Electric Porsche 944

Connecting 300V and Modifying the JLD404

by on Jan.13, 2013, under Electric Car

The CALB battery pack went together very smoothly and easily, in a bottom balanced fashion in series with the remaining A123 pack.  The final pack is 92S with the smallest cell being 61.08Ah giving me total pack size of about 18.26kWh (using 3.25v as the average voltage during discharge).  This gives me a 97-117km range at 100% discharge, which is not something I would do every day but represents the drop dead range of the car.  The safe range is 78-94km at 80% discharge which gives me a very usable car.  My commute to work is only about 16km each way giving me plenty of range for errands, going for lunch or being able to absorb a cycle or two of not being able to charge either at home or at work.  I have one CALB cell which has internal discharge, when it arrived I flagged it as having a lower open circuit voltage (OCV) than the rest of the cells, after cycling and bottom balancing the pack I revisited the OCV before I set up the charger.  This same cell is the only cell in the pack that now has a lower OCV than what I bottom balanced to.  I’ll have to contact Don and Keegan and see if there is anything they can do, I believe the cell will need to be replaced.  It’s other specs are quite good, it has low IR and high capacity, I would say it easily falls in the top half of the cells for performance.  The decreasing SOC with no load just doesn’t work though, I will be able to use the cell for awhile as long as I don’t discharge 100% and I keep an eye on it’s state of charge relative to the rest of the pack.  This won’t work long term and takes me back to the issues I had with the A123 cells (constant monitoring required).


With the decrease in available battery current with the replacement of A123’s with CALB’s I needed a new smaller fuse, I went with a 500V 400A Ferraz Shawmut semiconductor fuse that is mounted with the service disconnect in the rear of the car splitting the two halves of the CALB pack.  The service disconnect is the same one that Jack R. has started to use in his vehicles it’s made by Blue Sea Systems it has very solid 1/2″ terminals and is rated for 600A continuous, 900A for 5 minutes and 2700A for 10 seconds which is well above what the battery or motor can do.  The down side is it’s rated for 48v, so it’s not a switch that could be reliably used as a safety disconnect, one instance of breaking 300V @ high current would likely destroy the switch.  This switch will be used to break the circuit only when there is no current flowing, IE for storage or servicing.  You can buy it from EVTV or find it at some marine supply stores.




The EMW Dashboard is a bit finicky and although it is very attractive and provides a great deal of information (when it works) just isn’t at a state of hook it up and go.  I’m not going to give up on the system, however I am going to supplement it with what is already known to work.  The JLD404 from Lightobject.com or EVTV isn’t the most attractive meter out there, but it runs off a shunt, displays voltage/current/Ah and has been shown to be very repeatable and reasonably accurate.  I will be mounting this meter in the glove box to make it accessible but hidden.

Since I will be using the JLD404 as a backup system I’d rather not have to think about it, which includes resetting the meter to Zero the display after charging.  It’s likely the meter will be very close to zero due to the efficiency of charging LiFePO4 cells however I don’t want to have to open the glove box, hold the down arrow just to reset the meter after a full charge.  The first step in automating the reset is to bring out the terminals to the down arrow switch.  On my version of the JLD there are some unused terminals on the back, 11, 12, 13 and 14 aren’t connected to anything.

The first step is opening the meter, which is simply two tabs at the front and the rear cover slides off.  Next remove the front face by sliding it off the latches on the PCB’s.  De-solder the daughter board with the small transformer on it to gain access to the 2nd terminal on the down arrow switch from the inside.  (If you are uncomfortable de-soldering and re-soldering the daughter board, you could solder the wire to the front of the board directly to the switch, this may make it difficult to re-install the front or rear cover)  Once the meter is stripped down to this state simply solder in two wires as shown below:

JLD rear switch

JLD rear switch2


Then re-attach the front panel, this photo shows the relationship between the new wire and the location of the down arrow switch.

JLD rear switch3

Attach the rear cover and put some kind of marking on terminals 11 and 12 to represent the newly wired switch.

JLD rear switch4

This simply allows you to put a momentary switch or relay in parallel with the existing switch to activate the Ah reset function on the meter.  I will be looking at interfacing this with the “charge complete” signal from the EMW 10kW charger so when the pack is charged it automatically resets the meter.  I’m sure there are other chargers that have some sort of charge complete indication and this could be used to reset the meter.  If the pack is only partially charged the meter will continue it’s count uninterrupted.  Doing this for many cycles will accumulate an error over time that will not give you a true representation of SOC.  Allowing the charger to fully charge the car and reset the meter will counter this.

Terminals 13 and 14 also remain unused so if I don’t need to break out any other signals to make the auto reset work I may use them for the other switches, which would make it much easier to cycle through the Volts, Amps, Ah, Time display.  All of the switches share one common terminal so it would only require 4 terminals to break out 3 switches.  Using the Set, Right arrow, and down arrow would provide full functionality for the meter since all menus continue in a loop making the up arrow optional.


