|
May 23rd, 2018
Ok... Decided to make an electric Vehicle. In December I lucked out and found a 2003 Jeep liberty limited edition that had dropped a valve seat. The rest of the car was in great shape. It was selling for $500... I purchased it for $400 and had it towed home. I found on ebay I can get a new Head with valves for around $200usd so rebuiling the engine and seeing it should cost much nor be that hard.
I decided to make an electric bike first but realized an electric car was a better use of time and money for what I would get at the end of it all. A lot of planning has been done but not a lot of work yet. I'll explain what has been done to date in 4 sections. 1) Getting the Jeep ready. 2) Jeep hardware 3) Electronics and 4)Batteries
1) Getting the Jeep Prepped.
This is fairly strait forward. I need to remove anything from the jeep that has to do with a gasoline engine, and anything that won't be useful for the conversion. The first thing i did was to take out anything I could find in the engine compartment that was easy to remove that wouldn't be there when I was done. Some of the engine was already removed from the previous owner when he was trying to repair the jeep which made things a little easier and also meant that the coolent and engine oil had been drained. Also at this time I disconnected all the electrical components I could find. I also disconnected and remove the fuel rail and vacume systems. Next I went under the car and removed the exhaust system (muffler and catalytic converter). I emptied the Transmission oil and transfer case oil and removed the cross member and transmission mount, holding the transmission up via a floor jack. I removed the rear and fron drive shafts. The CV joint in the front had a broken boot and was damaged so I ordered a new one and will repair the shaft. I undid all the electrical, mechanical and vacume hoses from the transfer case and then lowered the transmission down. I unbolted and removed the tranfer case and put it aside. I will be reusing it. My thoughs on electric motor mouting will be later on.
I then undid all the electrical, mechanical and vacume connections of the transmission. The plan was to pull the engine and transmission out as one unit. It did take me a couple of tries to get the right length of chain and positioning as well as I had to remove the heads as the headers got caught on the engine mount. After a few attempts I was able to get the engine out and used a come-a-long to pull the transmission end up while doing it. Once the engine was out I took off the transmission and put that away. I also put the engine on an engine stand to start rebuilding. At this point the engine compartment is empty except for wiring harness, AC compressor and power steering pump.
I then removed the gas tank (thankfully it was almost empty), fill nozzel, and all the fuel lines. Lastly I removed all the heat shields as there's nothing producing heat anymore. I also removed the PCM and TCM modules as those wont be helpful anymore. Inside I removed the center cluster (including radio) as a new android touch screen will be installed. I took out the instrument cluster and the overhead computer as those 2 run on a J1850vpw bus. Although I can read and write to the bus there is no documentation that I can find for jeep device Id's and such so trying to sniff the bus and figure things out with soooo much on it would take way to long. I've decided to make my own electronic to run the insturments and over head computer. The only thing still on the J1850vpw will be the airbag module. This runs standalone and just sends a code to turn on and off the airbag light on the instrument cluster which should be easy to sniff.
2) Jeep hardware
So.. this is the hardware that will be needed not including electronics. So to begin with, with no engine some of the systems that run off the system wont work... most specifically; power brakes, power steering, heating, and air condtioning.
So lets start with power brakes. The brakes are assisted by a vacume system that is provided by the engine, but this vacume is no longer available. I went to the auto wreaker and purchased a 12v vacume pump from a diesel F250. This is a small pump and connects to a reservoir (I'm making a reservoir from 3" pvc pipe). The pump has a switch built in so it will pump up the reservoir and when the vacume lowers in the reservoir it turns on. Vacume us used in more then the brakes. Vacume is also used for the climate control actuators (opening and closing doors/vents).
For the power steering, I purchased a electric power steering pump from an mini cooper. It should work well. The only issue is it draws around 50A at 12V which runs all the time. This would have to run off a beefy DC-DC converter.
Lastly heat and AC. These will kinda go together. In a regular car heat is produced from the excess heat from the engine. As there is no engine I plan on adding a resistive heater. The benifit is that heat is near instant, however, the power draw from a resistive heater is high and shortens range. To deal with this I plan on replacing the AC compressor (which wont run stock anyway) with an electric compressor but replumb like the Nissan Leaf. I would need to add a small radiator to replace the heating core. The heat pump would supply heat at much less a cost to battery efficency. The AC compressor I decided to use is the compressor from a prius. The unfortunate thing is that it runs at 205v where as our battery voltage may be above 400v. However, this may be a benifit and will explain that in the electronics section.
