All set for inspection

Shield were made to prevent accidental touching of battery terminals and powered connectors. They were made from clear acryl sheets that were painted black on the bottom side. The shields also block some of the spashing water from reaching the battery terminals.

Also the motor controller assembly got an enclosure which was modified so that it can be opened by a single butterfly nut.

As everything works now I think its time to have the Corolla inspected to make it street legal as an electric car!

Updated main circuit diagram

Quite many changes were done to the initial circuit diagram along the way. The changes were governed by the idea of not needing to alter the original wiring. The circuit diagram shown above shows the current state of the Corolla eFX electrical connections. I am currently happy with it as everything seems to work well.

The biggest change to the original circuit diagram are the six parallel connected power MOSFETs that are located between the 12V battery negative terminal and the chassis. They were added to boost the discharge current capability of the PCM circuit from 20A up to about 100A. The gates of these MOSFETs are controlled by the same gate drive that controls the 12V battery pack PCM discharge MOSFETs. This way when the PCM notices an undervoltage situation the MOSFETs are turned off  protecting the lifepo4 cells from potential damage.

The other major change was the addition of the 72V to 12V isolated power supply. This change made it possible to use a common 12V relay to wake up the motor controller. It also provided an isolated 12V power supply to be used by the panel meters. The 72V to 12V isolated power supply chosen is actually a low cost plug type offline AC/DC power supply that needs only about 50Vdc to start up. Not all of them do, but many of them do start with a lot lower voltages than the typically rated voltage range of 96Vac...250Vac. Of course, if used with a lower than specified input voltage the output current can not be expected to reach rated values.

Electric brake booster mounted

The electric vacuum pump and the vacuum operated microswitch were mounted close to the vacuum brake booster. The pump was mounted using a rubbery tube clamp to damp some of the pump's vibrations. This electric vacuum pump is quite a shaker when it is on.

12V PCM tested

After successful tests with the 72V battery pack's PCM branded as Signalab, I was expecting similar behaviour from the 12V battery pack PCM. I do not know the manufacturer of this board which is shown in the picture, but it is still available in eBay.

When I first connected the batteries to the 12V PCM I had my multimeter connected to measure the resistance between the P- and B- terminals. I was expecting the resistance to drop close to zero when the last cell was connected. To my disappointment the resistance dropped before the last cell was connected to the B+ terminated. My first assumption was that the board was faulty. A new PCM board would take weeks to arrive so I though that I would examine this one a bit more before giving up with it.

I disconnected the PCM and took it to my lab room to examine it with a laboratory power supply. I found out that the PCM worked as it should with the cell over- and under voltages, but what was left undocumented is that if any cell voltage would be less than about 0,7V the under voltage situation is no longer detected and the P- terminal is enabled again (if the other cell's do have a valid voltage between 2.0 to 3.85V). This happened to me during the first test of the PCM when the last cell was disconnected and still the P- was enabled.

Now that I know how this circuit works I decided to install it anyway although the circuit can not be expected to disconnect the load if for example a wire gets disconnected between the PCM and the battery pack.

Placing the battery chargers

My original plan was to place the battery chargers under the hood, next to the batteries. It turned out that waterproof fanless chargers were very highly priced and thus did not fit in this low cost project.

The chargers that were chosen for the project are fan cooled and ment to be used indoors so they had to be mounted in the cabin. The most suitable place for them was found from the front passengers feet space.

Mounting the electric heater

The heater assembly was finished by adding a second aluminum plate on top of the power resistors using a clamp made from steel. The five 2.7ohm resistors were connected in parallel using 4mm^2 wire. After testing that each resistor and temperature switch worked the heater assembly was painted using black exhaust paint that withstands temperatures up to 650C.

The heater assembly was then mounted on top of the air vents that blow air to the windshield. The heater assembly is only 30mm high so it does not block visibility much. After it was painted black it fits quite well into corollas interiors. I hope there is now enough heat available to keep the windshield clear of fog.

12V system testing

The 12V 60Ah LiFePO4 battery pack was mounted on its place with a simple sheet metal frame and a belt. I wanted to test if all the original electronics still work after the original motor was removed. It was great to notice that all lights worked and only the cabin fan did not work. After some research I found out that its power wire was in the motors cable harness and got the fan also working just by connecting it.

Current consumption was measured for the different utilities and the main consumers were found to be the cabin fan taking 16A on full power while the driving lights took 14A.  It's good to know these so I can rate the new fuses and relays correctly. Some changes have already been made on the circuit diagram that was posted earlier. I'll post its updates along the way.

The Corolla eFX is now in a driveable condition with its lights fully functional. The following steps will be wiring the 12V battery pack protection electronics, finishing the heater and mounting the electric brake booster and chargers. After that I'll show the car to the local authorities and hopefully get the car street legal!

