interiorcabin heaterFront cell boxController mounted on heat sinkmid boxunder hood
OwnerTom
LocationReno, Nevada US map
Vehicle2001 Suzuki Swift
3 door hatchback
Motor AC50 3-Phase AC
HPEVS AC50 motor, about 64HP peak at about
3900 rpm with 115V pack
DrivetrainOverall gear ratios (including rear end):
13.45, 7.16, 485, 3.45, 2.88
ControllerCurtis 1238-7501
max current 550A, max voltage 130V
Batteries36 CALB/Skyenergy 180Ah, 3.20 Volt, Lithium-Ion
36 LiFePO4 cells connected in series,
nominal voltage 3.2V per cell.
System Voltage115 Volts
ChargerManzanita Micro PFC30
Max output about 38A DC at 120VDC.
Typical max in summer without thermal
cutback is about 27-28A. Can do higher
in winter. Takes a bit over 4 hours to
charge if discharged 70%. Typically
charge at 7 to 10A overnight.
HeaterFarnam heater pads and controller (kta-
ev), 10 total. Work very well, kept
cells at the 59F (15C) setpoint on two
successive nights at -5F (-20.5C). Slow to initially heat though, about 3F/hour increase in temp. Cells in steel boxes with 1/2" polyurethane insulation. This has a very significant impact on range and voltage sag in a cold climate in
winter. I see about a 20%-25% difference in range winter to summer with the heaters set at 60 F (16C) due to stiffer drive train and cabin heater use. In winter the cells typically remain at 50 to 70F, with ambient of around 20 F (-7C).
At 80 F (27 C) ambient, they get up to
90-95 F (~34C) and have lower internal
resistance and more capacity when
charged at this temperature. I had to
tweak the voltage limit on the charger
to get a full charge in summer. If the
cells are unheated in winter, range
will suffer significantly. The heaters
are only on when the car is parked and
plugged in. They will remain above 50
F for around 4 hours at 20 F ambient,
when parked outside unplugged after a
10 mile drive.

Cabin heater is two 1500W ceramic
heaters mounted side by side in place
of the original heater core. Gives
about 3160W with blower on high.
DC/DC Converter TDC-120V/12V
Hangzhou Tiecheng Information Technology Co., Ltd 400W isolated. Purchased from Cloud Electric in 2009.
InstrumentationCurtis 840 gauge, displays battery
voltage and (estimated)current, motor
and controller temperatures, motor rpm.
TBS E-xpert Pro gauge, displays battery voltage and current, SOC, Ah used, has low SOC, low voltage, high voltage alarms. Great gauge! Works just like a fuel gauge. I ran the car with only this, no bms, for about 3 months. The cells remained balanced to within less than 4mV of each other typically, and I only discharge to no less than 30% SOC.
I added a minibms after this, tied to
Regbus of the charger, relays tied to
buzzer, Check Engine light, and to
resistive divider to cut throttle 50%
for LVC event.
Top Speed90 MPH (144 KPH)
Calculated. Has pep in 4th gear at 75mph,
so I think it will do 90.
Acceleration0-60 in 16 seconds, 1st through 3rd gear,
measured with gps data logger.
Range75 Miles (120 Kilometers)
Generally get about 75 miles at 50%/50%
highway/secondary road driving, 60mph
highway/35 mph secondary, 70% DoD in
summer. Winter range more like 55-60 miles due to stiffer drive train and heater usage.
Watt Hours/Mile180 Wh/Mile
Average from the wall socket, ncluding
all losses, over the first two years of operation was 216 Wh/mile. Measured
with EKM meter. Typically use around
180Wh/mi at around 35 mph, and around
250Wh/mi at 60 mph, both on fairly
level/rolling terrain in summer.
EV Miles
Start:83,753 Miles (134,758 Kilometers)
Current:107,106 Miles (172,333 Kilometers)
Total:23,353 Miles (37,574 Kilometers)
 
    As of 2/5/2013
Seating Capacity2 adults
Curb Weight2,250 Pounds (1,022 Kilograms)
Everything stock, including all suspension(only about 350 lb over original curb weight)
TiresMichelin radials
Conversion TimeAbout 4 months (after about 1 yr of study, calculations, and searches on parts)
Conversion Cost$2700 for donor, about $20k for rest
Additional Features1/8" poly from front bumper to just before front axle to
protect motor from water and gravel. Bilge blower (150
cfm) to ventilate mid and rear cell boxes. Axial fan (250 cfm) mounted on 10"x11" heat sink with 2.5"
fins to cool controller - necessary to drive at highway speeds continuously in summer.
Controller now remaining below 45 C typically in 80
F (27C) ambient.
Update 6/10: cells still not going above 90 F yet in 80 F weather and no blower. Added Dimitri's minibms, sans shunts, several months ago and it works well (update: added shunts a few months later). But now that I have the charger voltage limit tweaked in it
stops charging at about 3.44V/cell every time, so the
minibms is just there as backup. I set it this way so
cells are charged to the start of the exponential rise
part of the V versus Ah or time curve. All 32 original
cells remain balanced typically to within 2 mV of each
other after charge. Replaced 4 due to over-discharge by
a different bms, and have balanced (individually charged)
these twice in about 5 months after they became
unbalanced by about 0.01 V from the others. Update: these are staying balanced now.
Update: changed charger limit voltage to 123V and timer setting to 15'. Most cells start shunting during the 15' period, giving small amount of balancing each full charge (sometimes do only partial charges). Pack V during last few minutes of charge (charger on, about 2A charge current) is about 125V.
More details at "SwiftE" thread under Builds and
Conversions forum on the diyelectricar site. Also in the
"garage" at that site. An Excel spreadsheet (referred to
as "ev calculator") can be downloaded at
electricnevada.org. It described this car's performance
very well after modification for 115V pack (gives 90lb-ft torque out to 3900 rpm), from the test data for 96V pack in the standard spreadsheet, and addition of equivalent mass for rotating parts. Largest error about 5% for predicted range and current draw at various speeds, and Wh/mile for summer operation.