1 Comment more...

Installing CALB’s and Some Good Customer Service

by on Dec.31, 2012, under Electric Car

I received 2 crates of CALB’s on Wednesday the 19th, it’s now Friday the 28th and all of the cells are tested, bottom balanced, and assembled into 7 cell groups for easy final connection when I move them into the car.  In contrast after receiving A123’s it was months before I had enough cells tested, grouped, assembled and usable, and that was just a test pack, not a complete battery system.

There was an issue when the shipment arrived, 72 cells fit perfectly in the two crates but that didn’t leave any room for the bus bars and hardware.  I emailed Don right away explaining the issue along with my requested delivery date.  Since I’m on vacation till the new year I had hoped to get the CALB pack installed in the car before going back to work, so I would need the bus bars on or before Dec. 27th.  Sure enough, mid morning on the 27th UPS arrived with a bubble envelope containing the missing items.  It’s easy for a company to look good when there are no problems, it’s how a company deals with a problem that really shows how dedicated they are to their customers.  Don Blazer and Keegan from CALIB Power came through, delivering what I needed when I needed it even during this very busy holiday season.

The 2nd issue was based on a lack of availability of the proper bus bars for the CALB CA60FI cells.  For side to side connection the cells require a 41mm bus bar, however only 46mm bars were available when I placed my order.  The simple solution was creating a larger “bend” in the center of the bar to reduce the center to center hole distance.  Given the laminated nature of the bus bars they are very easy to bend by hand, I simply clamped an aluminium bar to the desk using some bus bars as the perfect spacer.  The clamps are placed just outside the spacers to bend the aluminium bar up slightly in the center, making it easy to slide bus bars in and out for quick bending.

A fresh bar is bent “up” at a 45° angle, rotate and bend the other side up at a 45° angle as well.

Finally the bar is bent in the center against the square edge of the desk at ~90°.  The result is a new suitable center to center distance for use with the CA60FI.  Care must be taken to ensure the individual layers don’t move out of alignment too much to obstruct the use of one or both of the holes.

The final issue relates to the end to end connection of two CALB CA60FI’s.  I had purchased a metal punch when deciding how to make holes in the A123 tabs, it’s heavy duty and I didn’t end up using it on the A123’s but it was perfect for punching a new hole and slightly elongating the existing holes in the bus bars.  The bar was also flattened slightly to help increase the center distance.

A finished end to end connection.  The cells are shrink wrapped together to hold them in the desired cell groups for easy handling.  There are 10 groups of 7 and 2 groups of 2 cells making up the final configuration of the rear 74 cell pack.  The cells were assembled in order based on the bar code numbers the cells and documentation arrived with.  I will be able to easily match existing cell information with future data collected from the pack linking it to a specific cell.

In all of this there is one place where the 46mm bus bar actually works perfectly without modification.  When doing an end to side connection, IE transitioning from rows to columns or vice versa you need a (drum roll please) 46mm bus bar!  In my pack there are only two of these but perhaps it will let someone out there order the correct bars for every connection type and eliminate the type of modifications I had to make.

The bus bar sizes for CA60FI’s are as follows:
Side to side = 41mm
End to end = 51mm
Side to end = 46mm

The final layout of the pack is shown below, the negative most cell is the top right cell on the left block, the most positive cell is the top left on the right block.  This matches my layout plan for the A123’s making use of the existing 2/0 cables.  The bus bars not shown will be attached in the car, leaving simple to handle blocks to carry out to the car.  The bottom right cell in the left block and bottom left cell in the right block will be connected together through a fuse and maintenance switch.

This layout makes efficient use of the existing battery box, keeping the 10kW charger in the bottom center.  If I ever want  to add cells there is room in place of the charger to add 11 more cells, plus 19 in the front battery box would take me up to just beyond the limits of the Soliton1 at up to 104 cells.

The first half of the rear pack is installed, the charger needs to go in before I can install the 2nd half and connect everything together.

I’m finally putting the finishing touches on the 10kW charger.  The final panel has been milled/engraved, the wiring is almost complete.

There are still a lot of small tasks to get out of the way but the car is on track for being “complete enough” to resume daily driving early in the new year.

Speaking of new year, happy new year to all!




Leave a Comment more...

Installing the Front Battery Box

by on Dec.23, 2012, under Electric Car

The original plan was to have all of the batteries in the rear of the car in a single battery box.  The very compact A123’s would have made this possible.  The CALB CA60FI’s that will be going in the rear aren’t as compact.  The battery box that would have held 96S3P worth of A123’s will only hold 74 of the CALB 60’s.  Since my goals include a cell count in the 90’s I needed room for some more cells.  The “engine” compartment was never designed to hold a battery box, however there was a space in front of the motor that remained unused.  After some quick measurements I found enough space for the battery box that I built in the previous update.

