Recycled 18650 Bicycle Battery

In previous posts I’ve mentioned accumulating and testing used laptop batteries and other sources of 18650 size cells.  A project came up that makes good use of 91 of these cells (the range extender for the car may or may not still happen).  Many people have already done this, but the following is my method of assembling a 24v 26Ah bicycle battery specifically designed to fit in the existing case and charge using the existing charger.  The bike originally used a pair of 12v 12Ah lead acid batteries.

The first step was to layout the connection scheme for the 91 cells and create a drawing and then tool paths to drill and engrave the copper clad board which will be used for the bulk of the cell connections.

Bike battery - SketchUp Make_2015-04-08_14-59-26

Then drill and engrave the PCB material.


The layout of the PCB combined with the layout and orientation of the cells gives the desired series/parallel arrangement.


Tin the tops of the cells, take care not to heat the cells too much, others have said this can damage the cells.


I used resistor leads for the connections from cell to PCB, bend the lead as shown to make a good electrical and mechanical connection to the top of the cells.


Solder the lead in the very center of the cell.


Trim the lead, leaving roughly 1/2″ sticking up.  I found that each resistor would make 4 connections.  After all the cells are done, the PCB is placed overtop of the protruding leads, this takes careful alignment to make sure they all stick through, careful soldering in the previous step helps out a lot here.


Bend and solder the leads to the PCB making sure not to get too close to the engraved dividing lines.

Repeat on the other side with the appropriate PCB and as many modules are being assembled.  Due to the shape/size required I had to make 4 stacks of 5 cells to the side of the main blocks, these were simply soldered together with wire.


Complete the electrical connections (don’t forget to include a fuse!)


Connect the cell log 8 for monitoring and plug in the charger.


Once the pack passes basic tests use electrical tape or other insulating material to cover the exposed electrical connections.


For this application the pack will be balanced on the first charge and simply monitored and rebalanced as required.  If the pack won’t stay balanced using this method either a top balancing BMS or RC charger with balancing leads may be adapted for long term use.  The existing lead acid charger charges at 3A CC-CV to 28.82v which is 4.117v per cell on a 7S13P pack, this is low enough that minor cell imbalance over time shouldn’t be an issue.

DSC00501 sm

The whole system is a compact box with built in charger.  The original battery was a series pair of 12v 12Ah SLA UPS style batteries, at 50% discharge that’s about 6Ah if Peukert doesn’t decrease it even more.  The new battery is 25.9v nominal, giving a slightly higher average voltage during discharge and a capacity of 26Ah, though it shouldn’t be discharged 100% it can be done with slightly accelerated aging.  This gives 4.3x the Ah’s but with the higher voltage during discharge it’s about 5x the usable energy in a package that weights about 2/3 as much as the original lead.



The reconversion of a Porsche 914

The Porsche 914 is an iconic car from the 70’s that has stood up against the test of time.  It has great looks, great performance (not exactly powerful, but great handling).  Parts are available, but not always easy to find.  There have been countless engine transplants, and an abundance of EV conversions.  The car simply makes a great platform for customization.


This particular car was initially converted to electric drive using a Series DC motor, and lead acid batteries by the previous owner.  The car worked, but it never had very good range and was significantly overweight.  Later in life the car was then converted to LiFePO4 with the same motor/controller combo.  This made a significant difference, but range was still limited, and the DC motor less than ideal.


The next chapter in the 914’s life is a switch to a 3 phase AC induction motor and inverter combo made by Siemens along with the Lithium battery out of a Chevy Volt, two of them in fact totalling 30kwh.  With regenerative braking and a 30kwh battery, the cars range will be up around 200km on a charge.


Along with the EV reconversion, the car is getting a facelift.  Most of the rubber on the car has deteriorated, the tail lights and marker lights are faded/cracked, these will all be replaced with new parts.  The headlights will be upgraded to LED, this not only reduces power consumption but drastically increases the quality and quantity of light produced.  All side marker/brake/turn signal bulbs will also be upgraded to LED’s.  The brakes have already been upgraded, and the wheels are the lighter and more modern Porsche 911 style.


Some of the more mundane but important upgrades include a washer fluid pump, the original Porsche 914 used air from the spare tire to pressurize the washer fluid bottle and using a fluid valve in the steering column allowed the driver to spray washer fluid on the windscreen.  There are compatible parts from newer Porsches that allow the control of an electric pump which will simplify and modernize this oddly implemented feature.


The mis matched, broken and worn parts throughout the car will be replaced with original or better parts that will make this car equally at home as a daily driver and at a car show.  The original (and correct) door hardware shown on the left vs the compatible but incorrect parts on the right (passenger and drivers door respectively).

DSC00467sm DSC00468sm

The functional but incorrect Amps gauge in the instrument cluster will also be replaced with the correct “fuel” gauge found in the car originally.  The stock fuel gauge will be interfaced with the Ah counter to give the driver an indication of the remaining “fuel”.



