Battery Pack and Related Electrics Version 2.0

I spent the afternoon of my birthday playing around with a new Lithium ion battery pack design for the 914, if I ever return to civilization.

To save money and complexity, I'll keep the existing Curtis 1231C-8601 controller which is rated for a maximum 144V pack and 500 amps.  This is still sufficient for me driving through town.

Based on years of experience at EVTV with the China Aviation Lithium Battery Company (CALB) 180AH CA-series cells, I know how to bottom-balance them, charge them and discharge them safely.  This is in deference to the many YouTube guys building up Tesla-like 18650 packs from dodgy laptop battery packs.  I did the math to determine I need 42 of the CALB cells at a nominal 3.4V each.

I brought up an old SketchUp model I made a couple of years ago and worked out how to place these batteries and other components.  I'll fill up the rear battery box where the gas engine used to be and put the rest in the battery box where the gas tank used to be, with a good amount of space left over.

This means I can remove the 2 single-battery-sized boxes from the trunk and weld some steel plates into place, prime and paint and turn it back into a fully usable trunk.

I also gain all of the capacity of the battery box where the spare tire used to be right in the front.  I created shapes matching the sizes of the rest of the components I need and fitted them all into place.  This means the new TCCH 4KW or 5KW charger I'm looking at and its controller, the new GEVCU controller, the existing relay board and the existing DC-DC converter all fit inside the front box.  The TCCH is air cooled but the box already has an exhaust fan that switches on during charging, so it looks like it will be a perfect setup.

Cleverly I'm moving the 12V auxiliary battery from it's ugly, nasty, breaking-off-prone mounting bracket on the front of the front battery box to inside the gas tank battery box due to some newly available space.  I'll run the 12V wires down the passenger side of the frunk to the DC-DC controller and leave the high voltage battery pack wiring where it runs now down the driver's side.  There's even still a bit of empty space so I'll play with the position of the batteries to maintain side-to-side weight balance and fill the gaps with some stiff foam.

So, here's the layout.  I've also built up a list of to-do items, some must be done concurrently with this major upgrade, while the rest can be done at my leisure.  I'll bling up the installation by replacing the white plastic battery pack lids with transparent lids.

Charging Around Town and a Tesla in the Wild

After the Gathering of Friends car show, I drove leisurely through town and hit the Costco (for groceries) and when I came out, a father and son were standing next to the car taking pictures.  The 914 is an attention-getter and then when people figure out it's electric...

From there I went to the Target parking lot across the street to try their J1772 charging station.  Here we are, in front of their 2 charging spaces.  Unfortunately when I plugged in, the charger gave an error light and the charger didn't fire up.  I suspected a problem in the charger control circuitry or fuses.  Bummer.

They use Clipper Creek chargers.

Here's the label with full details.

 There are also a number of parking spaces reserved for efficient cars, according to the ACEEE rating score, which I had never heard of.  It's too bad you have to pay a subscription fee to see if your car meets the ratings, which I think will do more harm than good for the effort.

Sadly, my VW Beetle TDi which gets 49 mpg doesn't make the list, I think it was due to the particulates of diesel combustion.  I run the car on 99% biodiesel so there is very little pollution but I don't want to have an argument with someone in the parking lot.

From Target, I drove a hundred yards down the road to the Nissan dealership.  I pulled up at the front and asked if it was OK for me to do a test with their charger.  They thanked me for asking, as one day they showed up to find a Tesla Roadster plugged in and the owner nowhere in sight.  The Tesla was in the way of something they needed to move so they weren't really happy.  I pulled the car around the side and plugged in while 3 Nissan folks watched.  The charger tried to start up, then stopped over and over, and a relay in my charging circuit was buzzing on and off very fast.  I pulled out the cord and thanked the Nissan folks.  Sorry for the dark picture.

The following weekend I moved the enclosure for the charging system and rewired it and found a blown fuse.  After I got it all back together, I plugged it into my Schneider J1772 unit and it took the charge like a champ.  I'll have to try again next time I get back down to San Luis Obispo.

To wrap up this post, here's the backside of a Tesla Roadster I was behind.

The license plate is "SOLDVSI" so I'm thinking the owner sold a company or technology that goes by the name VSI and bought the car with the profits!  I wonder if it's VSI Products, which builds products for the cycling and  BMX  markets?