Now it’s time to figure out how to mount it.  I raised the front of the car on jack stands so that the battery box can be installed from the bottom.  There is no room to fit it in from the top because of the hood latch and pop up headlight mechanism.  Because of some careful measuring before building the box it fit the space exactly, I did get lucky as to where the top of the box sits though.

The top edge of the box is basically level with the “frame rails” of the car, it’s a unibody car but there are distinct frame rails front and rear.  The box is also basically centered below the mounting position of the pop up headlight mechanism.  By inserting a plate between the frame and the mechanism it allows me to suspend the battery box from the frame rails very easily.

I sketched a bracket design on a post-it turned that into a CAD drawing and milled a pair of them.  The CNC router is one of those tools that I simply couldn’t live without now.

The bracket fits nicely under the pop up headlight mechanism using the factory hardware and mounting holes.

Both sides test fit, and everything lines up as it should.  The top of the battery box will get a plate with 3 studs on each side that will align with the 3 holes in the aluminium bracket.  Raising the box up from the bottom with some guidance should put the studs in the holes, with some lock washers and nuts the top will be secure.  I’ll paint the brackets black for the final install.

The bottom of the box aligns with the bottom of the old radiator mount so a small bracket between those two points will ensure the box can’t move.  Although the box is fairly far forward it sits very low and should help even out the front/rear weight balance.

The whole reason for the front battery box is to hold the displaced A123’s from the arrival of 72 CALB cells that recently arrived (plus two I already had).  The cells were well packed arriving in two small crates, from CALIB Power out of the USA.  Thanks to Don Blazer for being the retailer in this transaction and providing a very good price.  You can find Don on the Diyelectriccar.com forum here.

I just hit the half way point in my testing, much faster than the A123’s since there are far less individual cells to test.  I’m testing two at a time using a pair of PL6’s with similar settings to the A123’s except charge/discharge at 40A and a CC-CV discharge instead of just a CC.  Using the CV portion on the discharge is providing the bulk of the work in bottom balancing.  Shown below is one of the only good uses for Lead, I borrowed a group of gel’s that use to power Wade’s Solectria Force since his EV got the upgrade to 56 of the CALB 60’s recently.

I really like the form factor of these cells they are tall and slim giving them a small footprint and making the best use of a deeper battery box.  They also match the height of the A123’s very closely making them fairly interchangeable.

As shown in the above picture I am now using kapton tape on the edges of the A123 cells.  This should help eliminate the chance of the edges shorting which seems to be related to the electrolyte leakage I was experience in many cells previously.  The aluminium heat spreaders in the bottom of the battery boxes also has a very tough 3M film on the top to give another layer of protection.  Anyone who still wants to use A123’s must tape all of the edges of all of the cells, this is not a suggestion it’s a requirement.  Kapton tape is available on ebay or from some soldering supply places.

I created a simple gland plate for the high current positive and negative wires along with a multi pin connector for the heater wiring and power for the Cell Log 8 Breakout Module.  The studs that align with the mounting plates are also in place, the last step is creating a watertight lid to seal the box, I’m deciding between a simple aluminium plate or plexiglass lid.


Leave a Comment more...

Another Battery Box and some new Cells

by on Nov.25, 2012, under Electric Car

As with all things related to this conversion there is a change.  The battery system is moving from just A123’s to a group of A123’s and some CALB CA60FI’s, the rear battery box is actually the perfect size for the CALB’s and will house 74 of the 60Ah cells along with the EMW 10kW charger.  This means there is a new battery box that will be in the front of the car to hold the A123 cells that will remain.  The 3P groups have been reconfigured to 4P so that all of the groups exceed 60Ah, the lowest CALB cell will determine the pack capacity.  As 4P groups that will never be fully charged I am hoping they will be trouble free.  In an attempt to go easy on the CALB cells, the battery current will be turned down to 500A for normal driving, (meaning just over 8C peaks) and possibly as high as 750A in a “sport” setting (12.5C peaks) based on Jack R’s 12C testing of the 40Ah version of this cell.

The battery box is fairly compact at 24″ x 8″ x 12 1/4″ and will hold either 20 groups of 4P A123’s or 19 of the CALB 60Ah cells if I choose to swap out the A123’s in the future.  This box will fit between the frame rails in front of the motor but behind the AC condenser.  Due to the placement of the hood latch and pop up headlight mechanism this battery box will need to be removed or at least lowered to access the cells.  It will be raised up from the bottom and bolted in place.  This box will be insulated and heated the same as the rear box.  Due to the lack of access, a Cell Log 8 Breakout module will be installed for easy measurement of cell voltages.  The cell log 8’s won’t be on all the time and will only come on by pressing a momentary push button that turns them on and allows the voltages to be read.

