Delphi-Led Team Developing Next-Generation Low-Cost, Compact, High-Temperature Propulsion Inverter
10 June 2008
 |
| The Delphi-led project seeks to leverage technology innovation in multiple areas to drive down the cost of future propulsion inverters. Click to enlarge. |
Delphi Corporation is leading an industry-government team to develop next-generation propulsion inverter technology to reduce the cost and size of current inverters by 50% or more. The new inverters are targeted for use used on the next generation high-efficiency hybrid electric vehicles (HEVs) and plug-in hybrids (PHEVs). In the longer-term, the inverter will be used for fuel cell vehicles (FCVs).
In addition to reducing the cost and size of the inverter system, the Delphi team is seeking to enable the system to operate at normal engine coolant-loop temperatures of 105-120°C to help reduce other system costs and the space needed to cool today’s inverters.
The team is also looking at improving manufacturability. Today’s power silicon modules contain too many parts, and therefore too many manufacturing steps.
Propulsion inverters provide phased AC (Alternating Current) power for hybrid vehicle traction motors and generators, as well as auxiliary pumps and drives. The propulsion inverter enables precise control over electrical power flow from the battery to the electric motor.
The primary team members for the project include: Delphi for the inverter design, packaging, thermal management, mechanical integration, build, test and assessment of cost to manufacture; Dow Corning and GeneSiC for silicon carbide-on-silicon power semiconductor devices; General Electric for high-temperature thin-film DC buss capacitors; Argonne National Lab for ceramic capacitors; and Oak Ridge National Lab for characterization of power semiconductor devices, modeling, simulation and evaluation of alternative inverter topologies, and system testing.
In the area of the constituent power semiconductors, the team is taking a dual path to identify power devices and packaging suitable for the application: silicon and SiC-on-Silicon (SiC/Si). The team is also taking a dual path to identify a capacitor design and packaging suitable for inverter applications: film capacitors and film-on-foil capacitors.
The Delphi-led research and design team will contribute $3 million and receive $5 million in funding from the DOE, resulting in an $8 million project that will run for three years.
This first year is focused on identifying requirements and developing the concepts. Testing and evaluation the component technologies is targeted for 2009, with the design, build and test of an inverter to meet DOE requirements slated for 2010.
Resources
When any new technology is adopted, there are many swift and predictable efficiency improvementsies to be gained. As the technology matures, improvements become more and more difficult and costly to achieve.
The low lying fruit is always easier to pick. Electrified auto technology is in that early phase of large gains; while ICE technology is getting toward the end-game of maturity.
Using phased synthesized AC, allows economies in electric motor design. Using efficient devices that generate little waste heat, and are cheaper to construct, have direct benefits when youu need to change DC from a battery, to synthesized AC current for the motors.
Posted by: stas peterson | 10 June 2008 at 10:45 AM
stas:
Do any of the current hybrids, PHEVs and BEVs use 3-phase variable voltage, variable frequency (VVVF) inverters and 3-phase motors?
Those inverters have been available from G.E., Thoshiba, Torsion Tec, Bombadier Mitrac and others for some time (15+ years).
Our recent Heat Pump (Friedrich) uses VVVF inverters in the exterior and interior units to get much higher system efficiency and much lower noise level due to the continuously variable speed.
Posted by: | 10 June 2008 at 12:18 PM
I had a whole bunch of negative comments to make, but I found that they had already been beaten to death, so I dropped them.
Go Delphi! At least they are announcing that they plan to do something with some people somewhere.
What is that saying about turning slowly but grinding exceedingly fine ... ?
Posted by: NCyder | 10 June 2008 at 12:23 PM
Anonym,
Tesla roadster (BEV) uses 3-phase induction motor and corresponding invertors to power it.
Posted by: MG | 10 June 2008 at 02:45 PM
MG:
Sorry about the missing name.
Thank you for the info on Telsa's 3-phase motors and inverters. Are the VVVF?
Would this R & D help to produce improved VVVF inverters + 3-phase associated motors for Telsa and others?
Posted by: HarveyD | 10 June 2008 at 03:52 PM
MG:
Sorry about the missing name.
Thank you for the info on Telsa's 3-phase motors and inverters. Are they VVVF?
Would this R & D help to produce improved VVVF inverters + 3-phase associated motors for Telsa and others?
Posted by: HarveyD | 10 June 2008 at 03:52 PM
HarveyD,
Those inverters in Tesla roadsters for 3-phase motor must be VVVF, it can be implied from articles on their web site.
That's the modern way of controlling 3-phase motors.
The described Delphi design is about improving efficiency, lowering price and size.
And what looks very important it allows the inverter to work at under-the-hood temps of 105-120°C without need for an extra cooler.
Given that Tesla does not use any ICE that would generate those high temps, one of the key advantages of Delphi inverter won't be neded.
But more efficient and cheaper components are always desirable.
Currently Tesla uses some latest high power modules (IGBT) that allow them to feed their AC motor with stronger curents, so they use their own design.
The latest powerful IGBT modules allowed them to increase motor torque and eliminate two-speed gearbox (that was breaking down after 4-5 kilo miles) and instead use a simple single-speed gearbox.
I don't think that the new Delphi inverter can provide necessary currents for 180 KW AC motor in Tesla roadster.
Probably some future designs will be suitable.
Posted by: MG | 10 June 2008 at 05:31 PM
Sounds like Christmas.
The smarts will be so important to all areas of power management and needs to be spot on so as not to waste energy or be unfriendly other savings.
Sounds like a lot of money but so is 1.3? billion each time a stealth drops.
I hope that Delphi can achieve value for money with this R&D.