To incorporate the necessary EV instrumentation into the 914, a 2″ round gauge made by EV Display will be located in the center console replacing the clock or volt meter commonly found there.  This can display Voltage/Current/Battery Temp/Amp Hours/State of Charge/Fuel Gage/Wattage/Watt Hours, well any two of these items at a given time.

EV Display V3 Kit medium (1)

To give modern functionality and a more informative display an Android tablet will be incorporated into the center console which in conjunction of the EV Display and Torque app will allow a customizable gauge cluster to be created in software.  Real time availability of motor power will allow the driver to choose the most efficient gear for cruising, unlike gas vehicles 5th gear may not be the most energy efficient highway gear.

minibms.mybigcommerce.comtemplatefilesUsing Torque Pro with EV Display

Having an Android tablet integrated into the car will also allow the use of GPS Navigation, Audio apps, and while parked even let you watch movies or play games.  The audio output will be routed through the indash stereo.


The Porsche 914 and it’s VW cousins of the era were never known for good heat.  The air cooled engines simply didn’t have an efficient method of collecting the waste heat and getting it into the cabin.  Many people augmented the stock heater with a VW gas heater which was a factory option in some countries.  Now it may seem counterintuitive to put a gas heater into an electric car, however there is one thing that gas is good at; making heat.  The propulsion of an automobile has a total efficiency in the teens for an average vehicle up into the 20’s for a very fuel efficient vehicle.  On the other hand home furnaces can reach into the upper 90% efficiency range burning natural gas or similar fuels, we typically don’t heat our homes with electricity for a reason.  The range impact on an electric vehicle from the heater can easily be 25% or even more in very cold weather.  The other problem with cold weather is most batteries don’t perform as well in the cold to begin with.  If the car is limited to 75% of it’s range just because it’s cold, then we take another 25% in order to heat the cabin we’ve significantly reduced the usable range of the vehicle.  The VW gas heaters are VERY fuel efficient, one of the most common is rated at 0.32 – 0.38 litres per hour with constant usage and in most cases the car is going to get too hot with constant use, so running the heater intermittently for 1/2 – 2/3 of the drive will be more comfortable and efficient.  Assuming a speed of 100km/h (typical of Deerfoot or Stony Trail here in Calgary) the car would burn 0.16 – 0.25 Litres of regular gasoline for a 100km drive with intermittent use (16 to 25 cents at today’s prices) the electricity cost to drive that 100km would be roughly $1.20.


Above you can see the vintage heater almost ready for testing before it gets torn down for cleanup and repainting.


The 914 is a very popular car for conversions partially because of  the large amount of space that can be used for batteries.  The front trunk, rear trunk, engine bay and fuel tank location are all suitable spaces for mounting EV components.  In this case, part of the front trunk will be used for batteries, holding just over 1/3 of the pack.  The Chevy Volt battery is liquid heated and cooled, this is a very efficient way to control temperature and something that will be maintained in the 914.  The liquid heater from the Volt will be reused and capable of bringing the battery pack up to safe charging temperature or simply warming the battery on a cold morning using wall power instead of battery power for better range and power.  The cars cabin may be included in this loop to preheat both the car and the battery to further reduce the need for gas or electric heat while driving.


Above the test fit and temperature monitoring of a small portion of the Chevy Volt battery pack.  The Volt battery comes as a complete unit configured for a very specific purpose as far as electrical and cooling connections are concerned.  Due to the size and shape of the 914, not all of the stock modules will fit the car as is.  In order to reconfigure the modules some extra parts are required, it’s not going to be as easy as dropping by the local Chevy dealership and requesting a very specific battery cooling part, the battery is likely not meant to be serviced outside the factory, only replaced.  While very specific the parts aren’t very complicated, simple plastic pieces needed to plug unnecessary cooling ports to start.  These parts were 3d printed using ABS material which has a suitable temperature range and chemical stability to work in the heating/cooling system.


The stock black gasket is on the left, with a 3d printed flexible version in white at the top, this white version won’t be used unless a replacement is required.  The center and right show the two types of blanking plates required to block off the cooling ports on one end of a battery module, one holds a gasket, the other seals up against the gasket in the last battery plate.  The 3d printed parts have studs incorporated into them similar to the stock pieces and are meant to be melted once installed to hold them in place.  The existing cooling ports and end plates are all interchangeable and can be configured as desired, I simply found I didn’t have enough blanking plates to seal off the ends of a few of the module groups in my planned configuration.  Unlike the Volt where the first two battery assemblies are in a row then feeding the final module through the “side” ports I found the need to have some of the assemblies side by side with cooling ports on one end only.