As usual I start with a simple rendering, this also lets me generate a cut list for the steel that I need.

Front battery box

Then the actual frame with a layer of primer/paint.  This time around it was much easier, I learned a lot building the first box.

front battery box

Then on to the aluminium skin, it’s riveted on just like the last box, but that process also went much smoother.

The box, ready for some rubberized undercoating to protect and seal it.

Test fit with some A123’s and CALB’s to confirm I got the sizing right.  There is 1/2″ foam insulation as well as a resistive heater in the bottom of the box.

The other notable change is the switch to bottom balancing.  Given the hybrid pack of A123’s and CALB’s I would prefer to use the bottom portion of the A123’s capacity and limit how much they are charged.  Each of the cell groups measures 76-78Ah but they will never be charged over 61Ah (estimate of the lowest CALB 60Ah cell).  Since the pack will seldom be discharged near 100% the A123’s will effectively be used only in the middle 60% or so of their capacity.  Along with the lower battery current settings and the move to 4P from 3P the A123’s should have an easy life.  This does change the normal charging scheme slightly, since the A123’s will never reach 3.5v a new calculation is required for safe CV voltage.  During charging the A123’s are around 3.35v as they pass 60Ah so 20×3.35 + 74×3.5 = 326V or 3.468V per cell.  I will likely charge to 3.45v/cell as a safe starting point and see if the charger terminates as expected when the CALB’s are nearly full, I may need to raise/lower the final voltage accordingly.


Leave a Comment more...

EMW EV Dashboard, and turning up the Heat

by on Nov.20, 2012, under Electric Car

The EMW Deluxe EV Dashboard hardware has arrived, mine is apparently the first copy outside of EMW’s test cars so documentation and instructions were a bit lacking.  Valery did provide me with a copy of the PCB layout which has some pin markings on it which points to some basics like HV Gnd, mV, bV, mC, bC, bCC, t_mC, t_bC +5v.  With these markings and a growing understanding of how Valery designs his products I got the basics sorted out and wired up, some quick tests on the bench had it talking to the tablet and displaying battery voltage, discharge current, power and wh/mile.  It looks like Valery took note of Jacks criticism on the floppy hall sensor that was attached to the basic version and has now created a very simple but solid mounting for the new shunt like hall sensor blocks.  There are two different sensor blocks, one for the motor loop and one for the battery loop.  The app of course doesn’t yet do anything with motor voltage or current so I haven’t hooked that up yet.  The set up is the same as the basic version, there are two buttons that guide you through the set up and enter configuration values.  The core of the Deluxe hardware is an arduino like microcontroller with a breakout board on top with a group of eight db9 connectors, two of which are used for voltage/current/temperature for both motor and battery.  This leaves six more db9 connectors that are wired to various digital/analog inputs and outputs along with other functions that appear to be preconfigured on the board.  There is a trim pot for tweaking the displayed battery voltage when compared to a trusted volt meter.  A bluetooth module plugs into this board which does the micro to tablet communication.  The high voltage monitoring for battery and motor voltage are isolated using a very similar design as used in the 10kW charger.  The system looks like it will have an incredible potential for expansion and customizability when the right people get their hands on this hardware.  The controller code will apparently be released for end user tweaking which will make this a powerful starting point for EV instrumentation and control.

The board is lacking mounting holes and is likely still a work in progress, I will be machining an aluminium top plate with some engraved wording that will mount and protect the PCB’s.

Since the hardware requires constant power I’ll have to revisit my 12v system to handle the constant draw without causing problems or draining the 12v system.  Since the car itself has virtually no draw on the 12v system when the car is off I hadn’t planned for an always on dc/dc converter or anything like that.  I may use a battery charger/maintainer or other small power supply on the 12v system that’s on whenever the car is plugged in.  I’m also thinking about an automatic cut off system that turns off the 12v battery if the voltage drops too low.  I would lose my state of charge indication but easily be able to restart the car and SOC would reset on the next charge cycle.  Better than being stranded because my electric car has a dead 12v battery!


I had some issues with the heater, in the end it turned out to be caused by too much heat.  I’m using the very inexpensive Kats tank fluid heaters meant for 1500W at 120v, it’s just a resistor so it doesn’t care if it’s AC or DC power.  I had tested one at about 140v DC and it worked beautifully, I didn’t have the final reservoir so I tested it in a closed loop sealed system.  The final plan is an open system so that pressure can’t build up and burst.  As it turns out at 150V the Kats heater can produce an incredible amount of heat, as measured around 2300 watts.  The problem is in an open system the fluid will boil at a lower temperature than in a pressurized system.  The 140v test likely worked because it was sealed, the fluid didn’t boil and the pump was able to keep the fluid flowing.  Once I had it all hooked up in the car with the proper reservoir the first test failed, I burnt up the first heater in a matter of minutes.  The heater was boiling the fluid causing the pump to choke on the steam and stop moving fluid, once the fluid stops the heater element heats up, trips the overtemp switch however this switch doesn’t work well on DC, I don’t think it actually opened the way it is supposed to.  The element itself then burned up going open circuit.