3) Electronics
Ok... the electronics.... I've come to decide on a few things, and a few thing have to be left open until further testing is finished.
i) Charger - So, I was able to get my hands on an opensource charger schematic that fits the bill and I can modify it to suit my needs. This is based on emotorworkz 12kw charger v14. I liked the charger as it was able to take in and output a wide range of voltages. I'm going to have to get setup the charger for accepting 450v and outputing the maximum (425v) if my battery option works out. Otherwise just using stock design would work. This charger uses an LCD with buttons (which I wont build into it) and an arduino as a control board. I plan on having the ECM control the charger through either spi or canbus information so the LCD and any buttons on the charger are un-needed. I've put the schematic in kicad and designed the 3 seperate PCB's which will be needed. I'm planning on etching my own PCB's so it's only a double layer and fairly easy to work with. The main heat comes from the IGBT and the inductors. The IGBT's will be attached to an aluminum box (on the outside). Inside will be a heatsink that has the main coolant running though it. This should be really effective to cool down the IGBT's The inductor will be placed in a box that has mineral oil flowing though it to a radiator. Using mineral oil as coolant isn't as efficient but it should be fine and it's better then shorting it out with regular water or antifreeze.
ii) Speed Controller (or inverter) - The inverter is a VESC based speed controller modified by a fellow named Marcos. Instead of a small unit that handles ~50A sustained and only 60V, Marco's desgin is modular. There is a control unit which has all the features of the VESC and some extra safety features. The build is almost 2x the price for the BOM but it's worth it to allow this to run like it is. Marco's Logic board connects to an integration board. This can be redesigned for whatever drivers/mosfets/power system one would like. I'm using Marco's demo integration board and design that uses 3 igbts and has a rating of 1200V and 600A. Now the integration board is only viable for 150kw due to shunts and other components. But 150kw is almost double the 80kw the Leaf motor runs at (yes, I am planning on using an EM57 Leaf Motor). Again, I'm thinking of running the logic board in mineral oil (although it may not need it and I'll keep it out) and running the 3x IGBT's the same as the charger to keep them cool. I have redone the PCB layout so it's 2 layers instead of 4. Marco's said the 4 layer design was due to ease of PCB layout and to protect from EMI, however, my logic board will not be right on top of the IGBT's like his design so EMI shouldn't be an issue (if it's mineral cooled it will be in a metal box anyway). I just havent decided if I want to etch these boards or send them to OSH Park to be printed.
iii) Mid voltage conversion/Second Charger - I've decieded I need to have a MidLevel voltage of around 210V for 3 reasons. 1) dropping the voltage from 400v to 12v will incure a larger loss in power and is more complex to build. 2)I need around 205 (210 is close enough) to run the prius compressor and 3) I alread have an 1.8kw 12v DC (so 140a) converter from 180-240v. I'm going to use a second carger for this application as the and connect the High Voltage battery feed to the second phase of the charger to bring the voltage down to 210v. It's also just easier to build a second carger to do this then pulling out the second converter circuit and redoing the board. An added perc is that under the hood will be a second charger that can be enabled on a second line meaning we can charge at 2x12kw if we do layer 2 or 2x30kw if we use a fast charger. When using the second charger as a charger the 12v system would have to run on battery alone, but that shouldn't be an issue
iv) Speed controller/inverter for Compressor - The Prius compressor is a 210v AC motor, so I'm building a simplified mosfet driven inverter to run the compressor
v) 12V/5V system Voltage - a lot of the car still runs on 12v. I have a 12v 140a dc-dc converter to use to step down from 210. This should provide more then enough power for all the 12v systems and I also have a 12v-5v converter for the logic boards that need it.
vi) ECM (Engine Control Module) - The ECM is a controller that provides the function to control all motor and drive train devices. This is the module that communicates directly with the Speed Controller. Basically anything under the hood is controlled by this device. The ECM brains is a arduino mega. So far it looks like I'll have enough pins so I wont need a daughter board but if I run out I'll have to make a daughter board to expand the pins available.