The Chevy Volt battery can be easily broken down into its 1kwh and 2kwh building blocks, 6 and 12 cell respectively then rearranged and bolted back together in the shape/size you desire just like Lego.  Changing the total length of a module does require different bolts to put it back together, but M6 x 1 threaded rod works perfectly for this.  Making a module shorter can be done two ways, either replacing the existing bolts with M6 threaded rod or cutting the existing bolts and threading the end with an M6 x 1 die (The stock bolts are only threaded about 1 inch on the end).  Remember you need four threaded rods per module.  M6 x 1 is commonly available in 1m lengths which is slightly longer than what is required to make a 4x 2kwh block using the “side” cooling ports.



Reassembly (compression) using a ratchet strap makes it easy to get the bolts started and using the built in wire guides and a 1″ strap keeps everything in place with minimal chance of the strap sliding off or out of place.  For longer modules like the one shown, using two ratchet straps, with a ratchet on each side make it easier to evenly compress the modules and get the nuts on the threaded rods started.

DSC00474sm DSC00476sm

Steel or Nylon straping used for banding skids for shipping should work well to replace the stock steel(?) band around the top of the battery assembly.  Nylon is probably cheaper/easier/safer for use around a battery and doesn’t require fancy tools to install.


In the name of recycling even parts of the battery assembly base can be reused to make new battery mounts and reuse the existing hardware and mounting method.  First find all of the somewhat camouflaged spot welds and drill them out with a 15/64th’s drill bit (or similar) being careful to get as close to center as possible.  Then enlarge the holes with an 11/32nd’s (or similar) drill bit to completely remove the spot weld.  This is slightly larger than the spot weld itself allowing for being slightly off center.



After all of the spot welds in a given section are drilled out the clamping portion of the base can be pried off.  There was some sort of adhesive in the bonding of these two pieces but tapping flat screwdrivers or other long wedges between the two pieces will separate them nicely.



The liberated plate now has the required mounting lip and studs to reuse the original clamping bars.  This plate can be welded to the bottom of a battery box or directly to the vehicle, I would recommend welding all of the previously drilled out spot weld holes to make a nice solid mount, an adhesive like liquid nails probably wouldn’t hurt, but might make removing it even more difficult than from the Chevy plate.


The finished product will be a smooth and quiet easy to drive 100kw, 200km convertible ready for the daily commute or the open road.  The combination of small lightweight car and a fairly large battery pack (30kwh) will give this Porsche 914 more range than almost all of the OEM EV’s available today (with the exception of Tesla).  Charging has not been decided yet, however a 10-12kw solution would extend the total daily usable distance by a significant margin allowing the car to charge from 20% to 80% in about an hour and a half.










18650 Li-ion range extender

For daily use my pack  provides more than enough range for what I need and want the car to do.  However there are situations where it would be nice if the car had 2 or 3 times the current range.  The Tesla Roadster and Model S have the highest range in the EV industry for production cars and they use massive packs of 18650 cells.  They benefit from the economy of scale  of the laptop and power tool industry as these are the most common cells used.  In the coming years this may reverse and the laptop and power tool industries may benefit from high quality cells developed for the EV industry.


Jehu Garcia is currently working towards an 18650 based pack for his VW Samba Bus using recycled laptop batteries.  I already have a few hundred 18650 cells on hand so I thought it just might be a worthwhile project to build a removable high capacity battery.  Although I have a few hundred cells, what I need to make this work is a few thousand cells.  If any of you guys have a stash of “dead” or unused laptop batteries and just want to get rid of them please feel free to send them my way to speed up this project.

The IT department at my day job was happy to give me the dead laptop packs since there wasn’t an established recycling process in place yet, I’m willing to bet the same exists in many companies large and small.

To put all of these cells together there are some inexpensive plastic holders available from

18650 Battery Holder


The initial plan is to use small  fuse wires just like Jehu, and I may use a copper clad board as the current collector, it provides stability and is easy to solder to.  The currents won’t be that high and with a massive width of copper sheet on the PCB it  actually has a fairly high current carrying capacity.

2014-03-22 13.24


The main purpose of this pack is not to drive the car on it’s own, but to supplement the existing pack in the car for longer range so the maximum current output doesn’t have to be as high.  I would like to build a pack with between 60 and 100Ah at ~86 cells which is a  voltage that would match quite closely the nominal and charged voltage of the LiFePO4 main pack in the car.  Assuming roughly 2Ah per cell on average from used packs I will need 30-50 cells per module and 86 of those in series, totalling 2580 to 4300 cells for the desired capacity.  This  pack would weight around 275 to 450lbs which although significant is worthwhile for 2 – 2.7x the range.

A New Direction

From the start of this build there have been countless changes, and with the exception of the donor car every single component has been revised/updated/eliminated due to testing/learning or simply something new becoming available that just didn’t exist earlier in the build.

I absolutely love the Warp 11HV and Soliton1 together, the torque is incredible and the car does 0-60 in 4.8 seconds.  I absolutely hate the brushes and brush dust in the 11HV.  (And the fact my 11HV requires repair/cleaning)   For this reason I finally have to agree with the OEM’s that some type of brushless motor is required for a daily driven, low maintenance EV.  The series DC motors still have a place right now simply because there are no high power AC systems available to the DIYer at a reasonable price.