Solar Carport and EV Charging

Yesterday I picked up my building permit for an additional solar array at my house.  My goal is 3-fold:

- Generate the remaining 15% of my house's power needs that I'm currently buying from PG&E
- Generate sufficient power to charge the 914 through a J1772 charging station
- Park the cars under a carport to block the direct sun and scorching summer heat

I bought the electrical parts for my system from my sister company AEE Solar and the carport from Protective Weather Structures in December of last year.  I've got a SnapNRack mounting system, 16 REC Group 240PE modules, a Schneider Conext 3300 inverter (which I will pull the data from and display in conjunction with my original Xantrex inverter), and a Schneider J1772 electric vehicle charging station.

A friend at work did my solar design and ran the calcs for my rebate application, and I worked with PWS's structural engineer on the final details of the carport structure.

Now that I have my permit, I can give PWS the green light to start fabrication of the metal, I'll call 811 to get them to mark any underground electrical services and I can start digging holes.

Here is a rough SketchUp model I made of the system.  The only SketchUp model I could find of a 914 was only half completed, so that's why it looks like it's missing most of the right side of the car.

Trusting the Charger

Alert readers will recall that I was having doubts about the effectiveness of my charger.  I was alarmed to see it was driving my 120V DC battery pack at 153V DC, when the sticker said it should be 143V DC!

After I called the Zivan repair service and described what I was seeing, they told me to send it in for testing.  After they ran it on the bench, I got a call from Mark, saying it was fine.  Here's what he said, any errors or omissions are mine.  I've been learning about the mysteries of battery packs on the streets and shady blogs, so I was happy to have another data point.

Mark said that the pack should be overcharged.  Specifically my pack should be charged aggressively up to 157V with the T-125 battery setting "4" on the rotary switch.  If I'm concerned about charging it too hard, I can set the switch to "3" which offers a slightly gentler charge for T-105 type batteries.  After 12-18 months the batteries should be mature and I can switch back up to "4" and drive them a little harder.  110% - 115% overcharge is OK.

The lower-voltage cells have to be allowed to charge up to the level of the higher-voltage batteries.  This is what Jack Rickard calls a "top balancing" scheme, which he says is wrong for lithium batteries, but he doesn't believe in lead-acid so it may be valid.

For the charging-state LED, a fairly-long yellow is OK, which represents the 80% to 100% stage.  I thought it should be much shorter, adding to my distress over the observed voltage.

Now having explained all this, I still have exactly the same symptoms that another person at Zivan diagnosed the charger as faulty and I should send it in.  This exercise cost me $70 plus shipping with nothing to show for it but a little more knowledge.

After I redid the whole charging system with the J1772 connector wired in, I did a short 3 hour charge and took it to the car show.  I noticed that the performance was rather sluggish, but the voltage was right about 120V DC and the PakTrakr state of charge was 100%.  Yesterday I decided to trust the charger and let it go through the full cycle without panicking and pulling the cord.  Well imagine my surprise when it put 13.11 KWH into the pack over the course of 10:33.  I had been doing about  8 KWH over 6 hours previously.  The voltage peaked at 153.3V DC and dropped down to 132.2V DC shortly after disconnecting the charger.

13.11 KWH is far beyond what I calculated for the 80% to 100% charge capacity, so I'm hoping that the pack was just very thirsty because I hadn't been charging it all the way up.  I did a short drive around the back yard today and the acceleration was neck-snapping like I remember from my first drive.  I hope I haven't damaged the pack.  If you know me, you know I'm a bit of a pessimist and I have this sinking feeling in the back of my mind that I now own a 10 mile pack.  Well, I'll do some range testing and see how it looks.  I just don't have the spare cash for a lithium pack so I'll have to make this work for as long as I can.

J1772 Charging - The Joy and the Agony

I continued with the work on the J1772 upgrade from yesterday.  Here I've tidied the wiring and zip-tied the wires in place.

I painted the gray epoxy repair holding the front battery in place with white plastic paint.

The box is now mounted to the side of the battery box.  You can see the nut holding the transformer bolt into place.  This will cause me some anguish shortly, I should have seen it coming...

Finally the battery box is back in and the batteries are dropped into place.

The batteries are wired up and the PakTrakr leads are all re-installed.

I did another 120V and J1772 test, both passed!  Note that I didn't tie the box's 120V output into the charger yet, so there's no load on the output.

Since the initial side-mounting for the relay box was inaccessible, I decided to mount it right on the top of the front battery box cover.  This will give me easy access to the fuses and the wiring.  I got another PVC enclosure and put the cable glands on the right side.  I attached heavy strength Velcro to the bottom, so if I need to get to the batteries, I pull the box off the Velcro and move it to the right side.  Now I just have to move the redo all of the wiring.