vii) BCM (Body Control Module) - The BCM controls most of the other body features of the vehicle, such as lights, door locks, Enviromental controls, security system, etc. Currently I have this designed with a Arduino Nano and a daughter card with 4x 74hc165 and 3x 74hc595 chips to expand the input/output pins available. The daughter card also have mosfets to switch 12v systems (kinda like a solid state relay) and voltage dividers to read 12v pins to the 5v that logic circuits require. If there's room (I havent taken the current BCM out of the jeep yet) I may switch to a Mega to provide more pins and to give more processing power.
viii) ICM (Instrument Control Module) - The jeep's instrumentation is all electronic and runs via the J1850vpw bus. This is not easy to hack so starting from scratch would be easier. I'm just going to make a backplane for the LED's and stepper motors. I havent decided if I'm going to etch the backplane or not but I am leaning toward doing a pcb for it (makes for easier and tidier wiring). The ICM will controll just the insturment cluster (gauges and warning lights) as well as the overhead computer. In the Insturment cluster there is a small LCD display that we can use to show warnings, odometer, trip meter, etc and has 1 button to push. The Overhead Computer has a number of buttons and 2 lines of lcd which can be used to display various information. The Odometer is stored within this module and this module will also connect to the colomb counter.
xi) CCM (Communications Control Module) - Havent decided if I'm going to do this seperate. The communications module will control all external communications. This will include serial connections, can-bus, Cellular data, and maybe GPS systems. It's the module that will communicate with a centeral server and provide access via a webpage or android app to the cars functions.
4) Batteries
So, my goal is a ~30kw of batteries. I'm going to start by testing the vehicle with 10kw of 12v batteries (30x 26ah batteries in series) which should give around 50km of range on a charge... this is good for testing but not for a final result. I can purchase lithium batteries (the type and configuration change everything I think about it). The cost for lithiums would run around $3000-5000 plus I would need to add a BMS (or multiple BMS's depending on the pack configuration) which would also add into the cost. Although Lithium batteries can be purchased and will work, I'm looking at some other options that have me a bit excited.
I'm looking at making dual carbon batteries using aluminim foil as a current collector, activated carbon dust in a binder as the anode and cathode and a non-aquious, non-organic (I believe) electrolyte. These technically wouldn't be batteries but supercapacitors which have some advantages and disadvantages. Major testing will begin soon to see if this is a viable option and what components to use. I dont what to list what I'm using as it may change, and even if it doesnt, it's kinda propriatory. The Supercaps will NOT hold as much energy per volume as lithium ion batteries (If they do I'll be rich), but they should hold double the energy per volume and weight as lead acid batteries. However, since I'm making these I can make them to exact size and specs I need and they dont need to be keep in one place. I'm planning on putting them where the gas tank was, in the engine compartment, and inbetween a skidplate and body (I'm planning on a skid plate for aerodynamics, but will also provide space for more thin batteries). If I need more I can even start putting them in door panels, under seats and in the headliner.
The following will be valid once testing is done. The test may come back different then expected. I will update more as the tests proceed.
Disadvantages of these batteries: So the main disadvantage for the dual carbon batteries is that I highly doubt it will hold the energy capacity of lithium batteries, but I am fairly certian it will be better then lead acid batteries. So it all depends on where between the 2 they sit. Although this may be offset by some of the advantages.
Advantages of these batteries: There are numerous advantages to the poposed dual carbon batteries. Although they hold less energy, I can probably "fit" more into the vehicle as I can control the size and can make it so theres no wasted space. Another thing using supercapacitors is that they have a different voltage curve. Lithium batteries (and most other types of batteries) are a little higher at the start then hold a small voltage range till they die and if you charge over that voltage or discharge under that voltage it damages the battery (and in the case of Lithium, they can ignite). For instance, a lithium battery cell has a top voltage of 4.2v and dead voltage of 3.2v. The supercapacitors will charge to their top voltage and then stop taking power, so you cant over charge them, and you can discharge to 0v without damaging the supercapacitor. For most EV designs this causes a problem as the vehicle is designed to run in a specific battery pack voltage range... so while the Supercapacitor is over or under this range you can't use them. This would lead to only around 40% of the stored power being available unless you did a dc-dc conversion but that incurs losses and is still problematic. My electronics are geared to allowing the battery pack to operate between 600v-0v. As the Voltage goes down the draw on the batteries will go up. As these supercapacitors can discharge near instantaniously without damage to the battery high current isn't an issue for the battery but a current limit will have to be placed to not damage other components like the motor. So we can recover most of the energy from the batteries (even more then lithium, I think) as we can operate over a large range. When the battery voltage drops below 200v some degration in performance will probabaly start to occur... and once the battery pack is at 50v I'm doubtful the car will continue to run. Some of the other advantages being that the battery is made out of non-toxic material. You can throw the battery away when it's dead and there's no enviroment impact. It can be charged/discharged rapidly without damage to the battery, it can even be shorted without damage (although whatever shorted it may melt). It can run for 10,000+ cycles. Lastly, the Battery is completely non flamible and will not catch fire or explode if damaged. Also it would be $300-$500 to do the battery pack as oppposed to $3000-$5000 for lithium.
