DIY Electric Car Forums banner

1 - 20 of 25 Posts

·
Registered
Joined
·
4 Posts
Discussion Starter #1
Hello -

Obviously I'm new to the forum. That said, I am not new electric vehicles, mechanics, or electricity. I got stuck on the battery configuration, so thought to share my design and get some feedback. The battery bank makes the price of the vehicle go through the roof, so I'm looking for alternatives.

My goal is to build a "poor man's" Rimac Concept One, with rear dual wheel driven AMR Dual Stack 250-90 AC motors. For the controllers, I guess the best bet is 4 x RINEHART MOTION SYSTEMS PM150DZ 720V 300ARMS. I'm going to start with a custom, reinforced, open, stainless steel chassis and build it from the ground up. Eventually I want to put a carbon fiber shell on it, but open air frame is fine for now.

I like the dual in-line AMR motors, but would consider options. Any input would be much appreciated.

Forgive my design skills. Here's a mock-up to show where I am at (google doc):

https://docs.google.com/presentation/d/1Rav1jdP223F6-tJhoA-cFWHhEPcTYicu2K4jMVbbtOA/edit?usp=sharing

And the image below....
 

Attachments

·
Registered
Joined
·
934 Posts
Go AWD to get all the torque down.

http://www.evdrive.com/products/evd-motor-controller/

Going for an package deal would be good. However sinking around 100.000$ in the drivetrain only is insane, when only going for rear wheel drive.

Id go for four motors per wheel.

However looking at getting max power to the motors, roughly 800kw total (depending on if you can get it down and other specs).

8c is do able with alot of cells however, depending on what you want to spend the weight will be determined. If you really want to get 1.6MW out of a 100KwH pack will take 16C, this will limit your battery choice drastically.

But then again when are you ever going to draw 1.6MW, and four of the mentioned rineharts wont maken 1.6MW.


It all comes down to budget and hard requirements.
 

·
Registered
Joined
·
1,007 Posts
This is doable, but would be very difficult and very expensive. Not to be a downer, but if you are worried about the price of the batteries, your budget expectations are probably not calibrated properly. I work for a company that actually has all of the technology developed that would be able to do such a car, and we don't think we could do a prototype for less than $250K (Edit: that would be for AWD) even though we wouldn't have to pay for our own NRE.

One thing to consider: You can get the same power from a single core AMR motor and a PM250DZ that you can with a dual core with dual PM150DZs.
 

·
Registered
Joined
·
4 Posts
Discussion Starter #4
Thank you both for fast response.

I realized the controller issue after I posted.... I meant to have the PM250DZ in there, not 150. It appears the PM250DZ will have 450ARMS continuous & 600 ARMS peak, which mostly matches the motors.

I'm fine with a $500k+ price, half the published price of Rimac Concept One, which is poorman-ish, I know. What's concerning is that 100kWh of batteries appears to be $125k+, before any cooling or labor. Finding a smart way to spend $125k on batteries is what my post was after.

Also appears that as Tomdb said, direct drive to just the back wheels is probably a bad idea, so let's double the motor / controller spend and put 2 up front.

So... 100kWh of batteries is only going to last me about 3 minutes at 2 MW consumption. Unlikely I can build a chassis that will sustain that power.

I guess overkill on the motors is good if you can intelligently balance the power consumption across the wheels. Is there a DIY controller to control the 8 controllers?

Roughly speaking, $330k in motors, controllers & batteries.
 

·
Registered
Joined
·
934 Posts
If you can keep the weight down I think even when spending less then your mentioned 500K. You can get a car that will blow the doors off anything. :D

Single remy 90r core with 250dz http://media.wix.com/ugd/28f2e3_a7428462d2824ec9a1c2a3c813fa850e.pdf

Four of these puppies with an planetary reduction on them, ratio dependent on desired top speed / acceleration balance.

I would strive for a car below 1700kg
-300kg drivetrain (motors inverters gearboxes ect.)
-700kg battery (high discharge 100KwH)
-700kg frame/car (adapated kitcar/replica/custom)

you would require a very big box of adult diapers when tuning the controllers on a puppy like this.

Controllers there are some smart guys out there that might help out with systems for torque control that act like an ABS, so not allowing one wheel to accelerate faster then others.

Are you looking for top speed or acceleration?

quick excel sheet calcs show roughly 300 km/h in 10 seconds, when geared for 300km/h top speed.

