H2 on board. This is half of the H2 capacity of the Honda Clarity FCV. For long distance, fill up the tank with methane and expect 450-600-mile range. The H2 is made from non-fossil-fuel sources, as well as the methane. The methane can be synthesized from hydrocracking of biomass (adding non-fossil-fuel H2 to biomass to double or triple the energy output of the biomass) You can see that ICE will be relevant with us way into the future and just as green as BEV's. Any badmouthing of ICE and ICE engineers is not warranted.


I agree with EP here and in most cases.

Using NG as a fuel hybrid best case efficiency would be about 40% well to wheels (50% efficient engine and 80% efficient compression) While an EV could hit 50%+ (60% CCGT with 85% transmission/charging losses) H2FCV would be no better than the hybrid.

Back to turbo v superchargers, I think the beltless turbo option is a better option. Ideally with an electric motor on the engine crankshaft to increase the response and low engine speed response. I like the idea of a small electric supercharger for diesel engines when they are caught off boost

Roger Pham

Thanks, 3PS for the comparison and the opinion.
The efficiencies of BEV and ICE-HEV are close enough that if green energy sources are used, other factors come into play, such as cost and user satisfaction and user preferences.

Electric-driven charger and turbocompounder would be most efficient in an HEV with already high-voltage electrical system onboard. Yet, people who buy HEV don't usually care for turbo, and people who buy boosted and downsized engine don't care for HEV drive train. It's the law of diminishing return.


@ Roger,
There are Schottky Diodes, now improved as Silicon Carbide Schottky Diodes. They have smaller voltage drop, about 0.2-0.3 V, which yields much higher efficiency than 90% at 10V, from your calculations for worst case rectifier.

I'm not an engine expert, but from what I can read on engine downsizing, the Miller cycle engine, with electrically driven supercharger seem to be unbeatable in terms of simplicity, responsiveness and reliability. Not sure exactly about efficiency and costs.
Valeo offers some electric supercharger, they bought that business from an UK firm recently. It uses switched reluctance motor, capable of high speeds. (Boeing uses this type of e-motor to start aircraft engines to 20,000 rpm, and also as APU for power generation.)
One thing that needs to be added to this is source of electricity for supercharger motor. Some sort of starter generator, that can capture part of braking energy, would do.

For this NuVinci supercharger there is no data on slippage losses (CVT in cars are reportedly below 90% efficient). Also there is no data on NuVinci's CVT speed ratio range. Cannot match direct electric drive for reliability either.
Nice thing about electric drives is that they allow short term overload, usually up to 30 sec, often allow 100% overload (depends of motor type). This matches passenger cars acceleration needs, when most power is needed, acceleration bursts usually last less than 10 sec.


Roger, two things:

  1. Cycloconverters REQUIRE substantial frequency differences in order to function properly.  The minimum IIRC is about 3:1.
  2. The connection between the compounding turbine's generator and the motor needn't operate at legacy DC bus voltages.  BAS-II's battery is 115 volts.  The induction motor can take a lot more voltage when running far above cranking speeds.

Roger Pham

Thanks to MG and EP for your contributions.

One of the excitement about the NuVinci is in the after-market supercharger installation. The mechanical device is easy for a small shop to work with, but an electric-driven supercharger may be too complicated for a small shop and may cost a lot more, including a significant upgrade of the electrical system. Roote-style charger is real inefficient, while twin-screw is real expensive, while centrifugal charger without variable ratio won't deliver the low-engine-speed torque.

Reliability-wise, there is no indication that it won't match the reliability of a direct electric drive. The balls and the bells are protected from wear with a special fluid film, while the bearings are like any ball bearings in any electric motor. Electric motor can fail from over heating from various reasons as well as from other malfunctions in the electrical system.


Well, now you have the source for the EV/HEV comparison. Why do you complain? Read it. There are answers to all your questions. Why do you ask me questions that are already answered in the report? For example, MIT use the 2030 electricity production.


Projections, guesses, 18 years into the future (2030) don't mean jack.


Sounds as you are a bad looser!


Peter XX, cite a source that's NOT BEHIND A PAYWALL so that EVERYONE can verify what you're talking about.