Regen was set up lower than I liked on the system as
shipped. Had to use mechanical brakes more than I wanted
on hills and had max of 15A to 20A regen. Purchased a
the Curtis PC programming sofware and changed this so now get up to 140A regen (limited by brake map parameters) and don't have to use mechanical brakes on most hills.
Control amount of electric braking by easing off
accelerator pedal, and if necessary by shifting (same as
"engine braking"). Control of vehicle speed is mostly by
the right foot. Can stop for a light using only electric braking most of the time, by easing off the accelerator more and more as I approach the light. A very reliable car. Just get in it and go. Plug in when done with it for the day, and go in the house.

ONE YEAR UPDATE:
Total miles: 7471
Total kWh used from wall outlet (EKM meter): 1615
Average Wh/mile: 216 (from wall, per EKM)
Maximum distance driven between charges: 75.1 miles, much
at lower speeds so only 64% DoD.

Roughly half of miles were driven at 50-60 mph, half at
30-40 mph. Maybe 1/4 of trips with bike on back and
mostly highway driving (higher drag).

Recently changed motor max speed to 8000 rpm from factory
setting of 6500 (HPEVS said no problem).
Now can go to about 35 mph in first gear and 70 mph in
second gear. Two results from this: (1)Now do almost all driving, including on highway at 60-65 mph in second
gear, only use first for fast acceleration, and (2)better acceleration since now can accelerate to over 30 mph in first gear putting me well up in the power band for second gear, get stronger pulling all the way to 60 mph as a result. Have not redone 0 to 60 time yet.
Update: After input from others, I decided it is likely
not good to run the transmission at such high rpm for
sustained periods of time, so now only drive in 2nd gear
up to about 50 mph (< 6000 rpm). Cells still behaving
the same, no problems with them.

Update 4/30/11: 10,053 miles and 1 yr 5 months in
service. See comment on Wh/mile in the Watt Hours/mile
section above. Added shunts to the minibms boards and
top balanced pack a few months ago. Working very well.
Got about 70 mile range in winter, almost same as summer, when cabin heater not used (cells heated to 65F when garaged). About 60 mile range if cabin heater is on much of the time (two 1.5kW ceramic heaters in parallel).

TWO YEAR UPDATE:
12/5/11: 2 years operating as an ev now. 14,617 miles,
3159.81kWh energy used from the wall (EKM reading),
216Wh/mile from the wall. No issues over the past year,
no down time.

THREE YEAR UPDATE:
Total EV miles: 22,456
Total kWh from wall (EKM meter): 4861.89
Wh/mile: 216.5, from the wall, around 200 Wh/mile excluding charger losses, about the same it has been each year.

Only one issue this year. A battery heater failed open circuit this fall. It was the one under the cell group that had the heater controller thermocouple on it, so the rest got heated to around 100 F before I discovered it next morning. No damage done.

Decided to check and redo all heaters. Made aluminum pans that fit tightly into the boxes against the insulation, with heaters taped to the bottom of the pans under each cell group with aluminum tape. bolted aluminum angle to the top side of the pans to help hold the cell groups in position. Strapped cells down tight, nothing seems to be moving over the last few weeks. Heaters should be ok now without the cell ribs scrubbing on them.

Also added 86F N.C. thermal switches, one in each battery box, mounted to a cell terminal on a tab. Wired AC supply to heater controller through them. There can be up to 21 F difference in open/close, and operating temperature is only accurate to +/-5F according to documentation. At least one seems to be opening at mid to high 70's and it appears all are not closed until around low 70's F, which is fine since I am using a heater controller setpoint of 65 F now. When I come home from a longer drive and plug in the heater power there is no power to the heater controller until later that night when the cells have cooled. Next morning they are all around 65 F. Should be no more overheating if the tc fails or the heater fails again.

No noticeable change in range. Only test I've done is the usual drive to around 28-30% SoC, then floor it to draw 550A (3C) and see if the minibms LVC alarms (threshold 2.8V for 4 seconds). Never has so far.

Other than that just drive, charge, drive, charge...

code by jerry