Controller programming for full power and release throttle regeneration

The traction battery is now ready to be driven with full power so the motor controller needs to be programmed to give its maximum motor and battery current. I also wanted to try the "release throttle regen" mentioned in Kelly Controller's PM72401B features. I had to contact their support in order to get the release thottle regen to work as I could not find the parameters for it in the configuration program. I case some of you have the same problem "release throttle regen" can be activated by setting the desired regeneration value to the parameter described as "Tps Mode And Max Allowed Regen Current[3]" shown in the picture above. I do not know what "Tps" means, but setting a value to this parameter certainly started regeneration when the throttle was released.

The release throttle generation was set to an amount of 30% which in practice felt a bit too powerful for easy gear switching and too weak to actually stop the car by regenerating. Also there seems not to be any ramp when regeneration is started resulting in a fast torque change which can be felt as a distracting thump in such a light car as the Corolla.

I am thinking of testing a combination of perhaps 10% release throttle regeneration and maybe 40% of brake switch regeneration activated by a microswitch that would be installed on the brake pedal. I am happy over the ease of Kelly Controller's programming so testing different alternatives is only a matter of time and ideas.

First smoke - CBM overheating

 

 

It was a good thing to follow closely through the first charging of the traction battery pack. When the first cell voltages reached the balancing threshold I started to be worried of the CBMs possible overheating as I had only tested the CBMs in free air in a vertical orientation with a very small charging current. I consentrated for a while into other things and when I checked the charging situation again about 20 of the 24 CBMs were active and the whole enclosure was clearly too hot to survive. I stopped the charging but it was too late. The heat shrink tube used as insulation between the CBMs melted resulting in a short between two CBMs. Smoke started to rise and I started to remove the fuses to disconnect the batteries from the smoking CBMs. Finally all 24 fuses were removed and no flames were seen. The molten heat shrink tubing and smoked CBMs are shown in the first picture. This was a good reminder for me  about how easy it is to underrate cooling.

Luckily the PCM board was not harmed. As seen in the picture the CBMs were removed from the enclosure dropping the balancing current to just 60 mA provided by the PCM board only. While charging the battery pack with 10A the charging was now stopped into a cell over voltage as one of the cells reached a voltage of 3,9V. Some other cells were clearly not yet full so I made a manual balancing for the battery pack by sinking current from the fully charged cells and charging the ones with lower state of charge. After a while of manual balancing I got the cells balanced enough to make the charger to stop charging normally; the charging voltage was about 87,6V and the charging current was reduced into about 1A.

Now the first traction battery charging cycle is done. I am curious to see if 60mA top balancing current is enough to keep the cells balanced. Anyway I am so far very pleased with the PCM board!

Traction battery charger close up

The charger was purchased from www.kellycontroller.com. It is an engineered product made for customers specs. I ordered this charger with the following specs that were needed for the order:

Mains voltage: 230vac
Output voltage: 87,6V (24 cells x 3,65V)
Battery capacity: 60Ah

This charger is ment for indoor use only so I cannot place it under the hood where it would be exposed to moisture and dirt. Therefore I decided to mount it in corolla's trunk.

The chargers manual does not say anything about its leakage current (current drawn from the battery while the charger is unpowered) so I measured it to be 1,8mA with 80v battery voltage. This is lower than what I expected so its well suited for onboard use.

The chargers manual does also not mention the chargers power factor which is claimed to be 0,96 according to www.kellycontroller.com. I will measure its power factor to find out if it really has an active PFC or not. Anyway I'm a bit dissapointed with the chargers documentation.

First time traction battery charging

Before connecting the motor controller to the traction battery pack I wanted to test the connections made so far as well as the PCM board and the charger bought for the 72V battery pack.

I first inserted the mini blade fuses in their fuse holders at the battery terminals. The fuses were inserted starting from the cell closest to the pack's negative terminal and then proceeding to the next cell connected to it. This connecting procedure was adwised by some PCM's and I decided to follow it as the PCM I received did not come with any manuals.

Before inserting the last fuse I checked that the mosfets in the PCM board were turned off. Once the last fuse was inserted the PCM's mosfets were turned on which is correct as all cells had a voltage of about 3,3V.

The charger was plugged in and a charging current of 10A was measured. I was beforehand a bit worried about the PCM's mosfets heating, but the heat dissipation with 10A charging current was measured to be just 0,5W divided by two TO-220 packaged mosfets so they only warmed up barely noticeable. I do not know which mosfets have been installed on the PCM as all of their markings have been sanded off for some reason.

I did not make a full charge as I want to be around with my multimeter learning about how the top balancing behaves and how the CBMs and everything else heats up. So I'll continue charging while working in the garage ready to stop charging if something fails.

New meters behind Corolla's steering wheel

Three led panel meters were installed to be easily seen by the driver. The panel meters were earlier introduced in the post titled "Configuring panel meters".

