Engine Factory

Visit our community sponsor

Page 1 of 2 12 LastLast
Results 1 to 40 of 70

Thread: Electric 818 | AC & LiFePO4 configuration

  1. #1

    Electric 818 | AC & LiFePO4 configuration

    The Electric 818 Possibility thread has grown rapidly and branched out into a number of technical discussions. In the interests of narrowing down the electric conversion options, I'm starting this thread to (hopefully) only discuss the possibility of using LiFePO4 (lithium iron phosphate) prismatic cells from CALB (China) and the AC family of motors from HPEVS (California).

    My co-conspirators and I have worked with this EV configuration (AC & LiFePO4) since 2009, covering some 20+ vehicles including 7 high performance sports cars, so I'd like to try and apply what we've learned from the school of hard shocks to Project818. Perhaps other EV pathways - wheel motors, different lithium chemistry, hybrid designs, BMS systems, etc. - will be pursued in parallel threads.

    The basic solution set compared to ICE for both street and race 818s is summarized in this table:

    model comparison.png

    MISSING INFO REQUEST
    One important missing piece of data is the weight that would be subtracted from the 818 if converted to electric drive. I hope someone tearing down a donor car, or a Subaru expert with access to accurate component data, can provide the gross weight (lbs) of the following:

    1. fully dressed Subaru Impreza engine, both stock EJ251-EJ254 and WRX turbo EJ255, without transmission [UPDATED: 260 lbs according to Kennedy Engineering]
    2. radiator
    3. empty fuel tank
    4. starter battery and engine wiring harness (not part of fully dressed engine)
    5. exhaust system
    Last edited by KrissMotors; 03-24-2013 at 09:14 AM. Reason: updated info
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  2. #2

    Help with weight measurements?

    bump ... can anyone provide weight measurements above?
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  3. #3
    Senior Member
    Join Date
    May 2011
    Location
    Largo, Florida
    Posts
    454
    What's the peak power output of the system that you are proposing? What's the battery's capacity in kw*hr?

  4. #4
    For the fuel tank, you want the weight of the tank FFR will provide, not the donor tank, which won't be used in 818. You of course also need the tank capacity, to figure out the weight of the fuel. You will need to get this info from FFR themselves, since no one has a kit yet that could measure the weight.

    You can't use the full WRX exhaust weight, as much of it won't be used. You want weight of WRX manifolds, up & downpipe, etc. Then weight of exhaust system from FFR.

    Also, you mention radiator weight. Don't forget the weight of the coolant, and the weight of the extra piping of radiator hoses from front of 818 to rear. Again, you would need to get this info (amount of coolant 818 uses and weight of extra piping) from FFR.

  5. #5
    Senior Member
    Join Date
    Feb 2011
    Location
    Orange, CA, USA
    Posts
    663
    I have no answers, but here is a long running thread on weights from nasioc.

    And a small list that you might find useful for a/c components.

  6. #6
    Quote Originally Posted by BipDBo View Post
    What's the peak power output of the system that you are proposing? What's the battery's capacity in kw*hr?
    Battery pack capacity is about 12.9 kWh (but see post #8 below for more detail about this calculation). See THIS THREAD (especially post #15) for more details about the AC-50 and AC-75 motors, including discussion of torque curve. The table in my first post on this thread compares the electric to Subaru ICE power output. Please note that horsepower is not very useful when discussing electric motors; torque is a much more relevant metric.
    Last edited by KrissMotors; 03-06-2013 at 11:05 PM.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  7. #7
    Senior Member RM1SepEx's Avatar
    Join Date
    Jul 2011
    Location
    Freeport, ME
    Posts
    2,598
    It's actually always about the area under the HP curve (torque x rpm /5252 over time)

    torque is the twisting force but it is applied by rotation over time and what is important is the torque over time at the wheels, the gearing is important. The advantage of an electric motor is that amazing peak torque at low RPM, it results in a large area under the HP curve

    I thought that the CALB batteries are 3.2 volt nominal per cell http://en.calb.cn/comm/?id-162.html http://www.thunderstruck-ev.com/ca40fi-en-2.html

    I can get you some of the other weights... header minimal exhaust etc... not sure what I can do re engine, I can hoist it but don't have a ready scale, perhaps borrow a set of corner weight scales from a friend...

    What range do you expect from the 36 cell pack

    I'd assume about 10 gallons for the fuel cell @ approx 6 lbs per gallon. 60 lbs

    are those 100AH CALB at 7 lbs each? 252 lbs

  8. #8
    Quote Originally Posted by RM1SepEx View Post

    I thought that the CALB batteries are 3.2 volt nominal per cell http://en.calb.cn/comm/?id-162.html http://www.thunderstruck-ev.com/ca40fi-en-2.html

    I can get you some of the other weights... header minimal exhaust etc... not sure what I can do re engine, I can hoist it but don't have a ready scale, perhaps borrow a set of corner weight scales from a friend...