Even with simple-cycle gas turbines, the LMS100 burning NG at 46% efficiency is going to have far lower CO2/mile than a gasoline-powered hybrid at ~40% efficiency.  This page has a comparison of CO2 per GJ of various fuels; gasoline is 85 kg/GJ, NG is just 63.

In other words, it's patently obvious that the claims you're making come from an erroneous source.  Stop using paywalls to hide your mistakes.


Once again, I have to repeat that the main report is public and can be downloaded via the link below. It is obvious that you know that but you just want to mislead other readers on this site. EVERYONE CAN NOW VERIFY THAT YOU LIE. Please everyone, download the report and read it. It is for free!

This is the most comprehensive study in the USA. It is made by the most experienced and well-renowned research group in this field in the USA. John B. Heywood is the most knowledgeable and respected professor in engine and drivetrain technology in the USA.

So, here you come as a Poet trying to dismiss this study with a few lines. I have much more faith in the MIT research group.


Peter XX, how many times do I have to tell you that the graph in the paper you keep citing does NOT list any sources or calculations for the data points on it?

A gas-fired CCGT has a pipeline-to-bus carbon emission approximately half of the TTW emission of even a hybrid (0.75 CO2/GJ * 2/3 heat rate = 0.5 emissions).  You can bluster and obfuscate all you want (as I can see you do), but pretentiousness only makes you ridiculous.

As for "slow pace of change in the electric power industry", in February the year-on-year consumption of natural gas in generation was up 33.6%, coal DOWN 13.6%, and that despite many more nuclear outages (source).

Note, that change happened in ONE year.


Well how many times must I tell you that all the calculations are documented in the report? The report is on 153 pages and it contains all the necessary information. I put in the link once again. Read the report!!!

Eventually, it is now you against Professor John B. Haywood. You lose!!!


What page are the relevant calculations on?  I've been through the report and the relevant appendices and not found them.  Section 8.3.2 refers to Figure 56 and not the reverse.


Well, you must be blind.


Way to not answer the question.


I have used data from the MIT report as basis for an own study a couple of years ago. It was possible for me to both follow the calculations and re-calculate anything I needed from the report. Why can you not do the same? I have used the same simulation program, Advisor, myself for calculations of energy use in vehicles and found very good correlation with experimental data, so I do have some faith in their simulations.

This time, you do not ask a specific question. How, can I answer something that you do not specify? Furthermore, I see no need to do that. All data used in the MIT calculations are available or else, they refer to a reference where they have received this material. Any assumption is clearly stated. What more do you ask for? I think you just want to uphold this discussion for some reason, instead of admitting that you have been totally wrong in so many of your previous statements on this site. If you want to prove that MIT has made errors in their calculations, go ahead and make your own calculations. And, by the way, good luck!

For the moment, I have other things to do; I have a job as well, so I will not be active on GCC until late this weekend.

I have used data from the MIT report as basis for an own study a couple of years ago.
Well isn't that special.  While I was digging for those calculations you said were there, I found Table 58 (just click to open the paper to that page) which gives CO2 emission figures which contradict the claims given on page 116.

I found some meat in Table 63 (Appendix 2, page 145) which provides a cross-check.  The baseline BEV has 40 kWh of usable capacity (Table 4, page 30) for a 200 mile (320 km) range (Table 5, page 31), for an energy consumption of 200 Wh/mi or 124 Wh/km.  But Table 63 claims 0.54 MJ (150 Wh) per km.  This is inconsistent.

The carbon emissions figures are also all wrong.  Table 3 (page 28) gives a WTT CO2 emissions figure of 213.6 g/MJ.  However, if we compare this against the carbon intensity of natural gas (63 g/MJ) divided by the efficiency of a modern CCGT, we get 105 g/MJ at the busbar and likely 150 g/MJ or less at the wheels.  At 124 Wh (0.45 MJ) per km, BEV emissions are around 68 g/km, around half of what's claimed by Figure 53.

Anyone can go through this and verify that what I've said is true.  The study is bogus.  I doubt that Heywood is directly responsible for this; he probably involved some underlings who did sloppy work.

You, on the other hand, should know better.  You're the self-appointed expert, with a PhD no less.  Much more of this and people will start wondering what correspondence school it came from.

The comments to this entry are closed.