When Azure Dynamics went bankrupt last year I picked up a DMOC 645 AC Inverter as well as an AC24LS AC Induction motor.  I never really had a plan for the set, however the AC24LS would never be able to give me the performance I’m looking for in the 944 without burning it up.  The motor that Azure Dynamics paired with the DMOC 645 was the Siemens 1PV5135 4WS14, an abnormally low voltage version of the typical Siemens AC Induction traction motor.  This motor should give me decent performance for daily driving.  I won’t have the 290-300ft-lbs of torque that I got from the 11HV but a more modest 220ft-lbs is still well above what the stock engine could produce.  The Siemens motor also rev’s much higher, there is good power out to 7500rpm, and usable power all the way to 10,000rpm.  I’m not sure how well the stock 944 Turbo transmission would like 10,000rpm on the input shaft, so limiting it to 7500 should keep things safe.

A new motor means a new adapter plate and coupler.  As it turns out the length of the existing adapter plate is pretty close to ideal, I’m going to add a 1/2″ aluminium plate that will adapt from the 11HV bolt pattern to the Siemens bolt pattern.  The Siemens motor is slightly shorter than the Warp so the extra 1/2″ does no harm.  The coupler will be much simpler, smaller and cheaper.  There is a steel coupler available that normally mates either the Siemens motor or a Remy motor to the Borg Warner single speed gearbox used by Azure Dynamics (and others).  This coupler is only $72US, mine is in transit but apparently it’s a good quality coupler.  This is only half of the solution though, I can now mate to the splines on the Siemens motor, however the 944 input shaft uses a different splined shaft.  For this I am using the center portion of a stock 944 Turbo clutch disc.

AC copuler


After grinding off the flange that normally attaches to the rest of the clutch I’m left with this small splined piece of steel.

The Siemens side of things is similar though a bit longer.  (image from the Vaxo website, click HERE to see more info from Vaxo)



The Siemens coupler is 40mm in diameter, the 944 clutch center is also 40mm in diameter.  I still need a pilot bearing, or to be more precise a piece to align the pilot shaft on the 944 input shaft, it doesn’t need to turn as both shafts will be permanently connected.  I don’t have a lathe, however I’ve been able to get good results using my CNC mill to make round parts.

siemens to 944a


Above is the aluminium piece that I milled from flat scrap.  It has a 15.15mm ID hole for the pilot shaft to align.  It has a stepped OD, one size matching the ID of the clutch center and the other matching the ID of the Siemens coupler.  This piece will be used for alignment of the steel sections during welding and remain inside as the pilot “bearing”.  It will also serve as a stop so that the coupler can not work it’s way off of the Siemens motor and along the splined section of the 944 input shaft.

siemens to 944c


The two pieces fit together nicely, it’s not quite as precise as could be done on a lathe, it would probably be better if there was a slight interference fit.  One day I’ll get a mini lathe to go along with the CNC mill.


siemens to 944d


The DMOC 645 also has some challenges, it’s CAN bus controlled meaning you can’t simply attach a throttle and other things to it directly.  There are a number of VCU’s in development from various people.  My favorite is in the final stages of development by Wolftronix.  This is a guy who is very familiar with both electric vehicles as well as Azure Dynamics/Solectria and because of this I believe his product will be the best first generation product out there.  Others will play catch up and likely incorporate some of  the features that come out in his first product.

DMOC adapter

The above image is from the Wolftronix website and I suggest you check it out for more information.

The Siemens motor is also on route, it was purchased from “Steve” thanks Steve!!!  This picture is from his for sale listing and may or may not be the actual motor I’m getting, he had purchased a lot (10?) motors from the Azure Dynamics auction.



If I didn’t already have a DMOC I likely would have just purchased the pair from Jack at EVTV, I think his combo price is quite fair considering he includes the required cables/connectors to connect the motor/controller as well as the main IO for the DMOC itself.  I would also suspect the connector for the GEVCU will be included when that is available as well.  Jack is also working with various groups on their versions of the GEVCU, I’m sure they will all catch up eventually but the feature set in the first generation Wolftronix is what I believe to be the best solution right now.  Jacks price on a single item is less attractive and I’m not sure if he would include any cables when only buying one piece.



In some respects this change could be considered an “upgrade” I’ll have a water cooled maintenance free motor with regenerative braking.  Those are all significant pluses!  There is however a negative or “downgrade”.  The peak power of this system is rated at 118kw, the peak power of my Warp 11HV/Soliton1 right now is around 200kw, and could be higher when the remaining CALB cells are installed.  I have no doubt that the system will still provide an exceptional driving experience, but it won’t be a sub 5 second 0-60 car any longer.