    What range do you expect from the 36 cell pack

    I'd assume about 10 gallons for the fuel cell @ approx 6 lbs per gallon. 60 lbs

    are those 100AH CALB at 7 lbs each? 252 lbs

    All LiFePO4 cells have a nominal voltage of about 3.4v fully charged (but the definition of "nominal" varies, so some vendors quote the typical operating voltage of about 3.3v or even 3.2v). These cells are typically charged to about 3.55v and then when the charger is disconnected, they fall to about 3.4v after the surface charge dissipates, which is the "float" or "nominal" voltage under no load. As soon as any load is applied to the cell, voltage will drop to about 3.35v. They have a very flat discharge curve, however, and will stay between 3.3v and 3.1v for 90% of the discharge cycle.
    ===
    The 818 could have either a 36 or a 38 cell pack. Mileage will depend on the amp-hour capacity of the cell, and I think a 100 Ah cell would work well. Since the pack is wired in series, voltage will vary with cell numbers, but capacity will remain the same: 100 Ah. Range should be close to 1 Ah per mile given a 1800 lb curb weight. Useful range will hopefully be about 80 miles, perhaps a bit more. Environmentals make some difference; range would be better in warmer temps. This range works out to be about 150 Wh/mile (12.3 kWh/150 = 82 miles).
    ===
    The CALB 100FI Ah cells are approximately 7.5 lbs each, so 36 x 7.5 = 260 lbs and 38 x 7.5 = 285 lbs.
    ===
    Calculation of "capacity" depends frankly on whether you ask an engineer or a marketing guy. The "float" capacity of a 38-cell pack is: 38 x 3.4 nominal voltage x 100 Ah stated capacity = 12.92 kWh which is what marketing would probably say. However, the true cell capacity is something like 111 Ah on average, the typical operating voltage is more like 3.25, so the engineer will answer: 38 x 3.25 x 111 = 13.71 kWh of which only 90% is practically available, or 13.71 x .9 = 12.34 kWh. The final answers are close, but describe different realities.

    Furthermore, Chinese specifications are pretty confusing. Unlike in the U.S., the Chinese tend to understate product performance. Of all the batteries I've seen from China, not a single one has ever been below spec. On the other hand, an American manufacturer -- A123 in Watertown, MA -- also makes LiFePO4 cells in a different pouch format, and I've never seen one of those cells ever meet the specification! I hope this long-winded explanation helps clarify what is generally a confusing product category.
    Last edited by KrissMotors; 03-06-2013 at 11:20 PM.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  9. #9
    Senior Member
    Join Date
    May 2011
    Location
    Largo, Florida
    Posts
    454
    Speak about torque all you want, but power is what actually matters. That's why we have gearboxes, to trade rpm for torque. I see that the peak is 67.21 HP at around 2200 rpm according to this chart. It would move around an 818 around town, but not quickly. This system might be more appealing for a hybrid build than a pure EV.
    75torque.jpg

  10. #10
    Administrator ehansen007's Avatar
    Join Date
    Mar 2011
    Location
    Aliso Viejo, CA
    Posts
    1,108
    I beg to differ on this one BipDBo. Just any one who owns a diesel how they feel about HP over Torque. And, while the subie engine might make the same amount of torque, it's only at peak. Maintained flat torque range using gears can be as effective especially with instantaneous torque. I'm loving what Eric Kris is throwing down here. Solid numbers and experience. Take a look at his site and you'll see what I mean. Details and experience is what he's offering. Also, guys, from my last discussions with them, FFR is planning on using the 33 Hot Rod Tank for this car or something similar behind the seat. At thirteen gallons at 6lb per gallon it's about 78lbs and I think it's futile to include tank weight since you'll need to build a battery box similar or larger in size.
    Last edited by ehansen007; 03-07-2013 at 11:17 AM.
    www.33machine.com

  11. #11
    Senior Member Xusia's Avatar
    Join Date
    Oct 2011
    Location
    Eugene, OR, USA
    Posts
    2,340
    I agree. Here's someone with EV experience trying to help and give us the benefit of his real world experience - at least that's how I see it.

    On a completely different note, I have an honest question for Eric: Why does the motor cost so much? For the same price you can get fully assembled engines which have many more parts, and take many more hours to assemble. It just seems wildly out of proportion to me, given the comparative simplicity of a motor.

  12. #12
    Senior Member wleehendrick's Avatar
    Join Date
    Jul 2012
    Location
    Encinitas, CA
    Posts
    1,114
    Blog Entries
    2
    Uh, oh, I feel the old horsepower/torque debate coming on!

    This is the answer right here:

    Quote Originally Posted by RM1SepEx View Post
    It's actually always about the area under the HP curve (torque x rpm /5252 over time)
    Sure, F=MA, and instantaneous acceleration is dependent on torque at the wheels. But over distance/time, power is the physical quantity needed to accelerate a given mass to a given velocity and dictates resultant acceleration; just ask Issac Newton.

