Saturday, August 21, 2010

This is my third conversion of a conventional vehicle to a pure electric.

http://www.metricmind.com/audi/main.htm
April 06 2008

Beginning of the project.

Audi A6 Avant Quattro This is my third conversion of a conventional vehicle to a pure electric. Back in 1995 I have done my first simple and typical conversion of 1991 Honda CRX HF. Outcome was good and I learned a lot building and driving it. Later on in 2001 I remodeled this car so extensively that I consider it another, more advanced conversion - every single electric and electronic component was replaced. Only the donor car stayed the same - see more advanced conversion. I installed AC drive system, tried several battery types (Lead Acid, Lithium Ion and Nickel Metal Hydride), ultracapacitors and few other novel at the time solutions. That vehicle has become known as "Honda ACRX" and it still serves me well. Back then first I built what I could to get my feet wet (DC system), then I built best I could with latest available components (AC system). I had to gain more experience in more advanced EV technology, and "EV bug" still bugs me - now and it's time to built what I really ultimately want. The idea of what this vehicle should be like kept materializing several months and finally in February of 2008 action started. I should tell reader that what is being described here is no typical conversion, nor it is by any means a guidance or set of instructions how it is suppose to be done. Choosing what to build and how to execute it is very individual choice considering skill level, availability of components, confidence, experience, taste, ultimate conversion goal, expected driving pattern, ambition level, stars alignment... and many subtle reasons that are hard to describe. Selecting a car to drive is somewhat similar to selecting cloth to wear - there is no right or wrong choice. There is better or worse one (optimal or lousy) as far as efficiency of converted vehicle and difficulty of execution, but anything on wheels can be converted with more or less degree of success. You may or may not be happy with the outcome and performance at the end, but that's just means miscalculation in a first place. Actually, the donor vehicle selection is relative simple process, considering few basic rules.

Rule 1. Pick what you'd love to drive and to be seen in. Any vehicle you choose will move, just some cars are better candidates than others - the lighter, roomier, more common they are, the easier process and better outcome. Unless you observe rule 1, you're going to start to hate your car before you even finish conversion. Therefore DON'T choose old clunker just because its engine has 400,000 miles on it and is going South. You will spend far more effort restoring the body, convert it and at the end will have electric version of still the same old clunker you'd want to get rid of. Trust me.

Rule 2. Pick based on planned utilization. What is the purpose of the vehicle?. Small, slick and efficient commuter, freeway flyer, roadster, drag racer, trusty work horse/truck or anything in between. Regardless, the lighter vehicle, the more room inside, the more weight it can handle and better aerodynamically shaped it is, the better.

Rule 3. If after filtering through rules 1 and 2 you still have few choices, pick the car in better shape to start with, straight, with more parts available for it. Avoid restoring old rusty car just because you already have it "for free" (Imagine, you've done conversion of it tomorrow. How long it will be before you'll get tired of this car?) .Also, depending on how comfortable you are with modern car electronics, pick more or less late model vehicle. The later model you choose, the more challenging it will be to integrate into its electrical subsystems to keep them working, since cars become more and more electronics intensive and like it or not this trend *will* continue. Newer vehicles designs tend to be naturally improved over old ones. However, older vehicles are simpler and not so challenging to work on. Very old ones (pre-80's) will have rudimentary or no ECU (engine control unit) and are wired pretty much as any appliance - point to point. For instance, battery - to the switch on the dash and switch - to the tail lights. Flip the switch (close the circuit) - tail lights turn on as well. Current flows from the battery through a fuse and switch to the light bulbs and returns to the battery through car's body. Simple. Today, switch just sends logic signal to the ECU or electronic control block. That block has some intelligence. In case of lights, it first analyzes condition of its other inputs and then sends signal over bus (or drives directly) to the tail light local receiver. That receiver actually turns it on and off. Why so much "complexity"? Because it saves material, weight and cost while adding functionality. In this example If you have rear brake light, parking light side marker light and turn light clusters on both sides of the car, it's 8 light bulbs. All lights lit in pairs except turn lights you must control independently. Common ground for them all is usually the chassy, so in old days you'd need a bundle of at least 5 copper wires from the front of the vehicle toward rear. At about 5m long each, it's 25m of wire. The gauge depends on the allowed voltage drop and typically is about 2mm2 (~AWG14). These days, with a light controller sending signal to a cluster wiring, it's only a bus of two 0.8mm2 (AWG18) twisted wires, 10m long total. Local driver decodes incoming signals and drives actual lights which have short connection to the 12V bus routed throughout the vehicle in any case. So you just saved 15m of relatively heavy copper wire per vehicle, just for the tail lights. Driver chips weigh very little; produced in millions cost pennies, and the software cost is spread over many vehicles (and weigh nothing). Add other rear side electronics like fuel filler hatch monitoring sensor, backing up sensor, trunk light, rear dome light or climate controls (if equipped) and you can see that each of these requires own connection to the dash or controls in front inputs and outputs. More wiring. With centralized controller, you still use the same 2 wire bus to send and receives digital signals to and from those blocks, meaning so much more copper savings for the manufacturer. Vehicles are not made for easy understanding, common people don't have to know how it works (though it pays to know). Like TVs, and other consumer electronics most of the complexity is in the software and no one these days repair their own TVs. Same with cars (unless problem is obvious mechanical one). Well, you get the idea.