When geared for 450 km/h it comes to 15 seconds to 300 km/h and 31 seconds to 450 km/h.
 

·
Registered
Joined
·
74 Posts
So... 100kWh of batteries is only going to last me about 3 minutes at 2 MW consumption. Unlikely I can build a chassis that will sustain that power.
Not to mention that with 2000 kw in about 10 seconds you'll be doing 500km/hr and in the process of total aerodynamic destruction.:)
 

·
Registered
Joined
·
74 Posts
On a more serious note.

Why not start designing the car around a Tesla model S 85kwhr pack. I know that it probably won't deliver the kind of power you are after but it's a start.
 

·
Registered
Joined
·
389 Posts
I would buy two Tesla power packs and run them in series to power four Remy motors powered by reinheart motion systems PM150DZ inverters setup for torque vectoring, if google serves me well a fully charged tesla pack produces 393 volts so two are 787 volts, which is just above the maximum rated voltage of the remy motors 700 and below the controllers 800v limit which is perfect.

the torque vectoring would be quite difficult, but i think could be very worthwhile.

if anyone has a cool 500k lying about the place gathering dust i would like to have it so i can build my car as well please, thanks
 

·
Registered
Joined
·
2,170 Posts
I think there is some number/units confusion going on or the project name is off.

100kwh = 360 megajoules

If you can sustain 278 watts (not kw) for an hour (1 amp @ 278v), you will have a megajoule.

A single ~$100 leaf module is 1.6 megajoules. So about $25k for 100kwh worth, 225 modules and 1900 lbs. I imagine the $25k could come down with some resourcefulness.
 

·
Registered
Joined
·
126 Posts
I think there is some number/units confusion going on or the project name is off.

100kwh = 360 megajoules
I think he meant megawatt, since he's also comparing with Rimac.

@noah: looking at your google docs, I see you've labelled the blank space between the motor and the wheel as "differential". what you need is actually a gearbox (single speed or whatever you want), because a differential will not work if only 1 side is connected. Yes in essence it's a gear reduction, but it doesn't work the same way.
 

·
Registered
Joined
·
4 Posts
Discussion Starter #11 (Edited)
Thanks for great feedback guys. Just what I need to keep the ball rolling. Some thoughts...

The Name: I was in a pinch on the name and went with megajoule because it sounds pretty good. Also considered centawatt, centiwatt.. None of them are that great. I guess it could be gigajoule (rounding up, save some room for upgrades).

Tesla 85kWh battery pack: Would be great to modify 2 of those, put them in series and get to 700V. I will dig in to that option. My guess is that Eon will not be thrilled with selling his packs off for DIY supercars. I started looking at that option, but couldn't find them regularly available to non-Tesla owners. Will look again. That would save some bucks and be a rather elegant solution.

Battery alternatives:
Without the Tesla option, my guess is that I am more than likely bound to 32 of these at ~$5k each:

AMR 26V 120AH HIGH POWER AIR COOLED LITHIUM BATTERY MODULE
http://www.elmofo.com.au/lithium-ba...-power-air-cooled-lithium-battery-module.html

Rounding up to 60 lbs each battery, including the cable mgmt + cooling mechanism, we are at 1,800 lbs for 100 kWh and $160,000. This is the rub that made me start this thread.

Speed: Thanks for doing the math Tomdb.. I was pulling my hair out this afternoon on the watts, HP, torque, weight, speed, tonnage and realized that the power / motor config is far beyond what the chassis could sustain out of the gate and top-end, assuming we allowed 500+ lbs of torque per wheel. Thus the power, batteries, and drive train will not be the limitation and keeping the chassis & supportive components from breaking is the inhibitor.

Keeping her sub 2.5 second 0-60 MPH would make me happy. Not sure I personally want to break 150 MPH top-end anyway, but having a qualified driver make the run would be fun.

Top end vs. off-the-line: Ideally, we flip a switch and optimize on-the-fly. I assume the controllers are smart enough to do something like that. But the controller to bring it altogether is the kicker. Not sure how that ties into the planetary reduction unit.

Planetary reduction:
Wow, how to kill 2 hours - google "direct drive planetary reduction". I have a ways to go there.