    That being said, a torquey motor (like a electric or diesel) can often have more power under the curve that a NA motor with higher peak power. This is especially true in 1st gear, when you integrate from 0 RPM! However, once up to speed and rowing through the gears, a lot of this advantage is lost.

  13. #13
    Quote Originally Posted by Xusia View Post
    Why does the motor cost so much? For the same price you can get fully assembled engines which have many more parts, and take many more hours to assemble. It just seems wildly out of proportion to me, given the comparative simplicity of a motor.
    First a side bar comment on AC vs. DC motors: Alternating current (AC) motors are more complicated to run than direct current (DC) motors, and require a controller that is expensive, with high end electronics. The cost of the AC50 (or AC75) includes the Curtis controller. The price of a DC motor often does not include the controller. An excellent DC motor is Netgain, and their Warp 9 sells for about $2000 without the controller. An excellent controller for the DC motor is the Soliton 1 and that sells for about $3000. So a comparable DC system is about $5000. You are correct: you can purchase a very good gas engine for $5000. The AC50 motor by itself (but it can't be used without a controller) costs about $1000-$1500 I would guess.

    One reason electric motors seem relatively expensive compared to gas engines is that they have been - up to now - mostly manufactured for industrial applications and they last nearly forever. There is no consumer market for electric motors per se. Also, a typical gas engine needs periodic maintenance whereas a typical electric motor can run for decades 24x7... without maintenance! So from a total cost of ownership perspective, electric motors look more competitive. Still, there is a price premium over what you'd expect.

    Hopefully as electrics penetrate the automotive space, prices will decline. It's still a cottage industry now, and that's the price of entry for early adopters. I certainly don't advocate an electric 818 as a cheaper alternative ... it's not at all. There will be a premium for quite some time, I imagine. But I think we are at - or very near - a tipping point where lightweight electric sports car will become increasingly attractive. Like any new innovation, you will likely see two steps forward and then one step back.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  14. #14
    Senior Member Xusia's Avatar
    Join Date
    Oct 2011
    Location
    Eugene, OR, USA
    Posts
    2,340
    Ahhh. Didn't realize that price included the controller. That is a very important piece, so I'm glad I asked.

  15. #15
    Quote Originally Posted by BipDBo View Post
    Speak about torque all you want, but power is what actually matters. That's why we have gearboxes, to trade rpm for torque. 75torque.jpg
    True dat. Interstingly, with electric drive, the gearbox (if there is one) kinda does the opposite, trades torque for rpm.

  16. #16
    Senior Member RM1SepEx's Avatar
    Join Date
    Jul 2011
    Location
    Freeport, ME
    Posts
    2,598
    And the gearbox decreases efficiency due to increased mechanical drag/losses... counter to what you want with electric drivetrains (remember that mechanical advantage doesn't come w/o a price)

  17. #17
    Senior Member bromikl's Avatar
    Join Date
    Feb 2011
    Location
    Milwaukee, WI
    Posts
    378
    Silly question about RPM:

    All motors have an effective RPM range.
    The transmission matches motor RPM to wheel RPM.
    Does anyone know the RPM of the wheels at say, 100 MPH?

    I ask because due to an electric motor's flat torque curve, a designer could decide to run without a transmission. How fast would the motor need to be designed to turn? And a slightly related question: having a flat torque curve, why not simply use a fixed gear ratio?

  18. #18
    Senior Member Bob_n_Cincy's Avatar
    Join Date
    Sep 2012
    Location
    Cincinnati OH
    Posts
    1,125
    [QUOTE=bromikl;92050]Silly question about RPM:
    Does anyone know the RPM of the wheels at say, 100 MPH?
    QUOTE]

    Assuming a 2 ft diameter tire:

    (5280 x 100mph) / (60 min per hour) / (2ft x 3.14pie) = 1401 RPM

    Direct drive to get 1/2 g acceleration you would need 900 ftlbs of torque on an 1800 lb car. This motor would weigh an additional 500 to 1000 lbs.

    Solution: Add 10:1 gear reduction and motor only needs 90 ftlbs of torque and motor only weighs 100 lbs. Motor would run at 14000 RPM. The weight saving of 400 to 900 lbs out weighs (pun intended) the 3-5% loss in gearbox efficency. This is what GM EV1 and Tesla does.
    Bob
    Last edited by Bob_n_Cincy; 03-08-2013 at 09:00 AM.

  19. #19
    Quote Originally Posted by bromikl View Post
    Does anyone know the RPM of the wheels at say, 100 MPH?
    I ask because due to an electric motor's flat torque curve, a designer could decide to run without a transmission. How fast would the motor need to be designed to turn? And a slightly related question: having a flat torque curve, why not simply use a fixed gear ratio?
    A single gear transaxle is possible but not the best solution for this application IMHO.

