Kyle’s Rewire of a 110 Volt EVSE into a 220 Volt EVSE

City Car Kyle was helping a friend who had somehow managed to chop the cord and plug off his only 110 volt EVSE for his Nissan Leaf. This left him with no way to charge his only car. Kyle gave him a EVSE in trade with the home of converting it from a 110 volt to a 110 or 220 volt EVSE charge plug. Kyle had nothing to lose; letting out a little magic smoke wouldn’t make it any more broken (as long as it wasn’t let out with Kyle in circuit).

After doing some research and examining the innards of the EVSE, it looks like three wires go into it, two of which are switched and monitored (the neutral and line); the ground is simply grounded. Rewiring it with a NEMA 13/50 three wire 220V plug should be possible by simply using the neutral as the second 110V line in.

It worked. Now Kyle has a portable 220V EVSE capable of 12 amps at 220 volts, or twice the charge rate of the original charge cord. There was one additional difficulty to overcome: the 110 three prong plug end had been cut in half. Within the plug end was some type of resistor. We think its a thermistor that is wired into the charger’s control board. The control board needs to ‘see’ a specific resistance, + or – some unknown amount, in order to continue charging. This is so that when it gets warm, or in this case too warm, the resistance in the thermistor changes and once it’s outside of the acceptable operating range the EVSE shuts off. It’s a simple but effective way to prevent the plug from overheating and burning the house down while charging.

We had to measure this resistance and add a resistor to the control board in the EVSE, which is not as effective to prevent overheating but is very effective in making the EVSE capable of charging at 220 volts.

According to a short update I received from Kyle, the 220 charger only works for a short period then shuts down. Although it worked fine on the Smart Electric Drive and IMW Seven, on a vehicle that can’t use a ‘cheater’ cord – a non-EVSE compliant charge cord that’s missing the proximity and/or pilot signals– the 220 cord only works for a few tens of seconds and then his vehicle stops charging. The Smart seems to need only the proximity not the pilot. and Seven doesn’t need either, so a ‘cheater’ cord works fine for them.

Our guesses? We think that it’s something to do with the pilot signal, other than that we haven’t confirmed anything yet. It’s a work in progress.

Adding an air filter to the intake on Jen’s Wheego LiFe

Jen’s 2013 Wheego LiFe has over 15,000 miles on it and is running OK. Like any small production vehicle, it has its quirks, but overall has been an OK electric car. It has an HPEV AC50 motor, Curtis controller, 3 Delta Q charger(s) (two of those pulling double duty as DC-DC converters) and some type of BMS all packaged into a vehicle originally manufactured for use in China

The car has had some issues compared to modern cars built for the North American and European markets: a few extra rattles, the fit and finish is more GM circa 1980 than Tesla. One of the axles is about ⅜ of an inch longer than it should be and this causes an occasional bang or popping noise in the steering linkage and lower ball joint.

The only other issue we’ve noticed with it is related to living in the country and driving on hard pack gravel roads. The AC50 is an air cooled motor. The Wheego’s 12 volt fan draws air from underneath the car and routing it to the side of the motor to keep it cool. The problem we’re running into is caused by a combination of things. The upgraded AC50 uses an optical sensor known as an encoder to monitor motor position, opposed to the previous magnetic encoder. Life in the country, where it’s dusty from agriculture, corn pollen, soil field erosion, and gravel roads, is not compatible with the cooling system drawing unfiltered air from the underside of the car.

Once the dust and pollen collects and blocks the encoder optics, the Wheego begins behaving erratically at higher speeds. When the encoder optics became dirty enough, low speeds also become inhibited. Which is to say, when the encoder is dirty, the car jutters violently as if it’s about to drop a motor or transmission at any moment.

Naturally, the answer is to clean the encoder. I had to put the car on the lift every three months or so. We added an aftermarket filter to the intake and turned it upright so it wouldn’t be drawing air right off the dusty gravel road. It worked. Now about once every 6 months it needs cleaning and that we think that’s mainly due to the re-rocking of the roads.

The air filter we used is a small old-school style, 4 inch round paper filter as opposed to a K&N type filter. Why, you ask? The K&N filters are covered in a light coating of oil, which really helps increase their filtering efficiency, however it also slowly coats any downstream sensors with a fine layer of oil…which then collects dust. Using an uncoated paper filter yields better results.

Eventually we may have a shorter axle made to address the banging/popping issue before it causes other problems.

Seven’s new dashboard

The new dash and console for Seven is progressing slowly. The Molex connectors used liberally throughout Seven’s wiring harness were never really meant to be used in a car for every connection. I have my doubts whether they were meant for much of any use in a car other than temporary. They are not automotive-grade connectors; they aren’t sealed from the elements and are not vibration-proof. They work well, are easy to hook up, cost effective and everything plugs into everything else…relatively. In fact, if it weren’t for our LED lighting and some of the connections slowly working themselves loose, I’d say they’re great for a car.

Our LED lighting is starting to have connection issues. With LED lights, especially our headlights, it’s necessary to maintain a minimum voltage, in this case 12.0 or higher, because unlike incandescent or halogen lights that merely dim a tiny bit if the voltage drops below 12 volts, LED lights turn off for a fraction of a second until the voltage rebounds to 12+ volts and the lights come back on. The voltage drops again because of the load the lights put on it and the lights shut off again. The fluctuation occurs quite rapidly, to the point where the road in front of you looks like a 1970’s disco with flashing lights daring you to either dance or go into an epileptic seizure.

All of our voltage supplies seem to be OK. The problem appears to originate at the Molex connectors, in the form of a 0.1-0.2 volt drop occurring across each connector. For one or two connectors this is not a big deal, but we used these connectors for every connection in the car and have rewired the car three times, splicing in more connections every time some component was changed.