And here is the final result!  This is now the overall front trunk layout.  Dual charging plugs are on the left, control circuitry in the middle in a NEMA-rated enclosure, and a twist-apart AC disconnect on the right side.

Here's a close-up of the control circuitry.  Moving it to this position makes it far easier to check for blown fuses and future upgrades.

After the wiring was complete, I plugged in for yet another charger test.  Note that this was the first one with the load of the charger filling the battery pack.  After a few seconds, I saw smoke coming out of the J1772 box!  Fudge bucket!

That led to an adventure of getting the J1772 box out.  I had two problems that I should have recognized during the design.

First, the bottom bolt is mostly inaccessible - I could touch it with my fingers, but couldn't get a wrench on it.  I ended up taking out the battery adjacent to the J1772 box and drilling out the bottom bolt from inside the battery box.  I won't replace the bottom bolt because it's not necessary for holding the box in place.

Second, the big nut on the outside of the cover on the transformer bolt is about 1/4" too long and the side of the headlight area was blocking the box sliding up.  I managed to get a ratcheting socket on the nut and got it off, but the bolt was still too long.  After a lot of swearing, I drilled a 1" hole through the wall of the battery box where the head of the transformer bolt is.  I then backed out the bolt through the hole.  The J1772 box then lifted right up.

My guess was that I had botched the wiring on the AVC1 board and melted it, but when I opened it up, I found the board seemed to be in good shape, but the small wire I used for ground was crispy where the insulation used to be.

I sent an email to David Kerzel of Modular EV Power at 7:00 PM on a Sunday night hoping for a reply sometime tomorrow but he wrote back in 30 minutes!  He asked some questions about the wiring, so I've sent him my schematic and we're going back and forth on the problem.  Great customer service!

I need to get the car rolling for the Warbirds, Wings and Wheels car show this coming Saturday!  Luckily I have about 80% charge and the show is a couple of miles from my house so the charging circuit doesn't really need to be working for the show.  I'll just re-mount the box and tell everyone it works perfectly!  Update: It did work!

Here is the schematic of the full charging system.

J1772 Charging Upgrade

What do you do with a working electric car?  Tear it apart and add more parts!

I bought the 120V charger with the goal of being able to charge up wherever I can find a 120V wall outlet.  That's great for home and any houses I visit or my work building, but electric vehicle charging stations with the J1772 plug are popping up everywhere.  I don't want to miss charging from one of these stations just because I don't have the right socket in the car.

I initially thought that I could install the J1772 socket and then run the power wires in parallel with the existing 120V socket and charging circuit.  Imagine my frustration when I found that you cannot derive the needed 120V neutral conductor from a 240V J1772 plug.  J1772 does specify a 120V option, called Level 1, but it's not supported by my Schneider unit and most other public charging stations.

So I had two tasks: step down from 240V to 120V and derive Hot, Neutral and Ground wires for the 120V circuit.  The solution comes in the form of a toroidal autotransformer.  I was given a pointer for this type of device to Toroid Corporation of Maryland.  They build a standard list of autotransformers, but I needed 240V in, 120V out and 15 amps of capacity.  As you can see from the table, they don't offer that as an off-the-shelf unit.  I contacted them and told them what I was looking for, and they replied back with a custom design, spec sheet and price quote.  I confirmed with my power engineer friend at work that this was, in fact, the part that I needed.  After confirmation, I sent my order in along with $228.41 and was told to expect it in a little over a week.

In the meantime, I worked on some prototypes to see where I would be able to fit it, in the tight confines of the front trunk.  I measured the available space with the battery box in place in the front trunk.  Then I built a cardboard mockup about the size of the real transformer.

I first tried a case with a clear removable cover from Polycase like I've used in other locations in the car.

It is definitely a tight fit!  Let's see how the plugs fit along the top edge... again very tight but not bad.

But my decision was made - the power sockets stick down too far to get the wires mounted and routed within the box.

I found this Cantex box at Lowes that just barely fits.  I held it in place with my endless supply of woodworking clamps, and drilled holes for the top and bottom bolts.  The top is a through-bolt but the bottom is a rivnut so the bolt head doesn't stick into the side of the battery on the other side of the battery box wall.

And here's the mockup with the transformer and both power sockets drilled and set into place.

Here's what the sockets look like from the top - just right.