    An average tire rotates at 10.4 turns per one mile per hour (exact rotational speed depends on wheel diameter, tire inflation, and other arcane factors). At 100 MPH, the tire itself would be rotating at 1,040 RPM.

    There are several single gear electrics, including the Tesla. They typically use an 8.28:1 ratio. At that ratio, the motor would be spinning at 8,600 RPM at 100 MPH. You'd need a different motor design to handle this higher RPM, with much higher voltage and lower amperage than either the AC-50 or AC-75. A higher voltage circuit is problematical for LiFePO4 prismatic cells because of their large footprint (only 38 cells in series). Tesla uses little laptop-style lithium ion cells and assembles thousands of them in complex series/parallel battery packs. You could build a LiFePO4 pack from A123 pouch cells, but I have yet to see a reliable pack made from these components (and A123 just went bankrupt).

    If you'd like to explore this further, an excellent single ratio transaxle is the Borg Warner 31-03 eGearDrive 8.28 transaxle:
    Screen Shot 2013-03-08 at 9.35.51 AM.png

    This sells for about $3000 with limited availability from EVTV. This gearbox was used by Tesla, Aptera (bankrupt), and CODA (soon to be bankrupt). Engineering is outstanding; motor mounts horizontally in parallel with the wheel axles.

    The electric motor torque curve is flat but does not extend across the entire RPM range of operation. Take the AC-50 for example: it's torque curve is flat from 0 to around 3,300-3,500 RPM depending on system voltage (higher voltage will extend torque to a higher RPM). By 6000 RPM, available torque drops 50% from peak. As speed increases with a single gear transmission, the ability to accelerate declines when you pass the flat part of the torque curve.

    Assuming a 2.88 5th gear ratio on a stock Impreza TY574*** transmission and an average tire rotation of 10.4 turns per one mile per hour, a speed of 100mph would be about 3,000 RPM, very comfortable for either the AC-50 or AC-75 (and at peak torque for the AC-50!). So the Impreza transaxle is a great fit with the AC-50 electric motor and the proposed voltage/amp circuit level.

    A single gear approach would require re-working both system volt/amps as well as a different choice of electric motor. One possible alternative motor is the Siemens 1PV5135 AWS14 motor ($3000 from EVTV in limited supply) with the Azure Dynamics Force Drive DMOC 654 inverter/controller ($2500 from EVTV in limited supply) combined with the eGearDrive transaxle above. But now the component cost is nearly $9000 before you even get to the battery pack which would probably be $10,000+. Furthermore, the Azure 654 device uses CANbus and requires its own Vehicle Control Unit (VCU) which is not commercially available (you have to build and program your own). The VCU communicates with the vehicle's CANbus. Using the above Borg Warner/Siemens/Azure combination is only appropriate for expert automotive engineers with access to proprietary programming and code information.


    The AC | LiFePO4 configuration I'm suggesting is about half the cost of a single gear drive like the eGear | Siemens and much easier to build.

    One further advantage of using the stock transmission: you replicate the feel of the gas-powered car, including nearly identical shift points.


    Last edited by KrissMotors; 03-14-2013 at 08:06 AM.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  20. #20

    System voltage and form factor

    In the prior post (#19), I briefly mentioned the form factor issue regarding LiFePO4 prismatic lithium battery cells and the relationship to system voltage. Here's some background for those less familiar with lithium cell options.

    cells-AC50.jpg

    The above photo shows 9 LiFePO4 cells in series and the AC-50 motor on the right (this is from my 1969 Saab Sonett conversion). A feed, the heavy orange cable, is coming from the rear battery pack and attaches to cell #21 on the upper right of the pack. You can see the battery straps go from terminal to terminal. After the rear feed connection to cell #21 negative terminal, the positive terminal of #21 is connected by strap to #22 negative, and so on. (The positive terminal on #29 will go to another sub-pack on the right side of the motor.) This is a series connection.

    In a series connection, the pack current remains unchanged (it stays the same as a single cell), but the pack voltage is additive. Thus 2 single cells at 3.4v combine in series into a 6.8v pack, etc. The way you increase system voltage is by putting cells together in series. Since the cell form factor is rather large, you can only fit a limited number of cells ... hopefully 36 or 38. This achieves a system voltage around 129v (38 x 3.4 for example). To get a much higher system voltage like the 215v design spec of the Siemens/eGearDrive mentioned in post #19, you'd obviously need at least 64 cells in series.

    To get 64 or more cells, you need to switch form factors and go with either small pouches, cylinders, or similar small cells. As mentioned previously, A123 manufactured pouch format LiFePO4 cells, but they proved to be very difficult to fabricated into reliable packs since delicate metal tabs had to be connected, and a single bad connection disabled the entire pack function.

    This is what the A123 pouch LiFePO4 cell looks like:
    A123 pouch cell.png
    Availability is now uncertain since A123 when bankrupt and has been acquired by a Chinese company.