I don’t want to enter the dark ages while driving down the highway some night at 70 MPH.

The connections are not strain-relief, and with a car that vibrates and a Kevin who keeps taking things apart to make them ‘better,’ we’re running into wiring fallout. The wires seem fine one day and then literally fall out of the back of the connectors the next. So along with building a new dash, we’ll be rewiring almost everything in the dash.

The Smart Electric Drive open source/hack project

As many of you know, we’ve been working on open-sourcing a drivetrain from a 2011 Smart ForTwo Electric Drive. We can control the DC-DC converter, the AC unit, the charger and make the motor turn. It was time to move from our complete drivetrain test assembly to a workbench assembly where each of the components are set up and wired both independently from, and to each other, verifying we can run each of the items successfully independent of the vehicle’s wiring harness and together as if they have been placed into a custom EV build.

After moving everything to a bench and chasing down lots of wiring issues, setting up liquid cooling and a high voltage power supply, etc., we realized that even though we had all of the components working (to some extent), we still had to:

  • map the motor data streams for some semblance of speed control over the motor (all we can do now is make it spin at one speed)
  • build a control board
  • write software to pull everything together into a single package and make it more than less usable.

All of this would be done by trying to control the motor through the existing control board via the existing software and whatever surprises that may have in store for us.

We were batting around different ways of making all this work, noting that none of us currently working on the project are what you’d call “programmers.” After much discussion and debate, we decided to take a page out of EV West’s book and build a new board that would control both the IGBT driver board and the DC to DC converter board and bolt into the existing control board’s location inside the motor housing. Now we can take advantage of work that has already been completed, and move the project along more quickly. Our board design is patterned after Damien Maguire’s Tesla build.

The new approach is to replace the inverter’s control board and software; this will make the motor easier to control, as we have access to the software that runs it, but slightly less efficient due to the type of control strategy being used. However, the control strategy we’ll be employing will work on almost any synchronous-wound AC motor and is approximately only 1-2% less efficient – a huge advantage for the small loss in efficiency.

Ideally, after we’ve built this board and have it working, we’ll be able to use it as a template for almost any other AC synchronous motor produced (i.e. Tesla, GM, Nissan), and quickly adapt it to run any of these motors. Of course the plan is to open source everything so you can build one, too. We may also make a small run of the boards for the Smart; if we’re going to build 4 or 5 for ourselves, we may as well have 50 made.

I’d like to thank Jarrod, Damien, Collin, the Software Guy, and all the other open source folks out there making this possible. Also thanks to EVWest who’ve been instrumental in making this happen and supplying the parts we need.

Here’s a link to Damien Maquire’s Tesla Project so you can see how he does it:

Open Source Documentation for Smart ForTwo Electric Drivetrain

We’ve been working on reverse-engineering a Smart ForTwo electric drivetrain with the goal of open sourcing the vehicle components for general electric vehicle conversions or other hobbyist uses. There are four main components in the Smart’s electric drive train: DC-to-DC converter, AC system, vehicle charger, and the drive motor with controller.

The DC-to-DC converter has the same function as an alternator for a standard gas powered vehicle: it supplies power for and maintains the charge of the 12 volt system. The air-conditioning (AC system) keeps the car cool on a warm day. The vehicle charger charges the high voltage battery pack from either a 110 or 220 volt standard household outlet, or from a standard J1772 vehicle charging station with the appropriate plug.

Note: All work completed so far has been to get the components functional while separate from the car, nothing has been done to decrypt CAN data packets. We are not addressing the battery pack for this vehicle. Any and all use of vehicle power sources and or components including the batteries, are up to the discretion of the user.

To accomplish our goals we needed a couple test drive trains, a working vehicle and some way to ‘sniff’ the CAN messages. The CAN capture and control device of our choice is an Arduino DUE combined with EVTV’s CANDUE. Along with this hardware we needed some software. There are huge cashes of Arduino sketches available. They’re are like apps for your smart phone and work much the same way. Each one is designed to do one or more specific tasks like turn on a switch or make your Lego Techniques do the Hokey Pokey.  Sketches are written, edited and uploaded through Arduino’s IDE (integrated development environment). We used GVRET software version 334 developed by Collin Kidder. This library can be download at

These are some of the basic tools you’ll need to either ‘sniff’ or run many of your open-sourced components:


The DC-to-DC converter takes the high voltage DC from the vehicle’s battery pack (for the Smart, this voltage ranges from about 280 to 380 volts) and  converts it to 12 volts (actually about 13.6 volts) to power all of the 12 volt systems like head lights, radio, power windows, etc.

That the DC-DC is a fairly straight forward unit to get running, although a bit hard to get at since it is a board mounted inside the controller casing which is a part of the motor/controller system. To use this unit independently of the inverter/motor you’ll have to install it into it’s own weather-proof housing and supply-appropriate cooling and electrical connectors and ports. You can see the high voltage plug wires which are labeled + and – (photo below). These labels can only be seen from the inside of the inverter case. Power must be supplied in the proper polarity and range between 280-380 volts DC (the full range of these numbers has not been verified). There are also two CAN wires, CAN high and low, pins 16 & 17 respectively on the main 23 pin inverter wiring connector. The DC/DC requires you to send the CAN message 112, included in the CAN file. This message will turn the unit on and off. Onboard software does all the necessary load sensing and automatically adjusts to the load being demanded.

Once you supply high voltage power and the CAN 112 message, that its, the DC/DC should turn on and supply 12 volt output for your system (note: the 12 volt output lines are located next to the high voltage input lines, they are the smaller plastic twist lock connector coming off of the motor/inverter casing.

When supplying high voltage power to the DC/DC converter, like most high voltage components that contain capacitors, a pre-charge resistor should be used.