Now we move on to positioning the real transformer.  You can also see the passive air vents I installed in the bottom and side of the box to help dissipate the ~94F heat the transformer will generate when I'm charging from a J1772.

You can see that it was custom built for me because it has my name on the label inside the wrapping!

Here's the test fit inside the box.

Here's the battery box back in place to show how it will look once installed.  What you can't see is I had to cut away some metal to allow the box to fit.  It was a curved piece of thin metal that was used to hold the spare tire in place.  No more spare tire, no more need for that bit of metal.

And here's how we access the charging sockets.  Both are waterproof in case some water gets in while I have the hood up during a charge.

I added some additional venting to the cover to make sure the heat isn't building up inside the box.

Here's the box, fully wired up inside.  Note the 3/8" hole drilled in the center of the transformer's potting material to allow me to run a big bolt through the box to hold it in place.  I also cut some steel spacers to size to center the transformer from side to side in the box to increase airflow.

You can't see it from here, but I installed the AVC1 board from Modular EV Power inside the box.  This board implements the car side of the J1772 safety charging protocol.  You could jury rig something up to fake it out, but for $37 on eBay you can't go wrong.  I hooked up 12VDC and ground, AC ground and a wire to the Proximity and Pilot pins of the J1772 socket.  I didn't take advantage of it yet, but it has normally-open and normally-closed relay outputs.  You could use this to light up an indicator that the car is charging or put an interlock so you can't drive the car while you're charging.  Not that anyone would every do that, of course!



If I had to do it again, I would use the newer AVC2 board, as it's enclosed in a nice box with better mounting tabs to protect the electronics.



I could have used twist-on wire nuts for the AC cabling, but I don't trust them enough to stay in place with all of the vibration this will be subject to.  I used Polaris connectors, which we use for high-current, thick-conductor bonding at my solar power company.  They're also dipped in plastic so they won't short out against anything else in the box.

Now that it was all wired up, it was time to test it.  I connected my Schneider station to a 240V breaker in my workshop.

This is the guts of the Schneider unit.  Technically it's not a "charger", it only delivers AC power to the charger mounted inside your car.  This is known in the industry as EVSE, Electric Vehicle Supply Equipment.  It consists of a couple of fuses, a contactor and a small circuit board that implements the EVSE side of the J1772 safety protocol.

I flipped on the breaker and the top green light lit up, signaling that it was receiving power and was anxiously awaiting an electric car to appear nearby.

The first test of my charging system was plugging in a 120V extension cord.  This tests whether the 120V hot, neutral and ground are wired correctly to the output cable, without involving the transformer.  Good news, it worked!  You can see the light in the extension cord and the voltmeter reading 119.8V at the output of the box.

The big test was now plugging the Schneider into my box.  For this to work, everything had to be wired right - the 240V from the socket to the transformer, the transformer to the 120V circuit, and the AVC1 board connected and properly talking to the Schneider safety circuit.  Happily, when I plugged in the J1772 and the thumb toggle clicked into place, the small LED lit up on the AVC1 board indicating it was ready for charging, then the contactor clunked in the Schneider, then the voltmeter showed 119.5V!  As soon as I pressed the thumb switch, the Schneider and AVC disconnected and power stopped flowing, just like it should, even with the plug firmly seated in the socket.

Declaring success, I will wait for Saturday to tidy the wiring inside the box, mount the box to the side of the battery box and install it back in the car.  Then I'll do the final wiring back up to the relay box and the rest of the conversion wiring harness.

Zivan Overcharging

I'm doing my 7th charge right now and have learned that I have to watch the Zivan 120V DC charger very carefully.

The first couple of charges went as expected.  There's a display LED on the unit that tells you what charging phase you're in.  When first plugged in, a red LED comes on, it beeps twice and two stages of fans come on.  When it reaches the 80% charge point, a yellow LED comes on.  When it reaches the 100% charge point, a green LED comes on and the charger shuts off.  This should allow fo unattended charging overnight.  The first 3 charges put in about 8.0 KWH to 8.5 KWH, ran through to the green LED and shut off around 6:45 hours to 7:00 hours.

This matches what I was expecting, given the characteristics of the battery pack:

242 AH * 6V = 1452 WH per battery.
1452 WH * 20 = 29,040 WH = 29 KW per pack.
Charge range goes from 70% to 100% of pack capacity = 30% of pack capacity
29 KW * 0.3 = 8.712 KWH.