    The other choice is to move away from LiFePO4 (lithium iron phosphate) chemistry to related, but different, lithium cells. These other cells are typically called "lithium ion" - a very unfortunate and confusing name - and have two major disadvantages: a thermal instability and relatively poor cycle life.

    Thermal instability must be managed with sophisticated cooling systems like Tesla has developed. A similar effort by Boeing in their 787s obviously didn't go well. IMHO this is beyond DIY capabilities at this point.

    The second problem about cycle is also significant. The LiFePO4 cells will last well over 300,000 miles. The cells Tesla uses, for example, have a much shorter life and Tesla is already making preparations to replace battery packs ... probably at significant expense.

    So, for three basic reasons, I'm proposing the large format LiFePO4 prismatic cells:
    1. ease of connection into robust packs
    2. no thermal instability and no need for elaborate battery management systems
    3. exceptional cycle life which eliminates the need for pack replacement


    The disadvantage of these LiFePO4 cells is that you are limited to system voltage below 130v. This eliminates certain high voltage motors, and probably makes impractical design choices like single gear transaxles.

    Hope this detail is helpful.

    Eric Kriss
    Last edited by KrissMotors; 03-08-2013 at 02:00 PM.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  21. #21
    Eric you have a gift, you make this complicated subject understandable even by a technophobe like me. lol

  22. #22
    Senior Member Xusia's Avatar
    Join Date
    Oct 2011
    Location
    Eugene, OR, USA
    Posts
    2,340
    Hi Eric,

    Can you explain something else to me in regards to batteries? I understand the difference between series and parallel connections when it comes to voltage, but how do each of these wiring schemes affect the battery pack's ability to deliver current (amperage), and the pack's rated capacity. I know you said series wiring doesn't affect capacity, but in a previous post you mention the overall capacity of the proposed battery pack as ~12KWh. A single battery is rated at 100Ah, so maybe the real issue is one of KWh vs Ah?

    I appreciate any light you can shed on this!

  23. #23
    Senior Member
    Join Date
    May 2011
    Location
    Largo, Florida
    Posts
    454
    This particular system creates 62 hp, which, I think gearhead would agree is a bit on the low side.
    The 12.9 kw*hr pack is also a bit low, probably really only good for around 44 miles of range.

    Just daydreaming without regard to costs; How about two packs in parallel feeding two motors and two controllers? One configuration could be one motor in front and one in the rear for AWD.

    Another configuration could be two in the rear, geared separatly, which would enable torque vectoring.

    Or, since cost is not a factor, let's imagine going Rimac style with a motor on each wheel for a total of 248 HP, AWD and 51.6 kw*hr!

  24. #24
    Senior Member Nuul's Avatar
    Join Date
    Sep 2011
    Location
    Indianapolis, IN
    Posts
    244
    Quote Originally Posted by KrissMotors View Post
    This is what the A123 pouch LiFePO4 cell looks like:
    A123 pouch cell.png
    Availability is now uncertain since A123 when bankrupt and has been acquired by a Chinese company.
    These are similar to what the R/C car guys are using today, right? They looks similar to some LiPo cells I've seen.

  25. #25
    Quote Originally Posted by Xusia View Post
    I understand the difference between series and parallel connections when it comes to voltage, but how do each of these wiring schemes affect the battery pack's ability to deliver current (amperage), and the pack's rated capacity. I know you said series wiring doesn't affect capacity, but in a previous post you mention the overall capacity of the proposed battery pack as ~12KWh. A single battery is rated at 100Ah, so maybe the real issue is one of KWh vs Ah?

    I appreciate any light you can shed on this!
    Poor choice of words on my part, sorry. Wiring in series impacts the pack voltage. Wiring in parallel impacts the pack current. Sometimes the term "capacity" refers to the amp throughput of the cell, and other times it means the amp-hour storage of the cell, which is confusing. The overall size of a battery pack is calculated by cell voltage x cell current x number of cells. As discussed above, this is 3.4v x 100Ah x 38 cells = 12.9 kWh. If you double the amp-hour size of a cell, from 100Ah to 200Ah for example, the overall size is then 3.4v x 200Ah x 38 cells = 25.8 kWh.

    For illustration, two 100Ah cells wired in series would yield a 6.8v 100Ah "double" cell. Two 100Ah cells wired in parallel would yield a 3.4v 200Ah "double" cell. They would both be the same size. The voltage of a cell is determined by its nominal, or float voltage, and indicates the maximum voltage the cell will generate. As discussed previously, this value is 3.4 for a LiFePO4 cell, although in reality, the cell is typically operating between 3.3v and 3.1v. Current (amps), on the other hand, is measured by something called the "C-rate". The "C" relates to the amp-hour capacity of the cell, a fixed physical characteristic of the cell. In this case, we're discussing 100Ah cells. Current flow is some multiple of this Ah capacity.