Weight / price:

What I have so far:

Batteries, wire assembly, cooling (?): Quantity 32 x 60 lbs each = 1,834 lbs (831 kg)
- price 4,600 each = $148,005 (bleh... 2 x Tesla @ 90k for 170 kWh would clearly be the way to go)
- AMR 26V 120AH

Motors: Quantity 4 x 242 lbs each = 968 lbs (440 kg)
- price 26,000 each = 104,000 (assumed a healthy discount)
- AM Racing AMR Dual Stack 250-115 AC Motors

Controllers: 8 x 22 lbs = 176 lbs (60 kg)
- price 9500 each = 76,000
- 8 x RINEHART PM250DZ 720V

Total weight so far: 3,074 lbs (1,394 kg)
Total price so far: 328,325

So based on Tomdb's weight schedule, that leaves me roughly 600 lbs to round out the chassis, etc. I think this works before we stick a fiber glass body over the open frame chassis. My guess is that it will get closer to 4k pounds.
 

·
Registered
Joined
·
2,170 Posts
what about calb cells?

I think your controller will need 800v to get close to its rated output fyi. which means you need ~250 cells (of 3.2v chemistry), that can deliver ~1250 amps to get to a megawatt.

CA60FI can be had at many places for about $80 apiece. a 250s2p setup would be ~$40k (can contact calb directly for a quote)
and weigh about 2200 lbs and provide about 1200 amps for short bursts (30 seconds)

I have no idea what is inside that $5000 ELMOFO box... $600 worth of calb cells and a fan?

edit: found a little info, not from the source though:
https://www.solaronline.com.au/elmofo-amr-24v-120ah-lithium-ion-battery-module.html
Discharge:
1800 A (10 sec)
960 A (Continuous @ 23°C)
Kokam Li-Po cells, not lifepo4 :/
 

·
Registered
Joined
·
1,007 Posts
Ok, so here is some advice:

1. Forget about 8 motors. There isn't a battery available with enough power density (yes, power density. C rating by itself is meaningless-- a cell that is 100Wh/kg and 15C is worse for power than one that is 200Wh/kg and 10C) to give you that kind of power and still fit in a car. Four PM250s is really the upper practical limit. (Disclaimer: if you were making a drag car, you could probably get a pack from John Metric or something that would technically give you those power levels for a couple of seconds, but you aren't building a drag car)

2. Forget about Tesla cells. They are energy cells, and don't have nearly enough power capabilities. Yes, you could get close on paper to having enough power by seriesing two packs together, but that's way too big of a pack. The cell weight alone is 1400 lbs and the volume is way too great--one pack alone barely fits in a Model S, which is a huge car.

3. The Elmofo packs are nice (also made by AMR) but super expensive...starting with Kokams, which are about the most expensive cells out there, then putting them in CNC cases. They do have good power density though.

4. Don't even think about CALBs. They are cheap and easy, but they are a great example of a cell with decent C rating but poor power density. To fully utilize your motors, you would need 800 CA60 cells. That's a cell weight of 3500 lbs.


I would restrict my battery choices to these three options:

1. Kokam pouches. Very very expensive, but they have good power density. You could try to save money by engineering and building your own pack instead of getting the AMR ones from Elmofo. You will need liquid cooling though.

2. Power 18650 cells. Look like Tesla cells, but with half the internal resistance. You could get by with a pack with about as many cells as a Tesla pack. These cells are much much cheaper than Kokams, but harder to make a pack out of. You might have to hire someone to engineer and build it, which will be expensive, at least for the first unit.

3. RC Lipo cells. John Metric sells these. They are kind of a poor man's version oft he Kokams. They aren't very safe and they won't last very long, but they are powerful and cheap.


Honestly, the batteries aren't the biggest challenge. The systems engineering for such a drivetrain (assuming you have a suitable roller, which is itself a major challenge) is not a trivial task. If you hired the company I worked for to design and build this for you, our quote would look something like this:

Motors: 4x $8000
Inverters: 4x $15000
Custom gear reduction: 4x anywhere from $10k to $50k depending on how many bells and whistles you want.
Battery: $1000/kWh for first engineered unit. Half that for subsequent units.
Chargers: $5000
$20k for misc ( your fuses alone cost $2500)

And then probably another $50k for NRE.

You can shave $100k off by buying the parts yourself and building it yourself, but it's not trivial. Do you have a VCU that can do four motor torque vectoring? Are you comfortable building a liquid cooled battery that won't arc over at 800V? Do you know how to make a gear box?