For the taxpayers dollars who owns the research?
Are the Insights gained then commercial in confidence? Will the co(s) shelve and trash anything they have no application for? Or share the findings and add the reputation to their Co's bottom line.
I do hope this goes well and adds to the knowledge pool.
the $ bottom line will no doubt follow.
Posted by: arnold | 11 June 2008 at 12:05 AM
MG, the ability to operate with 105-120 degC liquid cooling is NOT the same as the ability to operate without liquid cooling. Your assumption that heat only comes from combustion engines is not valid.
Posted by: doggydogworld | 11 June 2008 at 06:27 AM
doggydogworld,
I agree that I wasn't very precise regarding the liquid cooling, in the case of Tesla roadster there is no need to liquid cool anything, air cooling is sufficient.
And I answered for the case of Tesla car, similar thermal issues are supposed to be encountered on other BEVs.
I didn't assume that heat only comes from combustion engines, just that ICE is significantly greater source of heat than anything on Tesla roadster (or similar BEV).
Actually in the text it is said that they target HEV and PHEV, not BEV market - probable reason is higher env temp caused by ICE.
Of course there are several sources of heat in BEVs - motor, batteries, gearbox, then power transistors from the inverter.
Reportedly Chevy Volt will be using liquid cooling for batteries, and they'll probably keep LiIon batteries below 80 degC if not (much) lower.
Vehicles powered by batteries that operate at 300degC (? Zebra batts) are another issue.
Posted by: MG | 11 June 2008 at 10:37 AM
VVVF inverters have been around for some time now, and are used in applications where efficiency and precise control are needed. The difference here is the use of silicon carbide devices to enable higher efficiency, higher operating temperature, and smaller size.
For power switching, SiC technology is almost as big an advance over silicon technology as silicon was over mechanical relays. It's one of the core driving technologies that will be reshaping the power landscape over the next couple of decades. Among other things, it has the potential to make DC-DC converters as cheap and efficient as AC transformers. HEVs, PHEVs, and BEVs will be big application areas for it, but we'll also see things like homes and offices with DC wiring and battery backup systems.
Posted by: Roger Arnold | 11 June 2008 at 11:17 AM
@Roger Arnold, the last sentence of your post touched a nerve with me. I too wonder why residential and office lighting needs to be at 110vac. This requires licensed electricians and/or equipment like more expensive glass fibre ladders for safety. A 100W lamp takes an amp but most of the house wiring probably could carry 10A without getting warm. A replacement 24v lamp at 4A would also have a more robust filament. New electronic fittings would also be cheaper running from 24vDC. The use of a conventional car battery but with 12 instead of 6 cells would cost not much more than $100 and provide backup power for lighting. You can also work on 24v live. So you get a short and trip a 10A breaker so what ! Room switches also at 24v wouldn't need CSA/UL certification which opens the market to smaller firms. More complex switch functions could be handled by a programmable logic controller in the basement which could receive these switch inputs and dole them out to the relevant 24v lighting circuit. But more than that, a two way switch in the hall or on the landing could be carried out by the PLC instead of special wiring. The two way switches being replaced with simple pushbutton switches instead. Just one daylight sensor input can control any number of fittings, PLC's have real time clocks so timing functions are also available to any fitting. The programming possibilities are endless.... And now a word for our sponsors : -
Only $5m to Delphi ? Heck, in January 1999 the DOE awarded Satcon Technology Corp. $10m to develop power modules that would manage electrical power in vehicles. (In their words) The company was to cut the cost of management systems from $10,000 to $500. Prototypes should have been ready for the Big 3 within 15 months. Satcon (Cambridge Mass,) belonged to the PNGV consortium. Anyone at Delphi benefitting from that PNGV contract today ????
Well that aside, good for Delphi. I hope Ford gets their handout too this time.
Chrysler not so much, since they are a private corporation in the hands of the financial gurus of Cerberus who could sure find their own funding if they felt the cause worthwhile. Of course externalising has become very popular with the mining and energy sector.....
Anyone know if Tesla has received any of the DOE's largesse ?
Generally smaller companies are prepared to accomplish more with less. Larger companies tend not to want to do anything they haven't done already. Wasn't GM playing with the Electrovair II back in 1967 ? But only now may they be pulling all nighters to get the VOLT ready by 2010. Sheesh.
Look, I could write about how the Chevy Volt will be more demanding on its traction inverter pulling 400A at nearly half the voltage of the Prius while the Prius only has to touch 250A for a few brief seconds but I know some here would berate me for wasting good GM bashing time.
T2
Posted by: T2 | 11 June 2008 at 01:51 PM
in the case of Tesla roadster there is no need to liquid cool anything, air cooling is sufficient.
MG the Tesla battery pack is liquid-cooled. I thought they also use liquid cooling for the inverter, but if anyone knows for sure please speak up. BTW, about five months ago Tesla said they would switch to a liquid-cooled motor for single-speed gearbox vehicles, but they've since engineered their way around that and the motor will remain air-cooled.
Posted by: | 12 June 2008 at 08:41 AM
There are prismatic LiIOn batts from EnerDel that work at 33 degC, and need just air cooling. They will be used in 'Think' BEV.
At Tesla they kept the motor air-cooled, and say they improved it to handle higher currents. But didn't say if both stator and rotor were modified, or just stator.
Now compare that design with supercar Bugatti Veyron with ICE which uses TEN heat exchangers !!!
(Veyron goes 0-100 kmph in 2.5 sec, Tesla in 4 sec - like some Porsches and Ferraris).
Posted by: MG | 12 June 2008 at 03:27 PM