    The CALB LiFePO4 cells have a C-rate of 3C for continuous and 10C for maximum current output. This means that a 100Ah cell can output 3 x 100Ah = 300 amps for as long as its charge lasts, and it can output 10 x 100 = 1000 amps for 10 seconds or so. In practice, it's hard to use 10 seconds worth of max current output ... the vehicle is going to be travelling very fast or burning out a gear or two. It typically takes under 100 amps to maintain a vehicle at a constant speed. Thus, for a system that requires a maximum 650amps like the AC motors I describe, the 100Ah cells meets the amp current spec.

    If wired in parallel, two 100Ah cells would have a 200Ah capacity, and thus 3 x 200 = 600 amp constant output and 2000 amp maximum output. This is overkill, so wiring in parallel doesn't make sense at this Ah capacity level.

    Tesla does wire packs in parallel (as well as in series), but they use cells with much smaller Ah capacity.

    I hope I haven't completely tangled up this topic!
    Last edited by KrissMotors; 03-08-2013 at 03:58 PM.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  26. #26
    Quote Originally Posted by Nuul View Post
    (A123 pouch cells) ... These are similar to what the R/C car guys are using today, right? They looks similar to some LiPo cells I've seen.
    I'm not familiar with R/C applications specifically, but I know that small format LiFePO4 is popular for many small motor niches.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  27. #27
    Senior Member bromikl's Avatar
    Join Date
    Feb 2011
    Location
    Milwaukee, WI
    Posts
    378
    Quote Originally Posted by KrissMotors View Post
    The AC | LiFePO4 configuration I'm suggesting is about half the cost of a single gear drive like the eGear | Siemens.

    One further advantage of using the stock transmission: you replicate the feel of the gas-powered car, including nearly identical shift points.


    [/COLOR]
    Thanks. This does seem like the simplest solution.

  28. #28
    Senior Member bromikl's Avatar
    Join Date
    Feb 2011
    Location
    Milwaukee, WI
    Posts
    378
    Quote Originally Posted by KrissMotors View Post
    The disadvantage of these LiFePO4 cells is that you are limited to system voltage below 130v. This eliminates certain high voltage motors, and probably makes impractical design choices like single gear transaxles.
    Why not use a step-up transformer to get higher voltages? I understand the controllers create a variable frequency waveform, making a variable frequency transformer necessary -- and I've never heard of one -- but why couldn't we power an inverter from the battery with a step-up transformer and a rectifier to power the motor controller? Bug zappers and Tasers both create high voltages without extra batteries.

  29. #29
    Senior Member RM1SepEx's Avatar
    Join Date
    Jul 2011
    Location
    Freeport, ME
    Posts
    2,598
    Quote Originally Posted by bromikl View Post
    Why not use a step-up transformer to get higher voltages? I understand the controllers create a variable frequency waveform, making a variable frequency transformer necessary -- and I've never heard of one -- but why couldn't we power an inverter from the battery with a step-up transformer and a rectifier to power the motor controller? Bug zappers and Tasers both create high voltages without extra batteries.
    Simply because NOTHING is "free" every time you use gearing for leverage, electronically do something, convert from DC to AC etc... there are losses involved that directly reduce the power that gets to the road vs what you suck out of the battery pack. Simple is more efficient...

  30. #30
    Senior Member Xusia's Avatar
    Join Date
    Oct 2011
    Location
    Eugene, OR, USA
    Posts
    2,340
    As I understand the question, it relates to using more powerful motors, not efficiency. When it comes to making big power - whether gas or electric - nothing can be called "efficient"! I think the answer to the question really is that there is nothing to be gained. Stepping up the voltage would either result in fewer amps at the stepped up voltage (as I believe would be the case in this particular configuration), or fewer miles out of the battery pack. The latter could be a worthwhile trade-off in some usage cases, but I don't think it applies to the battery configuration under discussion here.

  31. #31
    Quote Originally Posted by bromikl View Post
    Why not use a step-up transformer to get higher voltages? I understand the controllers create a variable frequency waveform, making a variable frequency transformer necessary -- and I've never heard of one -- but why couldn't we power an inverter from the battery with a step-up transformer and a rectifier to power the motor controller? Bug zappers and Tasers both create high voltages without extra batteries.
    Quote Originally Posted by Xusia View Post
    As I understand the question, it relates to using more powerful motors, not efficiency. When it comes to making big power - whether gas or electric - nothing can be called "efficient"! I think the answer to the question really is that there is nothing to be gained. Stepping up the voltage would either result in fewer amps at the stepped up voltage (as I believe would be the case in this particular configuration), or fewer miles out of the battery pack. The latter could be a worthwhile trade-off in some usage cases, but I don't think it applies to the battery configuration under discussion here.
    This topic gets complicated quickly. Bottom line: there is no realistic way to transform the voltage/current profile of a battery pack. As RM1SepEx says, efficiency loss is one reason. Each conversion step, even a relatively efficient one, loses at least 10% of the total energy, and that translates into a 10% mileage range reduction.