The enclosure for the meters was made from a piece of white plastic 30x30mm cable trunking which was covered with black duct tape to better match with Corolla's dashboard.

These meters will be giving readings of the traction battery voltage and current as well as motor current. Only the speed meter will be left in use from the original meters, although rpm and motor temperature metering might also be nice in the future.

Battery pack connected to BMS

The next layer of wires connect the traction battery pack to its battery management circuits. The connections were made using 1,5mm^2 wire with a mini blade fuse holder as close to each cell terminal as possible. The mini blade fuse holders were soldered on each of the wires.

Traction battery power wiring

The 72V battery pack was wired keeping in mind that the wiring should be able to provide up to 400A for short periods. Most of the connections were done using power terminal connectors purchased togheter with the batteries from www.ev-power.eu. As the batteries were not assembled side to side, most of the power terminal connectors were too short and their holes had to be slightly modified to make the connections possible.

A 400A 75mV shunt resistor was mounted on the negative terminal of the battery pack to allow battery current monitoring. A 400A ANL type fuse was mounted as close to the positive terminal as possible. The cables used have a cross section of 50mm^2.

The battery management electronics will be wired next followed by the first charging of the traction battery pack.

Mounting electronic enclosures

A mounting stand was made for the enclosures including motor controller, battery management system and relays. The stand was mounted between the original motor mounting point shown on the left and the gearbox frame.

The stand was made using suitable steel rods that were welded together. It was then painted with zinc spray and a thick layer of black steel paint.

The wiring for the 72V battery pack can now be finished which will allow a first test drive with decent batteries! The wiring of the 12V system will have to wait until its battery pack and battery management system enclosure have been fixed on their places properly.

Wiring the battery management electronics

Both 12V and 72V lifepo4 battery packs will be managed by cell balancing modules (CBM) and a protection circuit module (PCM). Both are easily available and simple to connect. One CBM will be connected parallel with each cell and the PCM needs a connection between every cell. Although these connections are simple, the amount of them make this step quite demanding.

A half way wired 72V battery management circuity is shown in the picture. The PCM is shown on top and below it lies 24 CBMs. Once this is fully wired it will be mounted in a 240x190x90 mm enclosure. The connections to the battery pack will be done using 1,5mm^2 wire with a fuse located close to the terminals of the battery pack.

72V battery pack mounted

The finished 72V battery pack base plate was painted first with zink- paint and then with black paint to slow down corrosion. It was then mounted on its place where the radiator used to be. Next the battery pack was mounted on it cell by cell. The easy to reach location of the 72V battery pack will be handy while making the electrical connections on the cell's terminals.

Mounting the cells on the base plate

All of the cells of the 72V battery pack found their place on the base plate. The cells are sitting tightly on their places as the cells do nothave room to move in any direction on the base plate. The cells are tied togheter using bent washers that are shown in the close up photo below the wing nut and the rubber washer. When the wing nut is tightened, the bent washer presses both cells towards the base plate and also makes sure that the cells can not escape sideways.

Battery base plate under construction

The base plate for the 72V battery pack is being built from different steel materials that I had available. Most of the base plate is done using 25 x 25 mm steel square tube that is cut and welded. A tight place is made for each of the cells and the cells are fixed on their placed by pressing them against the base plate with a threaded rod and a wing nut.

The cells are slightly separated from each other to allow a better airflow which improves cooling. While the maximum continuous current of 160A is drawn from the battery pack every cell dissipates approximately 20W of power according to my interpretion of the datas given by the cell manufacturer Winston. If the cells would be tightly packed next to each other I would be worried that the ones in the middle of the pack would overheat. After getting the Corolla on the road it will be necessary to learn how the batteries, motor and the motor controller heat up so that any unpleasant suprises due to overheating could be avoided.

Heater assembly

A simple heater system was built from 5pcs of 2.7ohm 100W aluminum case resistors and 5pcs of temperature switches, bought from China through ebay. The resistors and temperature switches were mounted on a 1m long aluminum plate that will be mounted on top of the air vent that blows air to the windshield. The idea is to have just enough heat to keep fog out of the windshield.

Temperature switches were mounted next to each of the 2.7ohm resistors. Their job is to cut the power off from the resistor next to it if it reaches the temperature switches temperature limit. The temperature switches that were ordered have a limit of 55 celsius.

The heater will be powered from the 13,2V 60Ah LiFePo4 battery pack and all of the 2.7ohm resistors are connected parallel. This gives an initial heating power of about 320W.
If the heating power of the heater turns out to be too low I'll need to either improve the heat sink, install more resistors, or install temperature switches with a higher temperature limit. Practise will tell what will need to be changed.

In addition to the windshield heater, seat heaters will be placed on front seats. Anyway while travelling in this car during cold days, some extra clothing will definitely be needed as only a fraction of the original Corolla's heating power will be available.