The solution is a lower power version of the same heater, instead of the 1500W heater I installed an 850W heater, it’s the same size/shape but the element has a different resistance.  On 150V it produces about 1350W which doesn’t boil the coolant and things heat up and circulate as expected.  I’m now using a pair of these 850W heaters wired in parallel, with the fluid flowing in “series” which gives me about 2700W of heating power.  The system seems to be working well, but only a cold day will truly tell.  I have also revised how the temperature switch is wired in, it interrupts the 12v signal that turns on the contactor which it handles with ease.  I could probably use a pair of the 1000w version for a bit more heating power and still not boil the coolant, but I’ll see how these 850W units work when we get some cold weather.

As with all things in the car I strive for a stock appearance and function, at the moment the heater switch is simply a pair of wires and switch laying on the passenger side floor, however initially I will wire it up in place of the stock rear defrost switch.  This sits to the immediate left of the rest of the heater controls and is the most appropriate.  When I have more time I’ll test the behavior of the vacuum operated fluid valve that is part of the factory heating system, I may be able to install a microswitch on the vacuum valve that would turn on the heater system when the temperature dial is turned up above the cars ambient temperature and to a certain degree turn itself on and off as required based on the heat setting.  This would eliminate the need for a dedicated “heater” switch that isn’t part of the normal and intuitive operation of the car.  If I go that route I will have a hidden manual override switch that turns off the heater if there is ever such a need.  The other option is to put the EMW Dashboard in control of the heating system, Valery has already coded and assigned heating and cooling inputs/outputs designed to monitor temperature sensors in the battery boxes and cabin and turn on heating and cooling as required.  This would be a more complex but easily programmable system that would provide more flexibility and control to the frequency that the heating element turns on and off.  It would require some sort of control input to “set” the desired temperature however that might be able to be hijacked from the existing temperature control knob and be made to operate in a very intuitive way.

Well I haven’t even posted this update yet and plans have changed again.  Early on I looked at the RM3 and RM4 heaters sold by Metricmind, since that’s where I bought my vacuum pump.  At regular price they are very expensive so I ruled that out at the time.  As it turns out a batch of RM3 200-450V 4kw heaters with attached pump just came up on the surplus market, for a very reasonable $120US, buy a pair and that comes down to $100 each + shipping.  It should look like the one below except the orange connector and wire has been cut off. New the heater was over $400 and the pump was another couple hundred.


I will keep one of the Kats heaters in the loop for preheating the car off of 120v but the RM3 should nicely take care of heating the cabin while driving.


Leave a Comment more...

1000km later…..

by on Aug.29, 2012, under Electric Car

I officially have over 1000 EV Km’s on the car, it’s not finished, but it’s working beautifully.  I snapped the shot below just before going for a test drive, I’m now at 1006km.  The Soliton1 cooling system was a great addition and anyone using a Soliton 1 or Jr. should plan water cooling from the start.  I got a new AC switch and tested the air conditioning, it works very well when driving, since the motor stops when the car does so does the AC, I may have to visit the idea of a dedicated motor to run the AC system, but for now it provides plenty of cooling (which likely isn’t needed much more this year).  So the next important task is heating, I have another reservoir like I used for the Soliton1 cooling that will go along with my pair of 1500w fluid heaters and Liang D5 pump.  This will be an incredibly simple system I’m just looking at the idea of an insulated box to keep the cold outside air from steeling my heat, this would also make the system more efficient with most of the heat making it into the cabin.  I’m going to upgrade the dash lighting to LED’s at some point and while I’m in there I might try to “tweak” the odometer just a hair so that my 1006 EV Km’s reads as 61006 instead of the current random number.

I received the EMW deluxe hardware for the EV display, Valery is still working on it and the software, so it will take a bit before I can get enough info out of him to hook it up and start using it.  There are a pair of hall effect sensors, one for the motor loop and one of the battery loop.  The battery loop version has two sensors of different values I assume it’s for trying to keep the charging/discharging currents in the optimum range of the hall effect sensors.  I also got a replacement IGBT for the charger, I had an “issue”, I’ll elaborate more on that in the next charger update.

With my exposure on EVTV I met a local EV’er who’s in the process of upgrading his Solectria Force from Lead to Lithium, this is a shout out to Wade.  I’m not alone!