    But there is a more profound limitation imposed by physics, and I'll try to outline the technical issues below.

    TRANSFORMERS
    Transformers take advantage of changes in magnetic flux that induce current flow in a nearby conductor. So, if you run alternating current (AC) voltage through a primary conductor, the rise and fall of magnetic flux induces current in the secondary conductor. This is a typical experiment in high school physics class. The ratio of windings in the primary and secondary either steps up or steps down the voltage, and this is the basis of long distance power transmission that uses AC.

    Here's the physics limitation: the transformer requires a change in magnetic flux and you get that automatically with alternating current. But a battery pack outputs direct current (DC) with a constant flux. Thus, you can't use a transformer with a DC battery pack.

    CONVERTERS
    This does not mean that you can't change DC voltage. You can, but first you need to convert it back to AC, then run the current into a transformer, then convert back to DC. The device that does this is called a DC-DC converter. In fact, electrics cars routinely use DC-DC converters to take a little bit of the main pack voltage (say, 120v) and step it down to 12v for automotive relays, lights, dash instruments, etc. This avoids the need for a separate 12v battery and alternator/generator setup. For this rather simple job - usually about 40 amps at 12v - you need a DC-DC converter about 6x6x6" and it gets hot.

    But there are practical limits to DC-DC conversion. To run the DC motors we're talking about, you need about 360v at 300amps. Here's the math: 900 amps at 120vdc converts into 300 amps at 360vdc (actually, you would lose at least 10% in the conversion process, but I'm ignoring this inefficiency). First, the LiFePO4 pack I'm talking about can't generate 900 amps ... but assume for a moment that it could. You would then need electronics that can handle a 900 amp input. Probably two DMOC645s will do the job; they are extremely expensive. But you also need a transformer that can handle 900 amps. Typical residential power transmission lines have transformers that could do the job ... but look at them on the power poles! They're enormous, heavy (probably over 1000 lbs) objects unsuitable for automotive use. Laws of physics just won't let you get there unfortunately.

    The engineering lesson is that the battery pack must be designed to match the operating characteristics of both the motor/controller and the transmission/use of the vehicle. You can't transform or convert your way to a better place.
    Last edited by KrissMotors; 03-11-2013 at 02:50 PM.
    Developer of Comprehensive FFR Roadster Build Manual Mk4Build.com for FFR Complete Kit Mk3.1 & Mk4 (free)

  32. #32
    Senior Member bromikl's Avatar
    Join Date
    Feb 2011
    Location
    Milwaukee, WI
    Posts
    378
    Quote Originally Posted by KrissMotors View Post
    You can't transform or convert your way to a better place.
    Thank you for the explanation. Well said.

  33. #33
    EFI Rules and Carbs Drool Arrowhead's Avatar
    Join Date
    Feb 2011
    Location
    Stillwater, NY
    Posts
    365
    Quote Originally Posted by KrissMotors View Post
    One reason electric motors seem relatively expensive compared to gas engines is that they have been - up to now - mostly manufactured for industrial applications and they last nearly forever. There is no consumer market for electric motors per se. Also, a typical gas engine needs periodic maintenance whereas a typical electric motor can run for decades 24x7... without maintenance! So from a total cost of ownership perspective, electric motors look more competitive. Still, there is a price premium over what you'd expect.

    Hopefully as electrics penetrate the automotive space, prices will decline. It's still a cottage industry now, and that's the price of entry for early adopters. I certainly don't advocate an electric 818 as a cheaper alternative ... it's not at all. There will be a premium for quite some time, I imagine. But I think we are at - or very near - a tipping point where lightweight electric sports car will become increasingly attractive. Like any new innovation, you will likely see two steps forward and then one step back.
    I have no dog in this fight, in fact this the first thread I've read on the 818/electric subject. But I will throw in the additional money no one likes to talk about for a gas engine - dyno tune time, computer programing, coolant system, fuel system, misc repairs and part replacements on donor engines, etc. It all adds up. Not advocating one or the other, just pointing out that either way you do it, costs can run up pretty quickly.
    Arrowhead's '33 Hot Rod Build Site:
    http://sites.google.com/site/arrowheadshotrodsite/Home

  34. #34
    Administrator ehansen007's Avatar
    Join Date
    Mar 2011
    Location
    Aliso Viejo, CA
    Posts
    1,108
    Well said George. As for weight, most don't take into account the cooling system in which a fully loaded radiator can put on a extra 50lbs!
    www.33machine.com

  35. #35
    For those interested;
    The Siemens BorgWarner setup has just been discounted to 2500.

    ebay Item number: 261141303324

  36. #36
    Administrator ehansen007's Avatar
    Join Date
    Mar 2011
    Location
    Aliso Viejo, CA
    Posts
    1,108
    Jodie, can you tell us a bit more about this configuration and what your thoughts are for adaptation? Looks like it comes with a gear box as well. Is this CVT style or continuous setup with no shifting needed? It looks like it.