I have more cells to install anyway, so range goes up, with the regenerative braking I will probably get a tiny bit more range.  I live on a bit of a hill so I should be able to recapture some energy on the way down instead of heating up my brakes.  More than anything, my peak power will be lower, which to a certain extent limits the amount of energy I can burn accelerating away from a stop sign or light.  With all of these factors combined I expect my Wh/mile to drop from 300-325 down to 275-300, not a significant difference but still a difference.  The DMOC also has a higher input voltage range, instead of 350v max like the Soliton1 it has a 450v max rating.  It’s nominal cell count is 105 where the Soliton1’s max cell count is 100.

I will have 100 cells once I add in my remaining CALB’s, I also have a few more A123’s that could be installed if I have the room.  Each cell adds around 1km of driving range and I could easily go up to 110 or 115 cells providing I have the space to mount them efficiently, it’s never been an option so I’ve never actually looked into it.

Another small but interesting fact is my 944 will once again have a water cooled german motor.

It may be one of the hottest days of the year but I have heat!!

I’ve had a bunch of posts over the years talking about heaters, some of them were planning, some were installing, some were testing, some were blowing up heaters.  I’ve finally planned, installed and tested what I hope to be the final version of my 944 heating system.  Surplus components from Think EV’s when they switched from the Mes-Dea RM3 fluid heaters to a ceramic or other PTC heater.  The combo includes a 200-450v RM3 heater along with a high quality bosch pump.  During the removal at Think the HV power wires had been trimmed rather short.  These wires were replaced in a previous post, ready for installation.

4kw heater 3


New wires above, heater installed below.


Like most other systems in the car, it’s very important to me to retain the stock look/feel/function of the original Porsche system.  I’ve talked about it before, but a big thing for me was to NOT have a dedicated non factory switch to turn on the heater.  Some people have an aftermarket switch with some sort of label saying “heater”, this just doesn’t do it for me.  In the stock heating system there is a temperature dial, it’s blue on one end, red on the other, and has numerical values in between.  The most intuitive method of turning on the heat would be to turn this knob from blue to red or some value between, we can all figure this out if we get into a new vehicle that we’ve never driven before.  Why should it be any different in my EV?  In addition to simple operation I also wanted to maintain the “automatic” climate controls that Porsche was using in 1986.  It’s a simple system, it’s mainly vacuum operated based on some simple analog electronics.  To make this happen I installed two microswitches, one in the heater control itself, it’s basically a NC (normally closed) switch that only activates (opens) in the cold position on the temperature dial.  The second switch was installed on the vacuum operated coolant valve that would normally restrict the flow of coolant from the engine.  I don’t actually use this valve to restrict coolant flow, but the microswitch serves to turn off the heater.



The vacuum valve with added micro switch, and installed view below.


vacuum switch

I haven’t found a good program to create schematics but the system is quite simple:

Switched ignition +12v is connected to the NC microswitch on the temperature dial.  In any position above blue (cold) this switch applies power to a 12v relay which turns on the pump that circulates roughly 2L of coolant through the 4kW heater and stock heater core.  The same 12v out of the relay that powers the pump then goes to the 2nd microswitch connected to the vacuum valve that would normally limit coolant flow.  If the cabin temperature is lower than the temperature setting on the dial the valve is open and the NC microswitch applies power to the 12v terminal on the heater.  If the cabin temperature rises above the preset value the vacuum valve closes turning off the switch and removing 12v power to the heater.  The heater itself also has built in thermal limits that attempts to heat the coolant to 70 degrees celsius.  Any of these various systems will turn off the heater which should reduce the amount of power wasted creating excess heat and allow me to dial in the amount of heat I need or want while driving.

coolant tank

With a 4000watt heater the draw on the main pack is 12-13A while it’s “on”.  Only time will tell as to what kind of duty cycle I will end up with as the portion of time the heater is on vs off.  This duty cycle which will vary both with ambient temperature and the desired amount of heat and will ultimately determine the impact it has on my range.

I was going to make a short video on the operation and the various systems that control the heater however the ambient temperature this afternoon is too hot for the heater to even turn on.  There will be plenty of chances to do so in the coming months as the temperatures drop.  We have had some chilly mornings so time permitting I will demonstrate the system in action soon.

I still need to create some sort of indicator to make sure I know the heater is “on” and don’t have it on needlessly wasting power if it’s not required.


A weeks work on a lot of little (and big) things

I took a week of vacation to get some dedicated time to work on the car, I stayed busy the whole time and ticked a lot of check boxes off the to do list.  I also tackled a few things that weren’t on the list and didn’t imagine doing, but they are done and the car is better for it.

The week started with the manual steering rack, it turned out being a very simple and straight forward swap with no issues.  Since I had to remove the leaky rack I was up to my elbows in oily sludge which brought to my attention I was never able to pressure wash the motor bay before the install began years ago.

manual rack


The other very significant difference between removing the rack from the ’83 gas powered 944 and installing it on my ’86 electric 944 is the access to all of the parts, I had plenty of room to work with easy access to where I needed to bolt things up vs the incredibly tight dirty and awkward process in the gas car.