High voltage DC input to the motor/controller and DC/DC converter

Pictured above is the main connector for the motor/inverter and DC/DC converter. This diagram will be used for all CAN and interlock connections for the motor, controller and DC/DC converter. Only the 11 pins labeled in the picture below were actually used for our work.

Pictured above is the main connector for the motor/inverter and DC/DC converter. This diagram will be used for all CAN and interlock connections for the motor, controller and DC/DC converter. Only the 11 pins labeled in the picture below were actually used for our work.

The AC system follows the same basic strategy as the DC/DC converter to get it running. High voltage input (+,-), needs to be supplied to the orange high voltage plug (pictured below), CAN high and low inputs/outputs need to be sent to pins 2 and 3 and you need to supply 12 volts and ground to pins 1 & 4 (see pin out below). Again since this system runs off high voltage the same precautions, personal and electrical, apply; namely a pre-charge resistor and isolation between the high voltage and you are both a good thing to have.

The high voltage (see pic) must be within the range marked on your AC unit (since the Smart operates between 280 and 380 volts, I’d start with that range). The interlock wire loop to override the interlock should be left open (don’t do anything with them) sending the appropriate CAN inputs. In this case there are 3 messages being used: 0x244 is the start command for the compressor, 0x392 acts as the run command and message 0x452 is the response to the car from the compressor.

By hooking up the AC compressor in this manner and sending the commands listed, you will turn on and run your AC compressor. However, this has noting to do with the environmental controls, fan speed, compressor variable load control; that will all depend on the ‘car’ side of your AC system (e.g. thermostats, defrost settings, etc.) that you set up in your vehicle and will be specific to your system. I recommend using low tech to solve this problem, at least initially, to get it working and tweak it as time allows. To do this, combine the Smart’s AC with something like the Vintage Air evaporator and dash control system. They’re designed to be installed in vehicles that never had AC so they will not be tied into other systems, thus reducing the number of problems with getting the system up and running by some logarithmic amount (or a lot-ish if you’re interested in the technical term).

(See pin out pics) No, I have not put the Smart’s AC into another vehicle yet. We have only got it running. Let me know once you’ve fully integrated a system along with the info on how you did it and we’ll add it to our post as an option for others. (Keep it open source/keep it going!)

The 3.3 kW charger is slightly more complicated than the AC system…but not much.

The charger input has a 5 pin connector, the J1772 plug, that accepts power from an AC power source that can range in voltage from 100 to 250 volts, which includes the common 110 and 220 household voltages. The three larger pins are for power and ground. Line 1 and line 2, labeled as pins B and C, are the 110 power input lines (one from each leg of 110 power coming into your household breaker box. Pin A is the ground, see picture for pin out (this diagram is looking at the input of the charger with the cord removed). Of the two smaller pins, only one is used. It’s a communication wire used for the J1772 protocol handshake and interlock. This requires a known resistance, as per J1772 protocol, be put on the AC input hooked between the pilot and ground pins (I think it’s about 881.6 ohms, but don’t quote me on that, and don’t forget your diode). The high voltage power output ranges from 278-380 volt DC and goes directly to the battery pack. Some choose to put a charge fuse or Kilovac on the output side of the charger between the charger and battery pack. It’s up to you if you want to do this or not –safety is in the hands of the builder. ! Always check output polarity before hooking up to your system !

Next: the CAN communication wires and interlock. This system communicates using standard J1772 protocol. The CAN message you need to send to enable the charger is 122 (see pin out pic below). We currently have no idea what CAN message reports back to the vehicle for the charger. So far none have been required during our testing. The charger automatically adjust to load requirements. The only other thing you need to make the charger work is to supply power to the high voltage interlock. A 1.15 volt power supply is needed with the +1.15 hooked to pin 6 and the ground (-1.15) hooked to pin 1.

The drivetrain is a compact unit consisting of the motor/inverter/and DC/DC converter all in one. It has a 2 wire high voltage input with interlock (interlock on high voltage plug is not necessary to run system) from the battery pack, a 2 wire twist lock 12 volt output from the DC/DC converter and a 23 pin communication terminal that handles all CAN communication, 12 volt Power and Grounds, and high voltage interlock loop. Of these 23 pins only 11 need to be used, 2 of them are grounds (pins 14 & 15) and must be bonded to the DC/DC converter negative output (you may also want to bond the CAN shield, pin 18 to ground; we think it’s a good idea). 4 of the pins are 12 volts +Positive+ (pins 5, 6 ,21 & 23). Pins 2 & 10 are the CAN high and low. And finally the high voltage interlock, pins 16 & 17, which need to have 1.15 volt power supplied, +1.15 volts on pin 16 and 1.15 volt ground on 17.

To enable the drivetrain, a timed ignition signal must be given at the appropriate time with the CAN signals or nothing will happen. We used GVRET software version 334, which now has digital toggle functionality. This was been set up to inject a CAN signal when the ignition line is commanded to turn on from the CAN data we captured turning on an ignition relay supplying PIN 21 with the required 12 volt signal. (We created a CAN message, giving it identifier 101, that tells the ignition relay to come on, we did this by adding a line to the CAN data and building a breadboard with a relay on it which puts 12 volts on the ignition when commanded on).

So far we’ve got the drivetrain spinning on its own using the procedure above, and the CAN captures (link to files at bottom of post). There are many more steps before this will be ready to integrate seamlessly into a build, such as incorporating the acceleration/deceleration signal to acquire speed control. But with the information we’ve supplied here you should be able to run all of the components listed independent of the Smart ForTwo, including the DC/DC converter, AC, charger, and drivetrain.

As I said earlier, all work completed so far has been to get the components functional while separate from the car, nothings been done yet to decrypt CAN data packets. We are not addressing the battery pack for this vehicle, any and all use of vehicle components, including batteries, are up to the discretion of the user.