1 Comment more...

Battery Box Updates

by on Aug.26, 2012, under Electric Car

The battery box is one of those things that I’ve been working on slowly for many months.

Edit, I’ve had this post in the draft stage for so long I don’t remember my initial idea of what I wanted this post to say, but I thought I’d simply post it as I’ve had many inquiry’s about how I built the box and how much room there is around the transmission.


It started off as a rendering:

Then turned into a real box:

Then a test fit in the car:

Check the clearances:

Test fit the sides:

Add the final panels:

Most of the aluminium panels are covered in a white protective film that will be pealed off as they are being riveted into place.

The box is insulated with 1/2″ rigid foam and has a resistive heater built into the bottom that will let me heat the battery box in the winter.  I don’t know if I’ll need it but better to have it and not need it than to need it and not have it.

I currently have a 46S3P pack of A123 20Ah pouches along with a bunch of scrap foam taking up space in what was designed to hold a 96S3P pack.

I have a temporary charger in place while I finish building the EMW 10kW charger.  It takes a long time to charge, but it lets me drive the car every day and I couldn’t be happier with the performance which will only improve as I increase the pack voltage.

The box still needs a lid and completion of the batteries and wiring however it’s at a completely usable state and a pleasure to drive.

2 Comments more...

EVTV and Water Cooling the Soliton1

by on Aug.26, 2012, under Electric Car

I’ve been watching EVTV for years now, I don’t always agree with what Jack says but it’s still somewhat entertaining and usually informative even if you only learn what not to do from Jack’s experience.  Love him or hate him, Jack is doing great things for the EV community and his show EVTV is one of very few current information pipelines.  His shows tend to be long and boring to the non-EV type person however those of us who have done, are doing, or plan to do a conversion can relate to his stories, experiences and mishaps.  Jack had been playing with the A123 pouches for a very long time but never ended up with anything in a car, I had expected him to beat me to the finish and show everyone what the cells can do.  He’s busy doing other things and I’ve got my car on the road and I drive it every day so I thought I’d share my experience and build with his viewers.  My video is crappy and shakey, I should have narrated over or at least spoken during latter part of the drive, but I’ve had good feedback about simply sharing.

I’m confident that many more people find out about me from his show than would find out about his show from my blog, but if you haven’t already watched the August 17th episode of EVTV I would say grab a coffee and a comfy seat and enjoy the show.



The Soliton1 is an amazing controller, it’s powerful, feature rich and simply works.  I live up in Canada so our average summer temperatures here in Calgary are between 20-30C  (68-86F)with the odd day a bit warmer but seldom above 35C (95F).  Due to this fairly moderate temperature I thought I would get away with air cooling on the Soliton1, well it turns out that ambient temperature plays a part but doesn’t mean the Soliton1 will stay cool on it’s own.  I believe that this would apply to any conversion in almost any location.  The Evnetics staff as started to clarify that it’s true virtually every Soliton1 install requires water cooling, and a cooler controller will have a much longer life span.


There are a bunch of companies that have put together “water cooling” kits, some based on computer water cooling parts, and others using more industrial or automotive grade components.  There is a pretty big cost variance between the two, with the computer style starting in the $250-300 range and the automotive style in the $700-1000 range.  I wanted something slightly better than the basic computer stuff but less than the automotive style.  This meant putting together a better computer style cooling kit.


I had a few main goals:  Cost, Functionality, Silence, Appearance.

Starting with cost, I wanted a system that worked better than the low end kits, but every extra dollar was worthwhile in the system.

Functionality, It must outperform air cooling with only the pump running (no fans) and must be able to automatically turn on the fans for increased cooling when required.

Silence, I wanted a system that would work better than simple air cooling in virtual silence.

Appearance, It can’t look out of place and must contribute to the appearance in install of the system.

The system is optimized for high flow fan use, but still provide effective cooling with zero air flow.  I chose a fairly large reservoir to increase the thermal mass of coolant so that it takes longer to raise the temperature of the coolant.  I chose an inexpensive triple 120mm fan cooled radiator, this radiator is optimized for efficient high airflow cooling and I used three 120mm x 38mm fans.  These fans will run well between 24-48v with 200cfm (each) of airflow at 48v.  I’ll start with a 24v system and easily have the capability to run 24v, 29v, 35v, 39v or 48v.  The amount and effectiveness of cooling required will determine the final fan voltage.  The Soliton1 has a number of programmable outputs depending on the features you want to implement, control or monitor.  I’m currently using one output to drive the tachometer in the dash, I’m using another output to represent battery current by driving the stock oil pressure gauge and the final output is driving the stock temperature gauge to represent controller temperature.  Instead of “seeing” the controller temperature I am going to configure that output to turn on the radiator fans, this is controlled by the Soliton1 and corresponds to the temperature in which the Soliton1 would automatically turn on it’s own fans.