Final thoughts: Just trying to finish this up... as I carry on I'll try to keep more information posted and will go into more details on how things work.
May 27th, 2018
Ok... Been working on the electronics to run everything else in the vehicle. These are modules to replace the controllers that came with the Jeep.
1) Instrument Control Module- I had to replace the instrument cluster on the Jeep as it would take too long to figure out the PCM system that they have. This is a drop in replacement for the circuit board that they have (connector is different). It's going to be a bugger to print the PCB as all components have to line up exactly. I'll start that when I have the instrument cluster in hand
ICM-Schematic
2) Over Head Console - This is just an extension to the ICM above and connects with a 25pin cable (22 wires are used). Again, I'll need the console cover in hand to get this setup exactly.
OHC-Schematic
I'm just trying to finish off the Body Control Module.
June 9th, 2018
So a couple of Topics.....
1) Jeep
I've mostly been working on programming recently. I've completed the programming for the ICM which includes the OHC. It will run most guages and errors, but there is still more to be done and enhancements I can think of. The Schematic has been updated a bit as well but will post that when it's finalized. The Code for the microcontroller is so fare 1528 linces of C and I'm glad I selected teh Atmega2560 as the chip. I usually use a Atmega328 but that chip doesnt have enough memory (or pins) to support the code or IO. I have a bit of code (~1000 lines) for the BCM but I need to pull out the Jeeps current BCM to confirm the size and what I can do there. As of tonight I'm working on the code to the ECM. I'm hoping that I can connect the VESC inverter via the Can Bus but am still trying to figure out the signaling. I can hook it up via UART as the Atmega2560 has 4 serial connections but I'd rather reserver those pins if possible and connecting the inverter to the Can Bus means that any module can read the stats that it sends. I believe I'm going to keep the code private for the BCM's as I'm not quite sure what I will end up with at the end.
2) Battery
I've also started the preperation for some battery (Super Capacitor) testing. I have prepared on of the electrolytes I want to test and have purchased the ingredients to make the carbon Anode and Cathode. I want to try various electrolytes and see what results I get. After that I want to try various membranes and see what is the best. Once I know what materials I'm going to use I want to test various configurations and record the results. From the research I'm hoping to find the best case for power. I realize I havent written anything about what I'm doing with battery research so I'll probably start that seperatly here
A final note is I decided to start a gofundme campaign to help fund the research and construction for proof of concept. I'm hoping that once I'm done I'll have a commercially viable option for EV and conversions. I'm currently funding the research and development and will continue to do so, however, it slows down progress when I need to wait until I have the funds to continue. A link to the gofundme campaign is here. Any hep is appriciated.
July 29th, 2018
Well, after being distracted with work and some voluenteer commitiments I can finally get back to this project. I have been thinking about things. I originally looked at using an alternator converted to a motor but the VESC to get the amperature up I needed to add 6x the mosfets as the original VESC is only 60v. Even looking at raising it to 100v the cost of the controller and mosfets became too high. So I decided at the time to go with a 80kw leaf engine and a high power VESC. However, when looking at the discharge voltage curve for that dual carbon batteries and realized I need to keep a constant voltage of 400v which will also help with amperage that the system can handle. More on that later. With a constant 400v and assuming we can get 25kw out of the alternator that 's only 60amps. And that is max, ie. Foot to the floor acceleration. So I'm going back to using a small motor made from an alternator. The goal is 25kw of power and I plan on using 2 for front wheels and 2 for back wheels giving a total of 100kw power. The nice thing is I can start off with 2 and add more (up to 6 motors) if needed. For each motor a VESC speed controller will be used. The parts for the VESC will be affordable and as it will always run at 400v the current will be kept low. For 25kw we need 60a. The mosfet I plan on using is rated for 600v and 20a continuous or 80a pulsed (all depends on cooling). I'm going to design for 3 sets of mosfets (18 total) but only use 2 sets (12 mosfets) and leave the 3rd set of pads open so one can be added as needed. I think only one set is needed, I'm putting 2 in for a so I wont be running it hard. I believe 3 would be overkill.