As always, have fun and best of luck to you. I hope to hear about many successful builds stemming from this work. If you stumble across anything interesting along the way, please share and we’ll add those morsels to our blog so that others can continue to take advantage and learn from the work that’s been done so far.

What’s Next???

We want to make a board that ‘seamlessly’ integrates this info into an easy to use vehicle control module and see about making a portable supercharger– think trailer with a big plug, utilizing a stack of the smart chargers that will supply 90-120 kWs of fast-charging potential. Instructions on how to do-this-yourself are coming in a later blog post. Now where did I put my KCMIL2000 wire Krimper?


The CAN message baud rate was 500 kbps.

The high voltage you supply to the motor has to be within 1 volt (ish) of what the voltage was at during the CAN capture or your motor will not turn! What’s the voltage at which we ran our CAN data captures? That’s an Easter egg for you to find, shouldn’t be that difficult. 😉

Download files (800 kb zip file).

Hack-a-Thon 2016

This weekend members of the Illuminati, new and old, emerged from their long winter’s slumber and gathered from across several states, some in the physical and still others in the virtual, at IMW headquarters to commence our first official Electric Smart Car Drivetrain Hack (or Hack-a-Thon).

Saying that we’ve been slumbering since our last update is a bit misleading, it makes it seem like the projects we work on just sort’ve materialize on their own, like some sort of mystical Illuminati power brought them forth from one of our vast hidden treasure troves, (not only if you believe that do I have a bridge to sell you but for only $99.95 I will send you the Illuminati’s secret recipe for turning lead into gold! (Individual results may vary. Illuminati not responsible for accidental dimensional slippage resulting from use of this or any of our products) subliminal messages, like this one, are not included) when in actuality its just taken this long to get everything set up, the equipment, tools software, hardware and people that make up the team. Not all of this happens over night, and not all of it in one location, currently there are at least a dozen people that have had a direct hand in the project, eight this weekend alone working form three different states on parts ranging from the actual physical hack, to doing emergency CAN captures and surgery on a working vehicle while taking cover in a garage during storms in California too burning the midnight oil rewriting software, I think somewhere in deepest darkest Michigan, for the EVTV Can Due and writing something called parsing software all for data they had, as far as I know, never seen before, so that we could run the CAN data being collected in California on our Can Due and talk to a drivetrain that I know no one has ever attempted such a hack on before. Oh, did I mention that the software and the can capture and vehicle surgery were all completed within two hours!! To say I was impressed is a bit of an understatement. And just think, this is what these people do in there free time for kicks, think of what they can do when really inspired…or fed little Debbie snack cakes and caffeine!! The possibilities are, as I hope you now appreciate, endless.

Let me rewind here for a minute, for those of you who follow us already know that awhile back, last August, Nate showed up at the shop with a proud and cheesy grin just oozing mischief on his face and a motor controller combination out of a Smart Four Two electric vehicle in hand and said that he wanted to hack its controller so we could get it running and put it into his GreenVan. I said, “Sweet!” Knowing Nates’ more than just a little capable when it comes to electronics and not half bad with computers either…and knowing I’m more than just a little lacking in both those areas and that I truly believe there is no end to these seemingly magical abilities Nate possesses…I then, possibly foolishly but definitely naively, said ”Soooo why didn’t you get the whole drivetrain for us to play with?”

<cue ominus sounding background music> Bumm bummm bummmm

And from there is spread out like the arms of the mythical Illuminati octopus, and with the help and support of EV West’s Michael Bream where it again spread out to where I still don’t know all the connections and wheres whys and hows but I know and am thankful it includes: John Russelli, Richard Jones, Dimas Guevara, Linda Prettyman, Alexander Mandel, and Michaels intern at EV West who is also Richard Jones son who I can’t remember or find his name on my phone who has also been working on CAN captures and braved yesterdays storm to perform wiring harness surgery…who I will now simply refer to as Mr. Illuminati Jones Jr.…hmm, sounds a bit conspiratorial, I like it! And on another arm of the Octopus The Hack Team at EVTV that developed the tools and software we’re using that make the hack possible, Byron Izenbaard, Mark Weisheimer, Collin Kidder, and of course Mr. EVTV himself Jack Rickard.

And yet another arm, possibly the head…I get confused easily, the on site team: Byron Izenbaard, City Car Kyle, Nate Knappenburger, and a cast of thousands…or possibly just me and my imaginary friends…again, confused easy.

So now with our assembled cast, crew and victim (a.k.a Smart for Two drivetrain) we commence the hack. First we built a sort of rolling test bench for the drivetrain to rest on that would hold it’s wheels off the ground, allow access to its various parts and wires, hold our equipment while working on it and allow us to easily attach other needful components. What you might ask is the magical device that allows all this at once… from bottom up: a 4×4 table top with casters attached, the drivetrain strapped to a pallet set on top of the castered table, a pallet set on to of the drivetrain and attached to the lower pallet with wood slats from a third pallet screwed to the bottom pallet acting as legs and some scrap plywood for a top. With this set up we can wheel the drivetrain anywhere we want and attached equipment, wires, switches, etc by simply screwing them down. Oh, and we powered the whole thing off of Sevens high voltage charge port…the DC voltage requirement of the Smart was within Sevens high voltage DC bus operating range…convenient. With this setup complete Nate worked on tracking down wiring and hooking up the Can Due and Byron started setting up software to see if we could talk to the vehicles inverter through our setup. As an added precaution, they first removed the inverter board from another unit we had available and plugged it in as a test board…that we if we wired something incorrectly and burned out the board we wouldn’t be burning out the board buried within our test bench unit. With the test board hooked up and receiving and sending data we decided to hook up the real deal; all went well, no magic smoke was released. However, although we seemed to be sending and receiving data to and from the inverter, nothing was actually working, no wheels spinning, no power out of the DC/DC converter, nothing. Byron and Nate figured out that the CAN data we had was not in the format we needed for our hardware to run on the Can Due, so Nate, line by line, started ‘reformatting’ the data and Byron contacted Collin. Nate finished manually reformatting one of the files, but it still didn’t run the system. While Nate was doing that Collin wrote and update for the Can Due and a Parsing program that would reformat the data for us, of course he did this all in two hours from a remote location, impressive, most impressive. With Collins new version of the software for CanDue and the parsing program we were able to try and run multiple different CAN files, none of them worked, it looked like this build weekend was going to be a bust.