On charge #4, I noticed that it was still yellow but was up to 9.0 KWH at 7:00 hours and the pack voltage was 157V!  The PakTrakr was reporting overcharge on several of the batteries.  I pulled the plug and let it stand until the next day.  The pack had dropped down to 128V or so and PakTrakr said all batteries were now good.  I decided to check the water level and found that every cell, 60 of them, was about 1/2" down from full, but not anywhere close to the top of the metal plates.  This may be natural, but I certainly wasn't expecting to water the batteries every 4 charges.

Charge #5 was a partial, interrupted by a popped breaker as I was charging it from a different plug on a different side of the house.

Charge #6 hit 157V again, at 7:16 hours and 8.68 KWH.  Now I'm worried and I just can't trust the charger.

Here's the whole charger and the two stickers on the charger:

I do have T125 type batteries with 240AH capacity (actually rated at 242AH) , so hopefully the factory settings were done correctly.

This F5 sub-model isn't even listed on the Zivan web site, only ones starting with F7.  Also, there's no explanation of the U1, U1E, U2 and U2E voltage values and how the charging phases work in the NG3 manual.  The charging curve page is blank!  You can see that my physical model doesn't even look like the one in the manual.  Mine doesn't have the long bar of LEDs showing the charging state, just the circular 0-80%, 80%-100% and 100% charging LED display.

I just spoke with the staff at Zivan / Elcon in Sacramento and described the symptoms.  He had me check the small rotary switch and it's set to "4" which is correct.  He said given the model and switch setting, it was correct for my batteries and shouldn't go up to 157V and said I should send it in for testing.  When I asked if he would be able to look at it within a couple of days of arrival, he said "Jesus, no way", as he has about 150 chargers already in line for repair.  He gets about 3 more every week.  Even though I just got it into the car, it's out of warranty and if I pay a $35 rush fee he'll look at it soon after arrival.  I need to include a note stating what I'm seeing happening, the make and model of my batteries and my contact details.

Wow, I'm not thrilled that this is happening so soon after getting the car rolling.  I have to decide when the time is right to send it in, as I have a couple of week-long out-of-town trips coming up.  I get to pay for shipping two ways, repair and possibly rush fee.

I'm 4 hours into the charge now and it's at 134.6V and 5.35 KWH.  I use a Kill-A-Watt unit to watch voltage, current, AH and time of power delivery.  I'll watch it closely through the rest of the charge.  Next charge I'll log 15 minute data on volts, amps and KWH so I can see what's going on, and report it accurately to Zivan repair if necessary.  If the repair is prohibitively expensive, I may just go right for a 240V AC charger.

J1772 Car Charging Station

I have a plan to install a 3-car-size shade structure and put another 3.3KW of solar power on top of it.  This will get me down from about $30/month to a $0 electric bill, even with the burden of charging the car.

In order to move into the future I leveraged my employee discount at REC Solar / AEE Solar and bought a Schneider Electric J1772-standard EVSE (Electric Vehicle Service Equipment).  It's technically not a charger, but a "smart" device that delivers power to the car and works with the in-car J1772 controller and charger to intelligently charge the car.

I'll mount the EVSE to a post on the carport.  I will also add it to my profile on PlugShare to offer J1772 in addition to a 120V outlet to anyone who needs a charge in my area.

Checking out the charger

After the success with getting the motor turning, I moved to the charger.  It was all wired in, I just needed to plug in the heavy 120V extension cord and see what happens.

Success!  The red light shows it's in the startup phase of the charging cycle.

The temperature probe is deep inside the junction of 4 batteries in the front pack.

Here's a view of the charger in action.  Note that the charging system cuts off the 12V to the rest of the car and the state-of-charge gauge goes offline.

 

The battery box fan circuit came on, exhausting the hydrogen generated during charging.  Each battery box has a fan and a duct to the outside.

 

The DC voltage pushed out of the charger started around 139V and peaked at 148V.

 

I put my Kill-A-Watt between the wall plug and the charger cable.  It was dark by then but you can see it's pulling 14.64 amps...

for a total of 1262 watts.  That's a good size microwave or hair dryer.  Next time I do a full charge I'll get it to give me the total kilo-watt-hour (KWH) count of energy pulled.  This will help me estimate my yearly energy draw to see the impact on my solar power system.  I currently make about 80% of the power I use, so adding car charging will drop that down.  I have a clever plan to add a solar carport over my parking area.  This will shade the cars to keep them cool in the blasting heat of the sun, plus make some additional solar power.