As the max that each speed controller is doing would be 60 amps we can keep the cables between the power supply (charger) and the speed controllers on the smaller side. I'll be using my charger which can put 400v and 100a stock (I think I may need to up the current ability of the charger so that it can get to 200+, but I'll have to look at that.
I'm planning on installing 2 chargers in the jeep anyway. This means I can do a fast charge at 400v at 200a or 80kw. Meaning I could charge batteries from empty to full in approx 30min. As these chargers arn't used while driving I can use them for other purposes. One Charger I'm planning on using to convert power from battery to 220v which will run the AC compressor and also power the 12v power supply. Once battery voltage falls below 200v the charger wil be taken out of loop and these units will run directly from battery (no point in increasing voltage then decreasing voltage as there are losses involved). The second charger will be used when driving to convert the fluctuating battery voltage to 400v for the speed controllers. I guess if the charger gets pushed up close to it's 100a limit we can disconnect the second charger from provideing power to AC and to 12V system and can provide up to 100 more amps to the speed controller. This will temporarily disconnect the AC and 12v supply but we will have batteries (capacitors) for both for a short time. However, even at 200amps we would only have 80kw of power available to us. I'd like more... so I'll have to look at either a 3rd charger or change the charger to handle more current. I'm not sure which I'll use. These chargers will be connected and disconnected in certian ways via use of Contactors so when they change over the currents have to be non-existant to avoid a surge or arcing. Now if I go with a 3rd charger, do I actually build a charger or just do the 2nd stage (saves some money).
August 5th, 2018
So... I've been thinking more... which is usually a bad idea. So going to use the chargers as I posted earlier. I was thinking of the chargers. If the output is always the charge battery voltage (ie, 425 volts) there is no need for the second stange of the charger which is a buck converter. For instance... when charging that the batteries, the charge voltage is 425v and the input voltage will always be less (as the output can't go above 425v, if it could I would do that). After charging when the charger is used to provide a constant 425v to the speed controllers. So as the battery voltage drops the output voltage is always 425 volts. So I've been redesigning the charger. First I found there was a lot of extra stuff in the controller that isn't even hooked up. Also removing the second ibgt, driver, and inductor as well as all the Capacitors we free up a lot of space on the boards and also dont need as much for each charger. So going to do 2 chargers (or 80kw contious power). If I need a 3rd I can add one.
For those wondering what the use of the second stage of the charger is for. The charger design I was working from allowed 12v-425v input with 50v-425v output. To accomplish this extreme range of voltages it takes the input voltage in the 1st stage (boost converter of PFC) and raises it in the to 425v. The second stange (or buck converter) takes that 425v and drops it to the desired voltage.
Unlike the post before, with no buck converter I can't use a charger to drop voltage to 220v for the heat pump compressor or the 220v to 12v converter. However, the current of this buck converter isn't much (I think it's like 10amps) so we can use some inexpensive power mosfets to create a buck converter for this application and it can therefore always stay on.
Anyway... Design is redone, but the programming really needs to be cleaned up and I need to check that all the specs are in order. but should be able to start making the the boards.
August 14th, 2018
So.. litle update... trying to move forward with the build. Ive redesigned my charger/power schematic so it is only s buck converter. Also changed the arduino from a mini to nano (just want to standardize as much as possible). Lastly Ive included the CanBus module Im going to use as everything runs off a Can system. Im trying to make the system handle 300+ amps as thats what the ibgt is rated for so Im going to be paralleling 6 2.4in inductors in a mineral oil bath. Lastly, I was looking at the software side and was using code that was for the original hardware design.... but with all the changes including the CAN control Ill need to basically tear the code apart and start from almost ground up. Once I get a powersupply going I can make a speed controller and test out a motor using the 3 units.
As the programming and redesign of the circuit board will take a bit of time and is not to interesting to post, I probably wont be updating this for a bit. When I finish and more inteesting updates occure Ill start reposting.
|