Since the ’83 is now just a parts car I’ve started looking at pieces that I replaced to get it on the road for my wife, as well as parts that are still in great shape that can either be sold or kept as spares for my own use.  Since I had recently replaced the front strut inserts on the ’83 and mine was in need I swapped the entire strut assemblies to the EV.  They are missing a couple of small clips that hold the brake wear sensor wiring but a zip tie took care of that.  The other good thing is that if I decide to buy performance struts and an adjustable spring kit I can install it on the original strut assemblies and then install the completed units back on the EV which could reduce down time.  (See how old this update is… I already have the new Koni suspension installed)

After solving the dc/dc converter fan noise issue, the next “loudest” item in the car is the vacuum pump for the brake booster and hvac vents.  To solve this (or at least reduce it’s annoyance factor) I milled a new mounting bracket that made use of the included rubber mounts on the vacuum pump as well as one of the unused motor mounts still in the car.

vacuum pump


This new mount has significantly reduced both the vibration and noise created by the pump, but not eliminated it.

One of the big things on the to do list was installing the heater.  Initially I had hoped to install it in an insulated box however the shape of the space I had to work with along with reduced serviceability forced me to reconsider.  Instead the heater is mounted out in the open on some simple brackets but I should be able to insulate the hoses/reservoir to hold in some heat.


Once again I was going to take some pictures to complete this post, but now it’s getting old.


The heater is completely wired and integrated with the stock controls.  I would like to make some sort of indicator light to show that the heater is “on” however I’m not sure where I would like that or how it should look.  Ideally it would be an indicator on the tablet and not require drilling a hole in the dash for an LED or something.

I also decided on the Momo Tuner 350mm steering wheel, again pictures needed!  The wheel is awesome and I found it for a very good price.  I was able to take the steering wheel adapter off of the ’83 and modify it to work with the late steering column.  The “tab” that kicks off the turn signal when the wheel returns to center is too short on the early version, otherwise it’s exactly the same.  I used epoxy and a metal bar to extend it far enough to work properly again.


At the moment I’m driving the Infiniti again, (not by choice) I had a “Desaturation fault” in the controller, so it’s back at Evnetics getting repaired.  The people at Evnetics have been incredibly helpful and quick to respond to the problem.  I received an email back within 15 minutes on a Saturday when this happened.  More info to come.


Gear ratio’s and a possible transmission swap

When starting an electric conversion it’s tough to know exactly what gear ratio will perform the best and feel the best when driving.  I was happy that my car had the turbo transmission installed since it would handle the extra torque from the 11″ motor, however I really didn’t know what the ideal gear ratio would be, but having a free box with 5 to choose from is a good start.

After driving the car for awhile on 300v I’ve come to the conclusion that my ideal ratio wasn’t included.  The car does drive beautifully and it’s quite quick, however I wouldn’t mind a few slightly “shorter” gears to choose from.

First gear is too short (too high a ratio) and 2nd is pretty close to ideal but could be a bit shorter for low speed driving and fun acceleration.  3rd gear is also close to ideal but is a bit tall for a 0-60mph run or spirited but not high speed driving.  What I feel might be ideal is a 2.5 gear, with a ratio about halfway in between my existing 2nd and 3rd gear.  As it turns out Porsche did make that transmission, it’s has the tough internals like the turbo transmission to handle the hp and torque that I have, but has my desired 2.5 gear.  It came in the later 944’s namely the 944S and S2.  It also has a 2nd gear that’s shorter than my 2nd gear and would provide slightly better acceleration without being too short like my existing 1st gear which is too far in that extreme.

944 Turbo
1st – 11.81    34.5mph @ 5500rpm
2nd – 6.949 58.5mph @ 5500rpm
3rd – 4.725 60mph @ 3830rpm
4th – 3.490 60mph @ 2830rpm
5th – 2.798 60mph @ 2268rpm (top speed 5500rpm 145.5mph)

As you can see from the above speed vs rpm vs gear ratio my existing 2nd gear is almost perfect for a 0-60 run, of course I would accelerate through 60 for the best time which puts the motor slightly above the recommended maximum RPM.  So for a safe 0-60 acceleration I need to use 3rd gear, which is a bit tall.

944 S2

1st – 13.56    30mph @ 5500rpm
2nd – 7.979 51mph @ 5500rpm
3rd – 5.425  60mph @ 4400rpm
4th – 4.007 60mph @ 3250rpm
5th – 3.015 60mph @ 2445rpm (top speed 5500rpm 135mph)


The S2 info is above, and as you can see a 0-60 run in 3rd gear would take me to 4400rpm which happens to be around where the power starts to taper off anyway.  4th gear would also provide a nice RPM for highway cruising keeping the fan speed fairly high for good cooling.