    Also, this is in the UK? Not sure how many here are in that neighborhood!

    I see some information here on the failed project but it doesn't give much tech info.

    http://en.wikipedia.org/wiki/Azure_T...nnect_Electric
    Last edited by ehansen007; 03-13-2013 at 10:56 PM.
    www.33machine.com

  37. #37
    Quote Originally Posted by ehansen007 View Post
    Jodie, can you tell us a bit more about this configuration and what your thoughts are for adaptation? Looks like it comes with a gear box as well. Is this CVT style or continuous setup with no shifting needed? It looks like it.

    Also, this is in the UK? Not sure how many here are in that neighborhood!

    I see some information here on the failed project but it doesn't give much tech info.

    http://en.wikipedia.org/wiki/Azure_T...nnect_Electric
    Eric of Kriss Motors covers a bit about it on comment #19 of this thread.

    If you can be patient, there is heaps of info on www.evtv.me that cover this package as a driveline. Their weekly news videos cover the initial selloff http://media3.ev-tv.me/news112312-1280.mov to flashing the inverter http://media3.ev-tv.me/news121412-1280.mov

    www.evtv.me also sell the same driveline on their store.

  38. #38
    Quote Originally Posted by RM1SepEx View Post
    It's actually always about the area under the HP curve (torque x rpm /5252 over time)
    Quote Originally Posted by BipDBo View Post
    Speak about torque all you want, but power is what actually matters. That's why we have gearboxes, to trade rpm for torque. I see that the peak is 67.21 HP at around 2200 rpm according to this chart. It would move around an 818 around town, but not quickly. This system might be more appealing for a hybrid build than a pure EV.
    75torque.jpg
    Quote Originally Posted by ehansen007 View Post
    I beg to differ on this one BipDBo. Just any one who owns a diesel how they feel about HP over Torque. And, while the subie engine might make the same amount of torque, it's only at peak. Maintained flat torque range using gears can be as effective especially with instantaneous torque.
    First and foremost it is great to see Eric providing such great data and ideas. People with real world experience provide something that those of us who have not built and EV car need.

    However, it should be clear that in the end, the overall horsepower makes a significant difference. It is true that one could construct a theoretical motor that provides better performance then a second theoretical motor, but both of those would be pretty unusual cases. Sure there are engines with higher HP but produce slower results, but that happens at the margins. You will not find too many 1000hp engines that can get beat by a 12hp EV motor just because of the low end torque.

    To help illustrate this, I put together a graph simulation of a couple of EV configurations. The first is a close representation of the EV motor mentioned here (the EV75) in an 818 chassis weighing 1800 lbs with a stock WRX gear box. I compared this against a stock STI chassis (3300 lbs) using a WRX gear box and a tuned 2.5L Subaru engine with a VF39 (just because I happened to have that dyno data in this spreadsheeet).. I could instead use a real WRX 227hp dyno run.. but the difference would not really be that important. Take a look:



    This shows the RPM and Speed of the above mentioned simulated WRX/STI Hybrid compared to an 818 with the EV motor. (the lower lines). NOTE: This simulation starts at ~ 10mph, and assumes zero shift time... so it is optimistic.

    As you can see, the EV car is not particularly fast. It gets worse if you put that engine in the 818:

    Here is a comparison if you put the WRX/STI engine in the 818 chassis.




    First, the 818 with that kind of torque could not maintain traction.. so the STI part is not realistic... but at higher speeds is shows the difference. 300hp is much much faster then 75hp... even with the 'low end torque' of the EV.

    Jeff

  39. #39
    Senior Member Xusia's Avatar
    Join Date
    Oct 2011
    Location
    Eugene, OR, USA
    Posts
    2,340
    That's really enlightening. There is one thing that's making me scratch my head, and I'm not sure if it's just the EV system described here or something else. I've seen EVs literally leave in the dust cars with engines that make significantly more HP than the STi described here (and weight about the same). This was mostly at speeds less than 100 mph, and because of the RPM limitations of motors, I assume said car would eventually catch up. The point is, there was no denying the acceleration advantage provided by the torque of the motor vs. a more powerful engine.

  40. #40
    Quote Originally Posted by Xusia View Post
    That's really enlightening. There is one thing that's making me scratch my head, and I'm not sure if it's just the EV system described here or something else. I've seen EVs literally leave in the dust cars with engines that make significantly more HP than the STi described here (and weight about the same). This was mostly at speeds less than 100 mph, and because of the RPM limitations of motors, I assume said car would eventually catch up. The point is, there was no denying the acceleration advantage provided by the torque of the motor vs. a more powerful engine.
    ...which the graph does not appear to show. There is no 'advantage' off the line shown here that we would expect from the torquey EV. Of course, this is for starting at 10 MPH, not off-the-line, but still. It seems 'off' to see such 'linear' speed curve comparisons when one vehicle is all torque, the other high HP.

Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

Breeze

Visit our community sponsor