Back to the vehicle selection. In my case I wanted comfortable full size car. I want to take advantage of its room and utility. I happen to like station wagons, and another advantage of those compared to sedans is better weight handling (usually stiffer rear suspension). I don't plan to race on streets, but plan to accelerate quicker than most cars out there, definitely quicker than whatever stock car I get. I want to be able to demonstrate the vehicle, put 3-4 passengers in it and drive without any performance limitations. Finally, I wanted a touch or high tech luxury without being exotic, I'd like to have reasonable comforts offered by modern vehicles and of course safe car. It certainly had to be European vehicle, most likely German make. First, they use metric parts and I only use metric tools. I don't do any engineering in other units for many reasons. Second, I happen to have lots of respect for German vehicle engineering. Because this time I wanted smooth, shiftless, powerful and quiet acceleration from still to the maximum speed, I chose single gear reduction between motor and wheels - only via differentials. People often call such setup "direct drive", but this is technically wrong: direct drive is implemented using either hub motors where motor shaft is linked straight to the wheel or drive motor(s) with its/their shaft(s) extended directly to the wheels so the motor rotation speed equals wheel rotation speed. Any other form of transferring torque (gears, chains, belts, even with 1:1 reduction ratio is not "direct drive" it will be "single speed transmission". To increase stability and torque at the wheels and retain maximum grip without stressing differential, after careful considerations I decided to use 2 motors drive system - one runs front wheels and another - rear wheels through respective differentials, so the donor vehicle had to be 4x4. After few weeks of initial thoughts, consideration and Googling around, the selection was made - Audi A6 Quattro Avant. Lots of room, 4x4, large enough for me (I'm 195cm tall), very handsome looking and there is something else I can't describe about Audi's - they seem to have their own "spirit" no other vehicle (even more technically advanced) has. As they say, you don't sit in it, you wear it. Anyway, perhaps this is not relevant to the reader, just my personal preference. After about 3 month of monitoring local Craig's list, I finally purchased 2001 model with 131k highway miles on it. As long as the body was intact, suspension is not abused and all internal guts work, this is all I need. Knowing ahead of time what kind of drive system I will use, it made no difference to me if the vehicle comes with standard or automatic transmission - either one is going out anyway. Typically you'd retain transmission as convenient place to attach single electric motor to (whether you retain the clutch or not). Low gear selection allows to multiply torque at the wheels - you may take advantage of it. But, remember, in my case it is not "typical conversion". With two 200kW peak systems (543 "electric" hp and potential 860 Nm (633 ft*lb) of combined shafts torque, more on this later), I have plenty of torque for any kind of spirited driving, and I'm not planning to shred rubber. So, because I will direct couple to the front and rear differentials with two mechanically independent drive systems, not only engine but whole stock transmission and drive shaft goes out. That broaden my purchasing choices: automatic or manual would do.

NEXT

No comments: