June 24 '2013
Foreword to the reader:
This is very simple web site (really a draft put together in a couple of hours) and at the moment it only contains annotated photos of my professional upgrade of charging system in my Nissan Leaf 2011 model with options of additional 3.3kW, 6.7kW or 10kW charging power. The technical work is completed and all information presented here is released in public domain as open source project. I believe is can benefit technically inclined Leaf owners to carry out (and improve) such upgrade themselves. Work requires some specialized equipment (such as a vehicle lift) and basic knowledge of electric circuits. The upgrade is my personal project, I do not do this on hire as business. If you decide to roll up your sleeves, and try something like this on your own, please check out my disclaimer.
I should mention that starting 2013 Nissan Leaf comes with 6.6kW stock charger, so recharging could be done even faster - at 16.6kW total power. Electrically the upgrade described below is pretty much identical and still applies, but the location of charger, water cooling plumbing has changed.
Before starting I want to make sure you understand the terminology
right. It appears like more than 90% of people get this wrong. Just search for
"Nissan Leaf charger" and all kinds of home EVSE's (Electric Vehicle Supply
Equipment) will come up - you know, those boxes you mount o the wall of your garage that
have a cable with white charging handle you plug into your Leaf. Some also call them
charging stations. Why industry supports confusion of people is a mystery. Just remember -
what's on your wall is NOT a charger nor it is a charging station! It is nothing more than
a fancy relay safely connecting your home 240 VAC mains to actual charger, which is
integral part of your Leaf, it's mounted on the floor behind rear seat (in 2011 models).
This is the only device called charger.
If you notice, the output of a wall mounted EVSE is still the same 240 VAC as comes to its input - just measure voltage across both larger prongs of the J1772 plug or receptacle. So external EVSE cannot possibly be a charger! Leaf's battery is around 360V and obviously DC, so integrated charger converts incoming 240VAC into 360VDC (up to 390VDC near fully charged battery) - that's what chargers do. Now, fast 100 kW DC charger the size of commercial refrigerator on a concrete pad you see in public places is indeed legitimate external charger - it produces 360...390V DC directly applied to the battery terminals, just like integrated internal charger does. Thus those DC fast chargers are never called an "EVSE".
The 2011 Leaf's stock Nichicon charger is rated at 3.3kW output power (3.6kW input power), so with 10kW upgrade total charging power will be 13.3kW. This allows charging Leaf's 24kWh battery to 80% in less than 90 minutes if standard residential level II EVSE is used and 240VAC outlet is available. In addition to this, if only such outlet is available (not EVSE), the Leaf can still recharge fast - all you need is Leaf's inline EVSE that comes with every Leaf (everyone calls it trickle charger, but it is not a charger at all). In this case stock internal charger is limited to 1.4kW so the total charging power is reduced to 11.4kW and charging time increases by 10 min and will be ~100 min to 80% SOC). This means freedom to fast charge anywhere - in a garage, in the work shop where that portable welder is usually plugged in, in any RV park, shore power dock for boats, any home 240VAC socket (dedicated or - electric dryer) - anywhere where you can find regular standard 240V 30A to 50A outlet. Not as fast as from those gigantic public 100kW DC fasts chargers on concrete pads, but the next best thing.
This upgrade uses one, two or three OEM EV chargers made for the job, specifically - BRUSA NLG513-U1-01A-A01 (formerly known as NLG513-WA). Just like Leaf's stock Nichicon charger, these units are water cooled, with similar specs. Output power can be adjusted on the fly by a simple potentiometer. Chargers were programmed to run in automatic mode and loaded with booster charging profile. They are mounted under the trunk floor so no extra trunk space, cabin space or under hood space is taken - all interior and exterior appears stock. Regardless of the upgrade option, total boost power can be adjusted from zero to maximum available by a knob under hood where a power meter along with boosters' status indication is installed. You can monitor line voltage, adjust line current or line power to any desired value, and monitor total power factor. This ability comes handy if you can plug to shared circuit or the outlet limited to, say, 30A - in this case you just set line current low enough to prevent circuit breaker from tripping. The power and current displayed and adjusted as total for stock and boosters combined.
The concept of the upgrade is simple and will be clear from this block diagram. The booster consists of one, two or three BRUSA NLG513 chargers. The stock charger communicates with Leaf's BMS and timers as before, and knows when to slow down or stop charging. All booster does is looks at AC current consumption by the master (stock Nichicon charger) - if master is running at full bore this enables boosters at full power as well. Master does not know (and does not need to know) that additional current is being injected into the battery. The current measurement is done inside the battery pack and all the BMS sees is the rate of charge and controls master accordingly. When BMS determines that enough Ah was put back into the battery (which now happens quicker), it commands over CAN to master to slow down and eventually quit charging. The line current draw by master quickly decays and ceases. The added current sensor detects this condition and simply disables boosters as well. That's it. No CAN communication or pilot controlling boosters is required - they basically monitor master's current and double (or triple or quadruple it depending on how many booster units you install) - this is described well in the user's manual for the NLG513 charger. Booster #1 is somewhat special - it's digital outputs are used to indicate the progress of charging using three blue LEDs - very similar to the stock indication in Leaf. Other boosters, if present, can have the same charging profile, but their digital I/Os are unused.
The stock functionality of Nissan portable EVSE trickle charging from any 120 VAC outlet is 100% retained and no interference from installed boosters occur. If 240VAC is not supplied, booster(s) remain off and not visible to the rest of the vehicle, so you can charge from any 120VAC outlet as before. Moreover, you can take advantage of boosters working off of 120VAC for more than 50% faster bulk charge than 12A portable EVSE can do alone. In this case each booster is limited to about 9A total input current, or total of up to 27A in addition to stock EVSE. So you can plug the system into a 120VAC outlet and dial total line current to be just under 20A - 12A will be drawn by stock charger and additional 8A by the booster(s). This will recharge 67% faster than just portable EVSE at 12A alone.
One may initially choose less expensive option with one booster and see if this satisfies quicker charging needs. Because cabling and mounting provision is made to use one, two or three boosters, adding more units later on is simple plug and play procedure that takes about 1 hour.
Also, boosters will charge and maintain along auxiliary 12V lead acid battery topped off. There have been reports that if Leaf is left connected to an EVSE for extended period of time (weeks) to maintain traction battery, this keeps VCU, security, Carwings comm, timers, climate control and other subsystems on, drawing power from 12V battery which is not being recharged, thus runs dead. The suggested cure is not to leave Leaf connected to EVSE, but this will prevent traction battery from being topped off and ready after that long vacation time, although self-discharge rate is quite low. With proposed booster upgrade this is no longer an issue since as long as boosters are powered, even if they don't charge traction battery, they do charge auxiliary 12V battery so the car's subsystems can keep working if so desired. As each booster has isolated 12V output (actual voltage is 14.2VDC). Whenever AC power to boosters is present, each will output at least 0.5A into the aux. 12V battery even if main high voltage output is disabled. Because internal power supply is set up as current source, all outputs can be paralleled and total current is automatically shared between all the boosters equally. With 3 booster units you will get >1.5A of charging current.
As mentioned above, BRUSA NLG513 chargers have analog voltage input control, allowing to set AC line input current (and so output power) from zero to maximum with a single resistor. I installed potentiometer connected to analog inputs paralleled for all the boosters since they can share analog ground signal. This way I can dial any desired line AC current (0...16A per booster unit) simultaneously for all installed boosters. This is relevant if the circuit breaker is not rated for the maximum mains current all the boosters can sink, so you can dial the current about 20% less than this rating to prevent tripping. Monitoring is done by line current meter installed under hood next to the potentiometer knob. The pot can be replaced wit ha few fixed resistors and multiple switch to select predetermined line current from few common values, say 20A, 30A, 40A and 50A. Normally limiting power draw is done by CP (Control Pilot) signal from an EVSE, and it is trivial to make boosters to comply, but with just 240VAC outlet there is no CP source, so if you ever run into the 240VAC mains outlet which cannot handle 50A, adjustment can always be done manually.
The power for boosters is fed using extra inlet with standard 250VAC 50A twist lock connector. Even though the additional inlet is not visible, I wish I could get away with feeding all the power needed though the existing J1772 connector. While the standard allows up to 70A AC current feed, it doesn't mean actual hardware has to take advantage of full allowed capability - in fact the power contacts of Leaf's J1772 inlet are designed to handle up to 30A continuous current (up to 40A peak). It makes no sense to beef up the inlet contacts to allowed 80A if designers know the stock 3.6kW charger will only take 3600W / 240V = 15A (6.6kW - 27.5A). So why didn't I use second J1772 inlet, say, next to the stock one? Because three boosters will draw 48A from the mains and 30A rated contacts will overheat and fail in a hurry. Why then I did not split total 64A max current evenly between the two inlets? They each could handle 32A... Well, I could, but aside complications with routing, stock Nissan EVSE "Brick" would never provide me with 16A 240VAC power, and I would have to have not one but two wall EVSEs to take advantage of such setup, which would defeat the purpose of freedom from any EVSEs. But the main obstacle is I would have to reprogram stock VCU to signal first EVSE that stock charger is now 32A capable (which it is not) and come up with my own pilot signal controller for the second EVSE. So comparing the choice between adding one 50A twist lock vs. adding one more J1772 inlet + two 32A capable EVSEs + pilot signal control circuitry makes it clear.
I wonder if most people just want faster charging but don't mind to be tied to a home or public EVSE, or they rather want to be free from EVSEs of any sort. For the former, using existing J1772 inlet makes sense, but again, fast charging at 13.2kW (16.6kW with 2013 Leaf) is not guaranteed if an EVSE is not capable delivering that or stock inlet just can't handle it - it may well get too hot and fail in a hurry. On the other hand, a 240VAC socket is usually capable of 50A (with 30A being minimum), and the solution described here does not require messing with CP signal modifications, so I went for it as a first step. Sure not everyone will want extra connectors on their Leafs no matter what the benefits. For the time being those will have to wait and be tied to their EVSEs and slow stock chargers.
Also, realize, if you have 10kW booster, this doesn't mean every EVSE out there is capable of supplying ~15kW (~3.6kW on AC side for stock and 11+kW for boosters) through it just because Leaf asks for it. The handshaking between EVSE and Leaf includes indication by EVSE how much current it is rated at and able to supply, so if it is less than what Leaf can absorb, the stock charger is throttled back to comply. Common wall EVSEs are made to handle up to 32A mains current so that an EV can take full advantage of its stock charger (at least 6.6kW), but with 10kW upgrade this will no longer be the case - the more powerful charge you carry onboard - the less EVSE's around will be capable of handling it. I've never seen an EVSE handling 48A mains. In contrast, with home 240VAC outlets there is no such limitation - if you have 50A breaker, you can get 50A charging current. An electric range, stove, powerful cloth dryer, and such are all directly plugged into 240 VAC without any EVSEs in between - no problem. So can your Leaf.
As I mentioned, the 2013 model Leaf comes with 6.6kW charger that is moved upfront - this doubles charging power and halves the time required to achieve 80% of charge. 4 hours is pretty good, but still, with independent 10kW boosters you could have 16.6kW total charging power and recharge your leaf from empty to 80% in just 1 hour 12 min. But, again, this is not the only objective - the goal is to give you option not to use EVSE at all - freedom to go anywhere where public EVSE is not available.
The cooling lines will be extended toward the back of the vehicle where boosters are installed - much the same way as it is done for pre-2013 Leafs, possibly even re-using stock pre-2013 Leaf's plumbing. I will have to look closer into 2013 model design when one will become available to me for examination.
If you choose to duplicate this upgrade, I would advise seeking professional installation approach as the work requires to remove and re-install car's battery for efficient and neat OEM-like cable routing - something simple (the battery is held in place by 12 bolts) but impractical without vehicle lift. With all the supplies available the work can be easily done in one day by one person. It is possible to route cables under body without detaching the battery, but keep in mind the conduits it will be exposed to the road and the outcome will be quite amateurish looking (and potentially unsafe if you manage to rub conduit against something on the road to the point that conductors get exposed or smashed together).
Most of the photos are self-explanatory, but I will expand description next to each photo as I go along. In this description I will use two NLG513 booster chargers available to me at the time the work is performed, but as mentioned, cables and connections are pre-made for 3 booster chargers, so I will install additional booster just for the completeness at my first good opportunity. For now two boosters and stock charger with portable EVSE that came with my Leaf work together just awesome. My usual routine is to charge from 30 miles to 80 miles range in 30 minutes which increases usability of my Leaf to the point that I may not need my second EV anymore.
So what's next?
The feed back I'm hearing is that ability of recharging fast(er) is more important than that + independence from any EVSE, and use of additional non-1772 inlet is odd. Some people just want to plug in as usual and charge faster, that's it. So my next upgrade may address these issues.
Pros (of using just J1772 inlet to feed all the
power into Leaf instead of separate dedicated inlet):
- Your operation is identical to how you charge now - use the single inlet and J1772 compliant EVSE.
- No extra connectors in Leaf, the power for boosters is supplied through the same power pins as for the stock charger
- You loose independence from an EVSE and cannot just charge at any location from a 240VAC outlet.
- You may not always take full advantage of the booster's power capability (esp. with 6.7kW and 10kW booster) because EVSE will actually dictate how much power it is capable of supplying and only that much is allowed to be drawn even though boosters themselves are capable of more. With 240VAC feed you always get full power output as long as circuit breakers can handle it (if not, you can manually dial in any desired current)
- Your auxiliary 12V battery will not get charged if the main battery is not being charged (there is no power from EVSE at that time).
- You need to buy EVSE for home use - extra expense (compared to the independent 240VAC with own twist lock inlet solution)
- Likely more expensive upgrade as now the controller receiving info from EVSE about its current (power) capability and adjusting booster's power not to exceed it will be required.
- chance that with 3 boosters contacts of your J1772 inlet port may get too hot and fail prematurely - this has to be investigated in more detail though. With two boosters there should not be any concern and with one booster it is not an issue at all since with one unit total charging power is 6.6kW which is the same as for stock 2013 model with identical J1772 inlet.
Despite more apparent cons than pros, this variant is quite straight forward and might appeal to some Leaf owners more. It can be implemented without tapping into Leaf's CAN bus and disrupting it's stock operation.
Another improvement I did not think about before but now is implemented is to unbolt stock ChaDeMo inlet and relocate it under the hood between radiator and inverter (see update July 29 below). The twist lock inlet goes in its place. This is kind of afterthought - the hole in the plastic bumper is already made and swing open license plate worked just fine, but both most often used plugs (in fact always used together) makes sense to locate next to each other side-by-side. In rare occasions when I'd still want to use quick DC charger, I'll have to open the hood and plug ChaDeMo connector there, but I had to do this only once in 1.5 years of my Leaf ownership and even that was just for curiosity how this works rather than necessity. So for all practical purposes I'd say I never use QC inlet. So if I do need it, I can wait 5-10 min sitting in the car with open hood and it will get enough juice to always get me home.
The info on this page is open source, I will add design details here as I get time to do it. Anyone is welcome to use it, welcome to come up wit ha kit to sell to people, use information here as you see fit. Sorry, I cannot perform upgrade work for you - it's too much effort for one person to organize.
Finally, please get used to proper units. Don't display your ignorance claiming your Leaf stores 24 kWh worth of power in its battery or your car's power consumption is 4 miles/kWh (or worse - 4 miles/kW), or Leaf has enough power in the battery to cover 100 miles - it sounds as silly as saying you stored 300hp in your gas tank, or your fuel economy is 4 miles/hp or 60hp gas engine is enough power to cover 400 miles. Power and energy are two related but very different physical properties! If you care, see my rant about units usage here (ignore bottom portion of the page - the site appears in the identical style but actually belongs to my totally unrelated Audi conversion project).
One last remark - for obvious reasons all photos have embedded watermark - I tried to make it as non-intrusive as possible while still visible. Welcome to save and disseminate any material you find here with proper credits to this site and its author.
Enjoy the site and welcome to link it (or to it)! If you have questions or would like to suggest something or share notes, welcome to drop me a mail.
Trunk compartment (under trunk floor) - view from back - this is where booster chargers will be installed.
Trunk compartment view from front
Trunk cover off
Cooling hoses connection
Appearance of three NLG513 water cooled for 10kW booster option
Two NLG513 water cooled for 6.7kW booster on rails - initially this will be installed and tested
Back side (Will be facing up)
Trial fit - temporary support
Trial fit - side view
Two units installed
Cooling hoses links
Side view - installed at the angle
Cooling loop re-routed through NLG513 chargers
Another view - cooling hoses
J1772 inlet removed
Stock wiring - resistor and diode buried here
Harness to be modified to install measurement circuit
Interlock loop of internal connector
Harness taken apart
New connections made - splices are filled with water repelling compound
All crimp connections protected with anti-oxidizer (Noalox compound)
Harness modified - extra stub on top, otherwise - just like stock
J1772 inlet harness installed back
Inner connector from the harness toward the stock charger
Stock hardware - view from the front (rear seat removed) charger on the left, ultracapacitor on the right
Charger harness detail
Battery power and BMS connector
Battery connector unplugged
Battery case connector
Harness is shielded
Preparing to splice wiring here
Spliced high voltage wiring for NLG513 chargers feed
Stock harness restored
Underbody - no place for cables.
Battery has to be removed and cables routed together with stock harness
Vehicle is lowered so the battery lay on the battery support fixture
Twelve bolts around holding the battery removed. Vehicle lifted with battery remaining on the fixture.
Overview of this step.
Battery - front view
Battery close up
Battery - rear view
Underbody cavity with stock harnesses and cooling pipes
Routing AC power cable along cooling lines
Blue harness - new signal cable to the power meter and indication upfront
New harnesses and cables routed along the stock ones - not even visible.
Close up detail - black is AC power in cable (to the NLG513)
Battery and signal connections - all identified and marked
AC current sensor clamped around AC feed wire
Another view of the sensor PCB
Checking and logging charging voltage on the battery terminals (logging multimeter shows 1V lower than Agilent)
Precharging voltage ramp time measurement
Two boards quick through-hole design (display PCB on the top and current sensor PCB on the bottom)
Under hood right side
This is the spot for the meter and indication/control
Hole cut in the plastic cover
A panel meter installed (back view)
Panel meter PCB with current transformer
Close up - both stock and booster chargers fed through the transformer - meter will measure total current/power
Plastic cover is started printing on the prototype 3D printer
Cover as comes out
Meter PCB covered with this cover
Plastic cover under hood re-installed. A knob below will adjust NLG513 boost power (connected to "power indicator")
Testing with stock trickle charger from 120 VAC source - line voltage reading
Testing with stock trickle charger from 120 VAC source - line current reading (stock charger only)
Testing with stock trickle charger from 120 VAC source - line power reading
Testing with stock trickle charger from 120 VAC source - power factor reading
Location of the extra power inlet for twist lock 50A line power connector (this has been changed, see July 29 update below)
The power inlet will be installed here and covered with license plate (this has been changed)
Recessed inlet with flap cover - the license plate frame will be attached to this flap (this has been changed)
License plate attached to the flap cover - they will swing up open together on the flap's hinge, providing access to the inlet (this has been changed)
License plate final position - front view
License plate final position - 45 degree view
License plate final position - side view
Junction box with breakout strips and cable glands prepared
Junction box installed
Chargers' interface connectors crimp terminals and the tool
All three harnesses completed even though two boosters are going to be installed for now
Power in, power out, panel meter indication, current sensor, chargers controls and serial interfaces - all prepared for wiring
Half way wiring done - crimping ring terminals
Wiring completed - it may look visually messy, but electrically it's perfect (that really counts).
Installation complete - view from the front. The spot for a third booster is empty for now.
Same step - view from the rear
July 5 2013 update
50A 240VAC power plug and receptacle (this has been changed, see July 29 update below)
The plug is inserted and the license plate rests on it. At this point only aux. lead acid 12V battery is being charged (this has been changed)
Another view of this connection (this has been changed)
Both connectors are plugged in - side view (this has been changed)
A 120VAC receptacle mounted inline about 1.5m away from the power plug and portable EVSE is plugged into it (this has been changed)
Top view of all connections near Leaf - top view (this has been changed)
All you need to carry with you for the universal connectivity is this cord (and optionally - adapters for different 240VAC receptacles)
Preliminary test - line voltage as measured at the input socket
AC Amps consumed (all chargers combined) - this is true value and I can adjust booster's portion of it with the knob
Total power from the mains displayed*
Power factor is displayed
Auxiliary lead acid 12V battery is always being charged and topped off while 240VAC is connected.
Short video of the display in auto cycling mode and how the plugs are arranged (82MB .mp4 file)
Earned bragging rights are now expressed in this bold message
Experimental data plot obtained by data logging battery voltage with two boosters running (9.9kW total input power, 8.42kW total output power). The battery started charging nearly empty (6 miles remaining range displayed)
Downloadable charging profile loaded in each NLG513 booster charger
July 29 2013 update
The 240VAC 50A twist lock inlet took place of quick DC charge inlet, which in turn was relocated to under the hood. This proved to be so much more logical and convenient arrangement. This and other design details are below.
Schematic of new connections, current
sensor circuit, etc.
Layout of the current sensor PCB - I used through hole components and single sided PCB for simplicity.
STEP file of the booster chargers brackets along with hole pattern template
Another view of the current sensor PCB with split core CT clamped around stock Nichicon charger's AC input wire
The harness connecting data lines of the J1772 quick DC charger connector - unplugged before removal
Stock DC quick charging inlet and wire harness. Signal interface harness is too short for the new inlet location.
The data interface is extended by 25cm. Now the inlet can be relocated. (Granted, this wiring will be inside corrugated loom)
Twist lock inlet and adapter plate made from 10mm thick black nylon plastic. Its' shaped after stock inlet flange.
The twist lock connector is installed and AC power wires connected. Now it is going to take place of DC quick charge inlet.
Installed. Fits beautifully and looks great.
QC port is relocated right behind the hood latch. Short power cables actually dictate exact position
Side view of the port installation. It's held in place by two aluminum brackets. The water hose runs where it use to run.
Now, this makes much more sense. Twist lock inlet receptacle is just as convenient to plug into as into the J1772 inlet
Close up view of both connectors. Looks great and works as a charm!
Aug 30 2013 update
This is final part of the project which is completed now. A third booster charger was installed for completeness and demonstration of usability. Total booster power is 10kW now; about 28 A of charging current supplied in addition to 9.5A provided by stock charger. IF charging is done with stock Nissan EVSE "brick", the total charging power is 11.2kW, (or 13.3kW with home EVSE) - twice as much as newer 2013 Nissan Leafs with 6.6kW stock charger offer. This means my 0...80% recharging time is down to 100 min (~ 87 min with EVSE) respectively. Another way to look at it is gaining over 1 mile of range for every 1 minute of charging at home. I'm impressed.
So final steps of the project depicted here:
Provision was made for installation of 3
units including water loop, power cabling and signal harnesses, so...
... installation is a snap. All it takes is placing new charger under the rack line up mounting holes and...
... securing it with four M6 bolts. Blue LockTite was used to prevent loosening bolts.
Last step is to plug all cabling in place and fit cooling hoses onto charger's inlet and outlet fittings.
The job is completed. The only remaining step is to refill cooling system with fresh coolant. I didn't want to re-use drained one.
Another view of completed installation.
The BRUSA NLG513 chargers are so flat (88mm) that if I needed to I could fit another layer of them
Side view of installed units. Eye candy...
Yet another view of the installation
Close up view. Signal cabling is color coded for easier identification.
Stock plastic cover is back on. As if nothing happened.
The test reveal that all 3 units run at full bore as expected. The output power of boosters can be adjusted if needed:
This video demonstrates that if the mains cannot handle 60A total current, you can adjust booster's power down to zero leaving about 12A drawn by stock charger, as if booster is not there. (~20MB .mp4 file).
* The power readout is
accurate only for 120VAC. Because the meter has single voltage input to compute power,
with dual voltage feed (120VAC to the stock charger and 240VAC to boosters) meter is
switched to 240VAC line and computes power by multiplying line amps by line voltage as if
stock charger also gets 240VAC input. In reality only boosters are fed with 240VAC, stock
charger gets 120VAC from portable EVSE, so the portion of watts reading for the stock
charger is twice as high - 2640W (240V * 11A drawn by portable EVSE) instead of 1320W (11A
* 120VAC). So I have to subtract 1.3kW from the total to get accurate watts figure.
However, Amps drawn readout is always accurate and this is what's important to avoid
circuit breakers tripping.
This shortcoming can be fixed by using dual input voltage meter and processor to do math which might be on the improvements list. For now though, off-shelf panel meter works for me very well.
Copyright (c) 2013 Victor Tikhonov, Metric Mind Corporation