Signature Renewables | Vancouver | Welcome
Signtaure Renewables — renewable energy — smart-grid — electric vehicles — clean tech
Services Meet John Public Speaking Blog Contact
Signature Renewables

The Conversion

Others have compiled extensive websites documenting the project and process. Aside from reams of technical and circuit documentation and a basic owner's manual I collected and produced for my own reference, this narrative has been written to capture the main issues surrounding the components of this project.

Figure : ICE components removed: engine, fuel tank and fill tube, vapour canister. Not shown, muffler and engine coolant lines assembly.

This was a 5 year process that existed in my head for the first two; then in the final three involved active component selection, research, planning, prototypes, parts and components sourcing, bench tests, logic circuit building and detailed assembly. Breaking down my time as project manager and primary labour source, I estimate my time allocation was about 75% research and learning about the components, planning, sourcing of parts and components, integrating, and pure application of brain power. I spent 10% of the time bench testing subsystems outside the car, and the remaining 15% of the time actually integrating the components into the car. After the main systems had been bench tested and the primary logic box and wiring plan had been laid out, it was time to put the components into the car. I’m a thinker and risk manager, rather than a “get ‘er done’ type, so I spent some time figuring out the layouts, and changing my mind a number of times. However, there are always some things I’d do differently the next time. This was a one off project. The proto-type is the finished product. My engineering budget was at least 1,000 times smaller than Tesla Motors’. But after four months in a local machine shop, and with some skilled and experienced help of the guys who worked there, the main drive parts were machined, fitted, and installed, and I completed the remaining wiring, and auxiliary component installs. It was the end of April 2012 when the car rolled out of the shop on its own power, and with final adjustments and safety inspection, the car was licensed for the road.

Figure : Shop with hoist: Most components on a Porsche are installed from the underneath.

The final component selection and design came down to a number of key objectives and constraints. To be useful the car had to have a reasonable range. It had to be fully usable as a daily driver, and daily commuter. As a first time conversion project manager, I wanted to minimize risks. Therefore don’t redesign the stock transaxle – but fit the motor into the space I had. The Boxster rear engine compartment fits the stock engine like a glove. Fitting a motor in the centre of this space effectively ruled out putting requisite amounts of batteries in the dimensionally compromised balance. The design decision was to put twin motors in along the bottom of the compartment, connected by a torque combiner to the previous crankshaft from the engine. A sprocket gearing ratio of approximately 2:1 allowed the motor rpm corresponding to peak Hp and Torque to match the original engine rpm to shift through each of the 5 gears in the stock transaxle. A check with a trusted power electronics engineer gave me enough confidence to proceed. I'd half the internal resistance of the drive circuit, and I'd have more rated power than a single motor, and I'd have the given power and thermal heating distributed among the air cooled motors for better thermal control. In theory. There is an issue with combining two motors torque together. When mechanically linked one motor will always lead the other. Under regen however, each generates exactly the same power. This is not the ideal solution (that being one motor) but it solved a big problem of spacial allocation - a very limited resource. However some software tweaking to one of the controller settings has solved this problem under peak load conditions.

To maximize on board energy and power, thirty six (36) lithium iron phosphate cells were distributed between three compartments: Forward trunk where the spare tire normally is mounted (7); the fuel tank void forward of the cockpit cabin (6), and above the torque combiner/dual motor/dual controller assembly (23). They work fine through the warmer months of our northwest coast climate, but next winter I’ll put some thermal insulation around the forward two packs. A commercially produced car would have full water jacket thermal equalization.

I spent time on a logic circuit to ensure certain things could not happen at the same time. The Porsche key switch ignition circuit was fully utilized to start the controllers, running this circuitry through a series of relays and circuit switches to ensure that the motors cannot be activated if the car is plugged in, or if security is turned on, or if there is an adverse situation, such as a low limit fault from the battery management system, or worse, if the inertia switch has been tripped. I filled up the stock console switch positions of the Porsche and supplied the auxiliary latching relays to make them work for additional components: A “wimp” switch to turn on power steering if I need it for tight parking spots; A 4kW water heater for the cockpit cabin heat; a spare switch awaiting a use. When the ignition switch is off, no matter what the status of any of these switches are, the only draw on the main 12V system is the normal Porsche draw, after courtesy lights have turned off – just the security system. The only draw on the main pack is the DC-DC converter maintaining 13.61VDC on the 12V system, unless the main contactor is off.

Figure : Bench testing. Make it work here, and the install is much easier.

A vacuum pump turns on with the ignition and maintains a generous auxiliary vacuum reservoir for what little conventional braking is normally required in normal driving. The twin motor controllers have both an active and passive regenerative braking modes. As a sports car, it was my intention to make the car unique, interesting, and fun to drive. From my first sports car I remember the thrill of down shifting and hearing the engine drag against the forward momentum while coming to a stop. 35 years later, electric technology allows for the same thrill, but this time, it’s quieter, and the thrill is being able to recover that spent energy back into the battery pack whenever going downhill or having to stop at incessant traffic lights. Fingertip controlled active regen braking is achieved from a gear shift mounted single axis joystick – up to 150A or almost 1C of restorative power. Passive braking (when the accelerator is released) can supply up to 10A, typically 6.

A water pump sourced from a local computer supply shop, with its micro-processor controlled operation provides efficient and reliable cooling to the dual, back-to-back controllers through a single shared water jacket. A bi-metal switch at 35°C ensures the controller heat sinks never see more than 40°C and a secondary bi-metal switch at 50°C warns of an overheating situation (never seen). Repurposing one of the two stock Porsche radiators, and the expansion tank in the rear trunk complete the cooling circuit.

Figure : Test fits allowed for minor adjustments to be made and the installation procedure to be honed. The final install went almost as fast as the test.

In the design of the charging system I went back and forth where to install this component. On a cold December night in 2011 while unplugging a friend’s EV conversion it all became logically clear. Installing the charger in one of the trunks would cause excessive heat build-up from the air cooled system. By installing it behind the passenger seat (fits perfectly, without restricting seat movement) provided an in cabin heater to keep the dew off and pre-heat for those trips in cold weather. It was a natural and efficient placement. A few other serendipitous alignments also happened. The hydraulic power steering pump mounted perfectly within the fuel tank void in front of the steering rack. The vacuum pump also fit perfectly to the right of this unit, out of the way of the space occupied by the fuel void battery pack, and buried in the car to minimize its relatively quiet vibration. There were a few other things that worked out well on the conversion, bolstering my confidence of a successful outcome during that unknowing period leading up to the time you can actually drive the car.

To connect with the grid, SAE J1772 had been selected as the EV standard while I was building my Boxster EV, however there were no public charging stations of any number at the time of the build. I went with a readily available NEMA inlet port, but did confirm that the dimensions of my port were compatible with a J1772 inlet for future retrofit. Now, in the spring of 2013, just one year since my car hit the road, there are approximately 600 public charge stations in our Province, and I've made the interim step of purchasing a J1772 adaptor so I can now utilize any of these stations.

Figure : Charging port retrofitted into the stock gas inlet port. Includes status lights from BRUSA charger and a switch for slow charge. The dimensions of this NEMA L14-30 inlet will also support retrofit of a SAE J1772 inlet.

The curb weight prior to conversion was 2,826 pounds with a weight distribution 60/40 rear/forward. The conversion added 301 pounds, and shifted the weight distribution forward to 52/48.

Figure : Final weight and distribution - 300 pounds heavier and weight shift forward to 52%(R) and 48%(F).

Untitled Document