At the same time Kyle found a pic that was just posted by EV West, so we figured, although it was getting late on a Saturday night in Illinois, apparently it wasn’t quite as late in California, so I put in a series of semi paniced, semi wishful thinking calls to everyones number I had at EVWest hoping to get a hold of whoever was left awake and that they’d be able to pull some new can data for us…what CAN data I wasn’t sure, we were still trying to figure that out. So as I called and left many messages on everyones phone on the western seaboard, Nate, Kyle and Byron worked on figuring out what data we needed and which OBD port it would have to come from…so each of my messages got longer as this was worked out…leaving I’m sure more than a few confusing sounding voicemails that went something like….

“Hi, the CAN doesn’t work, but Collin wrote new software; what was that Nate?; so now it does but its still not right, the CAN, we need it off another bus, probably hidden, and what do you know about an interlock? Can you get us some new CAN from the hidden bus by tomorrow? Thanks. Oh, this is Kevin…Smith…from Illuminati Motors. Thanks.”

“Kevin again. It’s the right bus but the pins are different, can you reconfigure an OBD connect to swap pins 9 and 11 and 14 and 16?…I mean 9 to 11 and 14 to16. Thanks, oh wait and can you back probe the red and gray wire on pin 16…I mean pin B2 on the brown connector and the red and lavender wire on pin B2 on the black connector?”

So, then some magic happened, we slept and in the morning Richard calls us, by now our directions have increased in length and complexity in equal proportions, and Richard says in response to all this confusion something like, “Yeah, I can do that, got everything I need to make those parts and probe the wires in the car, just need to hear back from John on when he can get the car here and we’ll get you what you need.” And he did, and they did and POOF the info magically showed up in our in boxes. We made some changes to the test bench, ran the CAN file and we were communicating! With the right data! AND nothing happened.

Hmm, or did something happen? Just because the wheels didn’t spin like we thought they should doesn’t mean nothing happened. So, time to think and probe some wires and see what types of outputs, if any we got. On the output from the DC to DC converter at about 13,000 command lines on the CAN file into the run voltage jumped up from zero to 13.5 volts; interesting. We added a small inverter to the output of the DC to DC and hooked up a corded drop light to it and ran the CAN file again. This time when the DC to DC turned on it also lit up the light and our amp probe showed increased current flow in relation to the power draw through the dc to dc converter. Success! We had gotten something to work, and Byron was able to isolate the single CAN command that did it. One component down…a few more to go. But it was getting late and the team had to start heading home so they could go to work Monday. While cleaning up, we figured that there is probably something in the interlock circuit we missed that’s preventing the motor from spinning; as I type this John Russelli from EV West is looking into that issue, we may have something yet tonight, but we also think that there may be a lot of little wiring issues that may prevent us from getting the motor running, interlocks that you wouldn’t at first think of, so while John is back probing wires for us I’m looking into other avenues to get this data, one way or another we will hack this drivetrain into submission. It will take time, nothing worth doing is quick or free. But the experience and education, the fun we have while doing it, we all have the bug now and none of us will stop.

It’s kind’ve addictive, you should try it, you might be surprised a what you can accomplish.

In summary, we tapped into Sevens high voltage battery pack through it’s charge connecter, hooked this to the high voltage input of the Smart drivetrain through a fused connection that also incorporated a small kilovac and a precharge circuit. The Smart drivetrain was then controlled through and EVTV Can Due hooked to a laptop running CAN files captured in California earlier that day, converted with a parsing program written in Michigan the night before, to activate a DC to DC converter that powered an inverter to light up a light bulb. Admittedly, not the easiest way to turn on a light. 😉

Keep in mind, all of this babble you have just read is my annotated version of the story based on my limited knowledge and experience with things like software, programming, and the English language, and therefore is most likely only 50% accurate…but hey, batting 500 ain’t bad.

And that’s not all we had going on at the shop this weekend, Josh came out and brought not only a project to work on but two kids, his son Wyatt and Wyatt’s friend Conner, who brought along a couple projects of their own.

Josh is currently working on rebuilding a 1976 Dodge short bed truck. It’s a beauty, less than 50,000 original miles on an ex government vehicle, it’s in excellent shape down to the shiny black paint on the frame. Only two things really needed fixing on it; one, the front end had bumped into something braking the grill, bending one fender and the hood a little bit, and Two: it was missing an engine. The engine…well, he found a low mileage big block 440, once built it should fit quit nicely. The bodywork is pretty minor, the fender is replaceable, the hood can be straightened and the grill, he never liked anyway so he’s making a new one which was his project for this weekend.

Wyatt and Conner were working on a blue helmet with cup holders they had added to the sides…kind’ve reminded me of an electric blue bear hat, it was actually a hat they had seen on sponge bob. Honestly it looked familiar to me, but I couldn’t place it until they told me…and I’m at a loss for its name now, but it looked pretty cool and they were making some improvements, painting, reattaching the cup holders and adding clear tubing, it’ll probably be another weekend before they can finish it, the paint was drying slowly. Their second project, which was more of a show and tell since it was finished already, was a small electronic device they had named the radio hijacker. They had modified an automotive radio broadcasting device for an iPhone or similar device, made a power source for it, making it handheld so now it could be used anywhere and tuned into any radio station overriding the station normally being broadcast on that frequency and broadcast the music they wanted to play on it. Conner admitted the range was a bit limited but that that was a function of the power source, more power more range. I thought it was pretty cool when they held it up near the shop radio and the station that had been play was Hyjacked and their music started coming over the speakers.

So, two new makers in the making. I’m kind’ve anxious to see what they come up with next.

Plans, plans and more plans

We’re always planning something. The plan seldom works but generally something good will come out of it. Our current plans include:

  • A paint job for Seven (editor’s note: still waffling over the color)
  • Converting a converted van
  • Expanding into and the future of EVs and how hacking will be a part of it.


Paint color has been a question now for a couple years: when are we going to paint it? What color are we going to use and why? This summer I took a few minutes and a laser pyrometer out to the parking lot where I work and started measuring car hood temperatures. I always took Seven’s reading first to use as a baseline, only used a max temperature read on the hoods, and all cars measured had to be facing generally the same direction.

What I found is that the colors I like get hot in the sun. Really hot. Seven’s temp would reach 172 (all temps in ˚F) on an 85 degree day at noon while parked in the sun. In contrast, a white car would be about 120. The problem is that I’d like to put a deep blue paint on seven, metal fleck of course, but deep blue is only 5-10 degrees cooler than the flat black primer, and the epoxy used for Seven’s carbon fiber body panels is only rated to about 180 degrees. It’s not like it’ll suddenly catch fire at that temp or anything, but the epoxy, which is a plastic, will soften a little, making door alignment an issue. But I think we’re going to paint it a deep blue anyway, because dang it, I like blue! We’ll just have to do some magic and see if we cant find a heat reflective paint or IR resistant clear coat…both of which, if we can get them (if they exist), should keep Seven cooler.

Converting a converted Van.

Last fall Nate picked up a GreenVan from Jack Rickard of EVTV. It’s a small Chinese delivery van (not sold in the states) that was converted to an electric vehicle with a lead acid battery pack. Jack swapped out the batteries to lithium, had all the fun he could stand at its top speed of 25 mph, and sold it to Nate.

Nate is a whimsical guy and absolutely took to this van. It looks like a smaller, white version of the Scooby Doo Mystery Machine van. Nate and I brought this mini Mystery Van home and started working on it. All sorts of suspension components need to be replaced from its days of carrying in excess of 1000 pounds of lead battery everywhere it went. Nate translated a bunch of Chinese parts sites and ordered what he needed.

The plan behind this vehicle is to convert it from a very low speed neighborhood vehicle into a not-so-low speed neighborhood vehicle: something that will go 45 mph and hold 4 people and a bicycle. Therein lies our problem: how to make this van do that. Did it ever go that fast?? Where will we put the batteries? What will we use for a drivetrain? And how will we do it without spending a fortune on a van that cost about $3k brand new in China??

That leads us into the future of EVs and EV hacking. So how do we do this? It’s easy with parts we have laying around, like aluminum honeycomb to make a new battery box and entire new floorpan for the van and a drivetrain from an electric Smart ForTwo. You see, the parts we put in Seven, the motor, controller and DC-DC converter are all obsolete; the company doesn’t even make them any more. As with car companies, we’re likely to see many of the old tried and true but low volume, suppliers for EV conversions disappear as cheaper OEM parts become available either through the dealers or more likely from the boneyard. In fact that is exactly what Jack at EVTV is focusing on now: making it so you can use those high quality OEM parts in your build along with hacking your car to make it possible to use those parts in conjunction with the existing gauges and other power equipment already in your soon-to-be-converted vehicle.

How are we getting involved with this? We found a supplier for Smart ForTwo parts from decommissioned vehicles–all used and low miles. “Great,” you say, “how does that help me?” Jack makes these really cool boards that should enable us to hack the Smart drivetrain just like he’s been hacking the Tesla and others. Once hacked, the information will be available to use so others can put a 55kw Smart ForTwo drive in their conversion vehicle. So no, we aren’t writing the software or building any boards. Jack’s already done that and sells them for a good price at his store.

EVWEST is selling the drive system in their online store: motor control combos for under $1000. Once we’ve hacked them it’ll give people another drivetrain option and Nate a van that will go faster than the Mystery Machine ever did (it was only a cartoon, after all).

So wish us luck and we’ll put in the long nights and weekends and see what comes of it. Probably a little Chinese delivery van that can do wheelies (at least once).

The battery pack (7 years later)

I’ve been asked how are batteries are holding up. They are about 7 years old now and were abused a bit throughout their life, more-so early on when we were following the advice of others about BMS and states of charge and charge cycles. We’ve learned a lot since then and if you’ve read my previous posts you are familiar with much of it.

So what’s new? As of last year we removed 5 of our 99 cells and replaced them. Those 5 cells had less than 75% capacity left in them…or so it seemed. This year we haven’t done a thing. The pack was down to about 90% capacity last year. This year the slow dwindling continues and it’s down about another 2-3% depending on weather conditions (about 87%)., which reduces our usable range on a highway trip to about 180 miles. I don’t think that’s too bad at all; at this rate we’ll still have a 150 mile range when the car is 16 years old (old enough to drive itself). Who knows. By then we’ll probably be looking for the new, lighter, more energy dense cells that don’t exist commercially yet…so you know, Teslas’ old cells.

Upgrades and modifications on Seven

As I think back on my last blog, I hadn’t realized I had been quite so remiss in my own babbling way of relating new and used information to our followers. After meeting some of our followers at the Science Center in St. Louis last month, and having some of the other team members reminding me to get off my lazy fingers and start typing… here we are with a very long blog. Hmm, what to do? Where to start?

Why not start with what we’ve finished upgrading on Seven over the last year? The biggest changes where the addition of rubber mounts to the drive train, a new cooling system, custom A-arm bushings, and a new custom Quaffe differential for the transmission.

Seven’s always been very quiet and stealthy from the outside, except for the squeaky hub caps which we fixed last year with the addition of the new rims (the old rims were flexing and the hubcaps weren’t). From the inside, the motor whir vibrated at a harmonic throughout the space frame and the heads of the passengers; the faster we drove the louder it got. The solid mounts made for the XPRIZE race, for easy of manufacturing and efficiency reasons, were the perfect conduit for transmitting the high-pitched motor whir throughout the inside of the vehicle while keeping the outside quiet. So, very carefully, you know, with a Sawzall, we removed the old mounts one at a time, and sometimes modifying the frame as we went to accommodate the new mounts.

It was a simple design that would make a robust mount which would isolate the drivetrain from the vehicle frame: a cylindrical rubber tubing 1.5 inches in diameter pressed into a 1.34 inch steel tube, a steel sleeve in the center of the rubber for the bolt to go through, and some steel ears–two for mounting to the car’s frame and two for mounting to the drivetrain and voilà, you have a rubber isolation mount. Rinse and repeat with slight modifications due to issues with available room and six mounts later we done. With some careful planning and a little luck everything should be right back where it started and aligned, which it was (whew!). This style of mount is still really solid and rigid, not leaving much room for vibration. Because they are still fairly rigid, they will transmit motor noise into the frame, just at a lower frequency and energy than before, which is why a lot of electric vehicle converters will try to reuse the stock gas engine mounts already in a car; they are strong and not very rigid, making them nearly silent when used for a gas engine and reducing any noise to an imperceptible level when used for an electric motor. Our mounts, although not as quiet as stock vehicle mounts, worked quite well. Seven now transmits less than half of the audible noise into the passenger compartment than it once did. The whir that does still get transmitted is at a lower frequency making it much easier on the ears and pleasant to drive without losing any appreciable efficiency in the exchange.

The new cooling system is actually only half new. The temporary Fiero heater core that we’ve been using for four years was replaced with a slim transmission cooler and fan combo from Summit Racing and moved underneath the motor. The new position and new rubber drivetrain mounts required some mods to the original frame. It was time to clean that up a bit anyway. The two worked really well together and give us a much cleaner look under the hood. The peak temperature of any drivetrain component now is 145˚F, and that’s the inner CV joint on the driver side. The peak motor temp, which is the second hottest unit in the driveline after the CV, now tops out at 140˚F. Along with the new mounts and cooling system, we made a removable carbon fiber underbody panel to allow access to these units from under the car. Overall it worked out as we had hoped, keeping peak temperatures down and simplifying the motor compartment.

It’s kind of an odd story how we got to the point where we realized we needed new A-arm (or lower control arm) bushings. None of us had ever replaced them before on any other vehicle we’ve collectively owned, but now after the incredible results we’ve gotten, I’ll be considering it for all builds involving an older or custom vehicle from here-on-out.

Let me tell you about another upgrade we made this winter, a brand new Quaffe differential designed and built by our old friend and trans builder, John Frana of Frana Vehicles. This was an absolute beauty; a work of art in metal and oil. We have a very small Geo Metro gear box hooked to our trans that John has slowly replaced all the inners with until we reached the point where only thing Metro about it is the size and casing. We’re very pleased with all the upgrades; They just work like a trans is supposed to. However, every once in a while with untested custom parts, you run into a piece of material that has a flaw, a micro crack, something that didn’t heat treat quite right, a contaminant in the steel or an angle that might be off by 0.0001 of an inch, none of these possible trouble spots are identifiable until you put the part to the test. Then it either breaks or it doesn’t…and one of our new gears in the differential had a flaw, not a manufacturing error.

Let me tell you the story as it presented itself to us. We had just finished the winter upgrades and had been taking Seven out for some leisurely break-in miles. No reason to hurry as we had no scheduled events to attend. Dad drove the car for awhile, I tooled around in it some and then Josh brought friends out to give them a ride. Josh had been apprehensive about the old trans breaking on him and waited until we had the new parts in before chauffeuring a test ride. So Josh loads up the car with some of his buds from the Petersburg Terror in the Field Haunted Experience. With only 95 miles on Seven since the winter build, Josh goes to turn the vehicle back towards home, EV grins on everyone’s faces, and the steering wheel wont turn…this is odd. Then there’s a “POP” and the wheel now turns freely. Josh completes his turn and heads back to IMW headquarters (aka “the shop”). Back at IMW HQ, the steering wheel locks up again and frees itself with another loud “POP!” I can hear from inside the shop when he turns into the drive. Josh relays the story to me once Seven is parked and he’s concerned about what happened. We talk about it for a few minutes and remember something similar happening when a mounting bearing came loose on the rear steering linkage back when Seven was still just a frame. We figured that must be it: the bearing came loose which allows the U joint to become unaligned and then gets ‘locked’ into position, making a loud “POP” once it gets passed the ‘locked’ up or bound up position. No worries, I figured I’d crawl under the car the next day to straighten it and tighten the bearing mount.

The next day I squeezed my way under Seven and drilled out the rivets that hold the rear steering and wiring channel inspection cover for the rear steering, pulled back the cover and exposed the steering linkage. I inspected and lubed the linkage and asked Jen to help assist with turning the steering wheel while I was under the car trying to see what’s going on. It didn’t stick or pop on us; the only irregular movement it made seemed to be a slight shimmy in the steering rack mounts themselves. I continued checking out the steering, working from where we thought the problem was outward toward the wheels. When I got to the driver side wheel, I needed to jack the car up in order to inspect the steering components in action. Other than a slight movement in the inner tie rod end, which we’ll replace this coming winter build season, I didn’t notice anything and I started to think an outer or inner CV joint has gone bad and is binding under certain conditions. That’s when I saw something odd. Jen turned the steering wheel again, very patiently I might add (turning a wheel back and forth and back and forth and back and forth is quite tedious when it seems like the guy asking you to do this has no idea what he’s looking for). I saw something in addition to our locking steering noise. I noticed that when the steering reached the end of its travel the tie rod stopped for a second, but once a little more pressure was added to the steering wheel it continued its inward journey, pulling the A-arm in with it until the rear a-arm bushing was completely compressed and steel is touching steel. Not good. NOT GOOD at all!

No wonder it’s been so hard to do an alignment on the car; the bushings are shot. Part of that was our fault because of the way we’ve been strapping the car down onto the trailer when we take it places. Part of it was the old bushings, and part was the original design of the bushings. I found out later from John Frana that the only vehicle that he’s ever made custom bushings for is the Dodge Neon because they are notoriously soft. So I asked if he would make us some, too, and WOW what a difference! The car not only drives better and straighter, but its ride has been improved and quieted (which is odd since the new bearings are actual ball bearings in an aluminum sleeve and not rubber at all. They should be louder). Seven also used to make a spring slapping sound when going over big bumps. We’ve heard this before on lots of old cars, the thing is it doesn’t make that sound anymore.

But wait Kevin, what about the steering and popping noise that led you to this discovery? What was that? I’m getting to it…that’s another thing we thought we knew we didn’t really know. As it turns out, one of the new gears in the differential had a flaw in it. We had one flawed gear; the others didn’t break under load. A tooth came off a single planetary gear on the driver’s side of the differential. The broken tooth lodged in the differential but didn’t make any noise or prevent us from driving because the internal gears on the differential only spin when the front wheels are moving at different speeds, i.e. when you’re turning. With the differential lock, our front tires would not longer move at different speeds. This is no problem down a straight country road but a really big problem when coming to a turn. The power steering unit we added to Seven puts out about 1/3 hp max and the drive motor puts out 200 hp max, so you have a 1/3 hp motor fighting against 200 hp to turn the wheels. That’s not going to happen and the steering appears locked in place. If you put enough pressure on the steering, the broken tooth in the differential “POPS” loose and you can steer again.

Our beautiful new differential had broken. I had to tell John Frana the bad news and ask about A-Arm bushings.

John built us some A-arms bushings and made some replacement gears, which we’ll be putting in again during this upcoming winter build.

Other items we’ve added/improved since I last updated the site:

  1. Modifying the side mirrors so they are permanently affixed in place. No longer do they hinge up and down – which was never a design meant for gull wing doors that open up and jar the mirror out of place with every iteration.
  2. Adding a hood release to the inside of the car so we no longer have to pop the hood ‘Chicago Style’ with a screw driver.

Although we looked into mounting our regenerative shock this last winter, we ran into room constraints that would have required us to modify the frame. To do that we would have to pull out the drivetrain we just installed. That experiment will happen this winter. don’t worry, I’m just as excited (ok, probably more exited) than you are to find out how that works. I let you know the results…eventually.

Upcoming Events and Sad Cake

We have several upcoming events including taking part in National Drive Electric Week, a visit to Purdue university, EVCCON 5, and an upcoming TV show we’re featured in…

On September 16 at 6pm CST tune in to the Velocity channel and see Seven!! That’s right, Prime Time, Baby! Thanks to the folks from Translogic who have been doing a smash-up job for years now on their internet show and now have a slot on prime time TV!! Can you give them WOOT!??


Some thoughts about that guy you’re about to see:

Since Episode 156  of Translogic is airing on Sept 16 (yeah, it’s a repeat of the 2014 episode), Jen and I went back and watched it online last night and a realized a couple things:

1) How loud Seven was before the new rubber mounts were installed.

2.) Something that was left out. Ever notice how it seems like one guy put this whole thing together??? He must be some type of super human, the likes of which are only concocted in a lab or comic books. Or is he?

It’s like seeing a movie and this time, this one time, you stay to watch the credits, and you see just how many people it took to pull off that film to which we give the actors not only all the screen time but all the credit when the people who did the real work we only get their name and maybe a title in passing with hundreds sometimes thousands of others.

That’s what I saw the other night when rewatching episode 156…so when you tune in to the episode…or if its passed by the time you’re reading this and you’re about to catch it online, look in the background and you’ll see the cast and crew, sitting quietly at a table not making a peep: the team that did all the real work to make Seven happen: Thomas Pasko, Nate Knappenburger, Josh Spradlin, George Kennedy, Nick Smith. And there were many others like Jen Danzinger (who hides from press), Intern Matt Yochim and other volunteers who were the designers, builders, electricians, engineers, techs and programmers that made me look great and made Seven possible.

Yeah I helped. Yes it was my party…but it’s never much of a party if you’re there eating cake all by yourself. So when you read this or our other blogs, see past press or hear me talk…just remember all of what you’re not seeing, because a party of one isn’t a party at all, it’s just…sad cake.

Back to the Schedule:

September 20, 2015– we’ll be at the St. Louis Planetarium to help celebrate Drive Electric Week, so come out and see Seven.

September 26-29, 2015– We’ll be visiting Intern Matt at Purdue University. showing off the car, and giving rides. You might even catch me giving a talk or two (it might just be one that seems like two).

September 29-October 4, 2015– we’ll be at EVCCON 2015 in Cape Girardeau Missouri. Come out on the public event day Saturday, October 3 to see all the EVs and maybe get a ride in one.