I have many other things to work on first, but I’ll keep my eyes open for a good used S2 transmission for a future swap.
In the mean time back to suspension and motor updates!

The Weights are in and some Suspension Decisions

Early on the goal was to have a conversion that came in at the stock curb weight with similar front/rear weight distribution.  When I got to the point when I knew the A123’s weren’t going to be the solution I knew the car was going to get a little rear heavy based on the new battery choice.  Once I switched to CALB CA60FI’s I estimated the final weight at 2900lbs +/- 100lbs, well it looks like I just squeaked inside that estimate.  In it’s current state the car weighs in at 2992lbs with a F/R weight distribution of 43.5% Front and 56.5% Rear, this is much more rear biased than the original 50.7% Front / 49.3% Rear gasoline powered 944.  It’s still far from the ~39%F/61%R of the 911, and lands quite close to the new Porsche Cayman which is 44%F/56%R.  I think it’s a good number for now, but I may try to move some weight forward, I’m working on a new water cooled charger, and a water cooled dc/dc converter.  The new dc/dc converter will be completely sealed and go up under the hood (vs the rear side panel) and share the cooling loop with the Soliton1.  If I can make the charger sealed enough it will also go up front, but the required fan cooling for the inductors might mean it has to stay in the back.

near final weight

Above is the current appearance, you can see the front ride height isn’t bad, but the rear is a bit low.


Above is a pre-conversion picture showing the front slightly lower and the rear higher (but still not quite level).  The final goal is to have both the front and rear level at a height resembling the pre-conversion stance.

The options:

Reindex the torsion bars and tweak the final ride height with the built in adjuster.  This would likely correct the appearance, but wouldn’t address the fact the rear of the car has gained roughly 300lbs.  I also need to address the front end which has lost just over 100lbs.

Supplement the torsion bar system with rear coilovers, this would increase the load carrying capability while providing ride height adjustability.
Coilovers on the front would allow me to level the car and increase performance.

Piece together an improvised coilover like system with add on adjustable spring perches.

Replace the torsion bars with higher spring rate versions to carry the load, while setting the appropriate ride height.

Initially the torsion bar looks like the most cost effective solution (simply reindexing the torsion bar may not be a solution, when considering the extra 300lbs), however I believe the front and rear shocks likely need to be replaced as well, and since I’m there replacing the front springs with a 1″ lowering spring could be worthwhile.



Well as often happens with my posts, I don’t get around to finishing and publishing a blog post until things have changed.

I’ve already adjusted the ride height and I’ve purchased some larger torsion bars, and now I have a week off to play with the car, I’d like to get the new torsion bars in and the rear height adjusted to provide a suitable stance.  I’ve started to tear apart the ’83 944 which I bought strictly for the manual steering rack, but then replaced the engine and other stuff for the wife to drive it.  I had bought new strut inserts for that car so I’m simply going to swap them over to the EV.  They aren’t high performance struts, however they are basically new, and should let me evaluate the front spring rate better since the existing struts are starting to “bounce”.  This will help me decide what type of spring to put up front, or if I need to replace them at all.

I finally installed the manual steering rack today and it’s simply awesome!  A few short drives confirmed it’s got the feel and lighter steering weight that I’m looking for.  The stock steering wheel didn’t feel big enough to provide the required leverage on the depowered power steering rack, however it feels huge now and I could use a smaller wheel.  I’ve heard great things about the Porsche 968 Club Sport steering wheel, it’s a bit expensive so perhaps one day, although a nice Momo would actually cost less.



I’m a fan of the Momo Tuner wheel which is similar to the Club Sport wheel above.


momo wheel


I’ve also done my best to level the car using the existing adjusters, and as it turns out it was lowered as far as it would go, so there was quite a bit of travel to bring it up.

near final weight ride height adjust


As it sits now before I change the torsion bars above, and to compare the pre adjusted level below.

near final weight


I’m happy to be at a point where car stuff is running in parallel with finishing the electric portion of this build.  The car itself is wonderful to drive and getting better every day as I address some of the normal car stuff, including the steering rack and suspension.

Building another EMW 10kw Charger

I’ve already built an air cooled version of the EMW and it’s working beautifully in my car. I was going to build a 2nd air cooled version for a friend, but because the kit was different enough I couldn’t use the same CAD files and panel designs for the 2nd charger so it was on to plan B.


Since I had to start from scratch (or at least revise all of the panels) I told my friend he was getting my air cooled version and I’m going to build a water cooled one.

I bought an 8″ x 8″ x 0.75″ aluminium plate for another purpose, however it’s just about ideal for the EMW.