I have not yet reconfigured the Soliton1 to control the fans, I’m still watching the temperature of the controller on my stock dash gauge, the immediate trend is that temperature changes are much slower, and peak temperatures are much lower, I will use the logger program to collect some data about the Soliton1 and compare it to some earlier air cooled logs.

I was not able to find a single source for all of the required pieces to put together this system but I was able to narrow it down to two.  There is an online Canadian store called Dazmode.com that has a nice selection of computer water cooling components, fittings and tubing that let me put together most of the system.  The fittings required to connect to the Soliton1 came from an Ebay store along with a roll of thread sealing tape.  Service from Dazmode was very good, there was one small error in my order, I received all 3/8″ tubing and no 1/2″ tubing but this was quickly rectified through the RMA process on their site, the 1/2″ tubing is in transit right now.  You will see in the pictures the clear tubing was improvised and less than ideal but let me test the system while I wait for the proper black 1/2″ tubing to arrive.

The pump is mounted level with the bottom of the reservoir so that it is always primed simply with the force of gravity on the coolant in the tube.  I used the variable speed version of the common Laing D5 pump, I like the amount of flow on setting “3” but this could be tweaked to optimize cooling with some trial and error.

I used a combination of 3/8″ ID and 1/2″ ID tubing to connect to the various components, this was simply based on the best way to connect to each piece and not for any type of flow optimization or pressure.  The type of fittings required depend somewhat on the location where these items are mounted and where the tubing is to be routed but I will make a list of the components that I used and where I bought them to make it easier for others to put together a kit that meets their specific needs.

I mounted the reservoir in a way that gave me three fitting holes in the top which allowed me to use one for the input from the radiator which completes the cooling loop, the outlet is on the bottom so that gravity does a good portion of the work in keeping the pump primed.  The other two top holes make for easy implementation of a fill hole and vent hole.  I don’t want the system to pressurize as heat is built up I simply want it to raise the level of fluid in the reservoir but remain at normal air pressure.  The vent was easy to make out of the included clear plastic “LED” holder that is already drilled 90% of the way through, I just finished of the hole with a small drill bit which will let the pressure equalize.  The other fitting is a metal plug that makes a good plug for a fill port.  I used a small funnel and filled the system while letting the pump run, it quite quickly moved the fluid through the complete system and removed all of the trapped air bubbles.  I was going to make a video of the system in action but once I had the system filled up and all the bubbles removed you can’t even tell the system is on.  The input to the reservoir from the radiator goes through a 2″ long tube that is below the water level in the reservoir so there are no bubbles or even moving water visible and the system doesn’t make any noise so it didn’t make for good video.

I will be using the same technique for the heating system and have already ordered a similar but smaller reservoir to be using in the heating loop that will allow the same style fill hole, and vent hole to keep the system from pressurizing.



2 Comments more...

Charge Port and A Few Test Drives

by on Jul.19, 2012, under Electric Car

I’m still working steady on the battery box, including the wiring for the cell log 8 breakout modules.  The battery pack is not quite half done however it does make the car drivable, I haven’t finished the 10kW charger, however I’m using a small 600+ watt charger made from 3 small Mean Well 48v power supplies that I’ve modified with CC/CV adjust-ability.  It charges to 154v at 4A and terminates when the charge current drops below 0.75A.  To allow easy charging I’ve finally had to install the charge plug in the old fuel filler opening.

This was much simpler than expected using a piece of scrap aluminium I had laying around that only took a few minutes to modify to make it work.

I still need to seal a couple of extra holes and weatherproof the perimiter but it fits great and doesn’t look half bad either.

With the ability to recharge easily I’ve had to chance to do a few test drives, everything is working well and the car has a very nice feel and is fun to drive.  The tires squeal in first, but I start out in 2nd which gives nice acceleration and I don’t need to shift when driving around at up to 50km/h, any higher and I drive in 3rd.  As soon as I get the deluxe EV display hardware from Valery at EMW I’ll be driving it to work every day.  I’d rather not drive without voltage/current/fuel gauges to help me out.  There is still a lot of work to do on the car, but once it’s driving I can pick away at the important tasks one at a time.

I’m happy with the donor, I know I made the right choice, the car itself is fun to drive.  I still have some reinforcements to make to the battery box then I’ll test the car out in some corners and see how it feels.  There is a self weigh station not too far from my house so I’ll take it down one evening and see how far off I am on the weight as well as check the front/rear weight balance to see how close to stock I am.  I only have half the battery installed so I would expect it to be slightly front heavy but end up slightly heavier on the back when it’s done.

2 Comments more...

EMW 10kW Charger Kit

by on Jul.01, 2012, under Electric Car

When I started the project there were no chargers on the market that I was completely happy with.  There was the Manzanita Micro that was left over from the Lead Acid days, it will charge Lithium but isn’t optimized for it and it has a bad habit of changing the CV voltage when you adjust the current.  There are a bunch of Elcon type chargers and knock offs, Elcon being a knock off itself.  They work, but if you change your pack you ship it back to be reprogrammed.   That simply doesn’t work for me.  There are the high end chargers like Brusa, but that’s not in the budget.  There’s also the fact that virtually the only time LiFePO4 cells can be dangerous is if overcharged, this is where the dc/dc charging concept came from.  This system is safe and reliable but not optimized for efficiency or speed of charge.  What the market needed was a charger with user programmable outputs for voltage and current that will run from 120V or 240V without having to change any settings, one that will do accurate CC and CV stages and terminate at the right point.  It doesn’t seem like too much to ask now does it?  Well until recently it simply didn’t exist.


Valery at EMW has changed that, with the help of others he has developed an open source 10+kW charger platform that is available prebuilt, as a kit, or simply a BOM/Schematic and you are on your own.  I went with the kit version because it needs to fit in a very specific space within the battery box so I need to build the charger to this size requirement.  The version of the charger I chose has a PFC front end which means a couple of things, it doesn’t really care what the input voltage is, so it will run off of 120v or 240v just fine, as well as maximizing the power that can be drawn from a breaker (Power Factor Correction).  The PFC front end takes a small hit on efficiency but is required for the battery voltage that I need to charge so it’s not really a choice anyway.  By the time I was ready for the charger, and the charger was ready for me.  Meaning it had gone through a few revisions and has gotten to what I believe is a fairly mature product.  The electrical and mechanical design is quite nice making good use of 3D space, meaning for a 10-15kW charger it is very compact.  There are two cooling options, air and water, I went with air because I hadn’t planned on plumbing the battery box for coolant lines, however if I had started the design with this charger in mind I would have gone with water cooling.


I’m a visual person and I can put just about anything together if it has pictures, however Valery’s instructions are simply text so it took me a bit before I fully understood where things are meant to go and how they connect with each other.  I started with the basics soldering the PCB’s.

Then moved on to the high power components and their layout.  I drilled marker holes in the heat sink using the CNC router to give faster and more accurate results.

I revised a few of the assembly instructions, including the output diode wiring.  I drilled some copper blocks with the CNC router again, to directly solder 8awg wire to the blocks which bolt to the diode.

Inductor mounting is also revised, instead of the leads pointing up for use with flying wires to connect to the PCB, one connection from each inductor can be soldered directly to the PCB.

Mechanical mounting of the inductors is also revised and I created a c-channel beam to hold the inductors in place.  This beam also becomes a mounting point for the enclosure walls.

The layout is quite nice allowing for a small overall footprint.

The charger will face upwards in the battery box of the 944 behind the transmission between the two battery banks.  In normal use I can simply plug in the charger and it will just work, if I need to change settings the switches and display are front and center for easy adjustment.  I designed the front panel using V-Carve pro, it’s nothing fancy but simple and functional.

Progress of the panel after drilling marker holes.

After milling and engraving.

And finally after drilling, paint fill and switch mounting.  The display is simply sitting behind the panel just to see if all the characters are visible through the display hole.

A test fit on the heat sink to ensure everything is going according to plan.

The bottom panel / fan bracket was designed the same way.

With fewer holes and no engraving it was much simpler to produce.

Some 120mm x 38mm 48v fans were mounted for mock up only.  I will be using 120mm x 25mm 12v fans for the actual build.

Test fitting the panel to the rest of the enclosure along with the required power brick and finger guards.

The enclosure shape and size are starting to show, unlike the assembly instructions where the charger/heat sink is mounted inside a steel box, my version uses the heat sink as the main frame OF the box.

I’m in the process of designing the side panels which will be the legs and mounting points for the charger.  The left side is quite simple but the right side is also the mounting point for the main control board of the charger.  The input/output connections will be made on a terminal strip on the front of the charger (the top as the charger is viewed in the above pictures).  The external wiring for BMS input and J1772 compatibility will also be mounted on that panel.

EMW is creating some great products to fill some important holes in the market.  I will also be using the Deluxe version of the EV display tied to my Samsung tablet that I’ve mentioned previously.  EMW is also developing a controller, and although I love my Soliton1 there is potential for a higher powered street controller closer to the power level of the Zilla 2K HV/EHV but with configuration and controls possibly closer to that of the Solitons.  The Evnetics Shiva is amazing but 3000A and 425V is well above my requirements and budget for the Porsche, I would love something in the 1500-2000A range.



Leave a Comment more...

Looking for something?

Use the form below to search the site:

Still not finding what you're looking for? Drop a comment on a post or contact us so we can take care of it!