Test fitting the silicone gasket, I’m using a 220mm x 3mm gasket in a 1.9mm deep by 3.5mm wide channel.

cooling block 1

Drilling and tapping the G1/4 threads, since the rest of my cooling hardware is computer stuff, standard fittings will thread in.

cooling plate3

And a test fit of a 1/2″ barb fitting. I may go with 90 degree fittings, depends on the final mounting location in the car.

cooling block2

Both barbs and the silicone gasket. I still have tons of holes to drill and tap on the top and bottom as well as around the edges. The asymmetrical design of the cooling channel is to leave enough room for the mounting holes on the other side for the IGBT’s, diode, input bridges etc.

cooling block 5

Test fitting everything on the block.

cooling plate plus emw2

The final dimensions including mounting feet will be:
10″ x 8.3″ x 5.5″ This doesn’t include the barb’s or cooling hoses that will run out the side.

cooling plate plus emw

Daily Driving and an Excellent Car Show

With the motivation from The Future of Transportation Symposium and Car Show I was able to get the car in a complete enough, tidy, safe and drivable state.  So being the first day it was above zero I drove it to work.  The JLD404 wasn’t quite set up correctly, it was a decimal point off on the Current and AH’s, but it still gave me an idea of energy usage.  Since I don’t have the EMW deluxe dashboard hardware set up again I’m using the JLD404 as my only instrumentation.  So to implement the auto reset I’m simply using one of the built in relays and a voltage alarm.  My pack charges to about 317v (92 x 3.45) so I have set an alarm that when the voltage reaches 317v which is my full charge and the charger terminates very quickly after that point.  This will likely give me a small error since it will continue to charge after reaching 317v, to increase accuracy I could set a 2nd alarm for a low current level and put the relay contacts for the two alarms in series.  That would mean it would reset the JLD404 when current has dropped to zero AND the voltage is at or above 317v, the only time this happens is at the end of charge.  I’ll start with just the voltage alarm and reserve the other for an AH warning or disable function.  If this proves unreliable I will revisit using the end of charge signal from the charger.

Since it takes awhile for me to put these blog posts together I actually have had time to run with the auto reset on the JLD404 using only the built in voltage alarm.  Typically I have a reading of -0.01 to -0.03Ah on the JLD when I come back to the car, this is actually far more accurate than I had expected and should work very well long term.  I may have to see how this plays out at higher and lower temperatures, but it’s currently working perfectly and I am as close to 0Ah as I would hope.

I had my first “break down” this past Friday, but luckily it happened while I was at work, which gave me access to the resources required to patch the problem and allow me to drive home and fix the car.  This will mean more to the fellow 944 owners, the rear side panels behind the wheel wells have tubing and drains that allow water to flow through and out of the car.  I spent a lot of time ensuring that all of these were clear and the tubing was in good shape and connected correctly on the passenger side of the car.  The previous owner had a CD changer installed in the drivers side cubby, and for whatever reason I had assumed that the drains were working because the changer still worked.  I installed my DC/DC converter in this cubby and as you might be guessing the main drain was plugged.  We had a few days of heavy rain and that side panel filled up with water, well not exactly full but there was 4-5″ of water in it.  This meant the bottom 1 – 1.5″ of the DC/DC converter was actually under water.  This included the fan and the 300v output rail of the PFC section of the power supply.  When I turned on the car there was a very loud POP and the car didn’t turn on.  The pack voltage was fine but the 12v dash gauge didn’t read anything.  I quickly turned off the car, opened the hatch and turned off the maintenance switch.  I confirmed the pack voltage which was fine, but the 12v system read about 3v.  My 12v battery is the original battery that came with the car.  It’s been drained to zero volts more times than I can count, and it only stays alive because the demand on it is so low.  So there is no chance I could use this battery alone to power my 12v system and get home.  Fortunately we had a couple of 4P 100Ah thundersky battery modules that were used as a battery backup on one of our products at work.  I had checked them and left them at roughly 55% SOC for long term storage a few months earlier so I knew one would have plenty of energy to power the car and get me home.  I disconnected the dc/dc converter from the 12v and 300v batteries and temporarily connected the TS pack to my 12v system.  The voltage came up quickly as it brought the lead acid back to life.  I drove home without issue, though being much more conservative in my driving and use of the wipers/fan etc.

dc-dc toast

The dc/dc converter that I use is similar in concept the the Vicor Megapac that many people use, it’s a PFC front end that will accept AC or DC input over a wide range with a motherboard style layout that allows you to install a number of different modules for your desired usage.  The motherboard was toast (literally) but luckily all of the modules were just fine.  I plugged these modules into a spare motherboard I had on hand and the dc/dc converter was back up and running with no downtime.  I also took this opportunity to “Fix” the fan in the dc/dc converter that sounded like a small jet turbine and the noise/airflow was driving me crazy, as it was by far the loudest thing in the car.  I played with the voltage to the fan and found a nice compromise between airflow and noise.  The car is much nicer to drive now.  There is a good chance I will be able to repair the motherboard and lowering the fan noise was a big win so all in all this “break down” wasn’t a bad thing.