Anyway, if you are serious about being fine with a 500k price tag, it's probably doable if you have a roller. But you have to be serious about it...the company I work for charges $6k just to do a formal quote. And we're fairly cheap as those things go. I can only imagine what AMR charged Rimac to design their drivetrain.
 

·
Registered
Joined
·
1,664 Posts
Honestly, what are you hoping to achieve with this ?..
...other than spending a small fortune and taking a lot of time to do it.
..Keeping her sub 2.5 second 0-60 MPH would make me happy. Not sure I personally want to break 150 MPH top-end anyway
.... for what you estimate you will spend on motors, you could buy a 85 D P ++ spec Tesla S, and have some super smart tech guy tweek the software to give you the same performance !
Some of the remaining $350k you could use to have a customised body fitted by a pro body shop.
 

·
Registered
Joined
·
934 Posts
The biggest issue here will be the time that will need to be invested in this one-off. Time for design and manufacturing will cost you the most money.

The cost of parts becomes trivial at this point. Unless you got a roller built already this will take more time then designing the electrical part.

If you are really serious about this I would look into partnering with a engineering firm.

BTW. where are you located?
 

·
Registered
Joined
·
251 Posts
Why dont you go for two Tesla Performance drivetrains.
Get from wrecked cars.
EVTV is showing really good progress at controlling the P-drivetrain.

Each of the tesla drivetrain is about 150kg

With two of them it will probably move..

You would not have torque vectoring with that setup but it would probably be as fast.

I think that the Pike Peaks winner is an intresting car to look at.

Millens car had AWD but with normal final drives.

The YASA motors could also be of interest to you.

The Koenigsegg Regera is going to use YASA motors.
They are stackable.
Millens car had six YASA motors if im not mistaken.

Stacked in two stacks with three motors in each stack.
One stack for the FWD final drive and one stack for the RWD final drive.


Look at these batteries for a nice Tesla like product.
https://endless-sphere.com/forums/viewtopic.php?f=31&t=71951

If you get that module done with 3500mah cells you get 105ah/module.
Pair 2 and have 120 in series for 440v nom and 210ah giving about 91kwh
at about 360kg and good terminals for easy mounting.

Price of the battery about 240 pcs time say $200 gives some $48000

Should be a good start.


REgards
/Per
 

·
Administrator
Joined
·
6,156 Posts
Another alternative
Buy a new Tesla "Ludicrous" and make a light weight two seater body chassis but use all of the Tesla drive train - think about using the suspension and possibly the floorpan

You can probably lose 25% to 40% of the weight of the Tesla if you are not interested in creature comforts
 

·
Registered
Joined
·
389 Posts
Check out http://www.evdrive.com/ they can supply the reinhart controllers and motors of your choice and they can also provide you with a controller for torque vectoring "off the shelf" so to speak which will save alot of time!

I have to say i kinda agree with what people have said about tesla cells and packs, it would probably be good for a road car as it would go long distances but you do want high discharge cells.

The new Panasonic prismatic batteries that are designed exclusively for EV use would be far easier to engineer into a pack
i haven't got a data sheet but im guessing they are quite powerful
they are used in the prius plug in and some Audi electric cars. I have bought some of these cells myself.

As they are Panasonic you are buying a quality product that you know will be reliable. maybe you should contact a battery company directly and tell them what you are looking for, you may be able to put a sticker on the side of your car and get a cheaper price.

there is a company called Power Japan Plus who produce a "dual carbon battery" that looks very promising indeed but they are very much in the r&d stage i believe.i would talk to these people or Panasonic and see what they say.
PJP is testing their batteries in a racing car as they have very good thermal stability.

stay away from RC stuff unless its top of the range, there is a company that make very good cells but i forget their name maybee someone here knows?
 

·
Registered
Joined
·
389 Posts
Check out http://www.evdrive.com/ they can supply the reinhart controllers and motors of your choice and they can also provide you with a controller for torque vectoring "off the shelf" so to speak which will save alot of time!

I have to say i kinda agree with what people have said about tesla cells and packs, it would probably be good for a road car as it would go long distances but you do want high discharge cells.

The new Panasonic prismatic batteries that are designed exclusively for EV use would be far easier to engineer into a pack
i haven't got a data sheet but im guessing they are quite powerful
they are used in the prius plug in and some Audi electric cars. I have bought some of these cells myself.

As they are Panasonic you are buying a quality product that you know will be reliable. maybe you should contact a battery company directly and tell them what you are looking for, you may be able to put a sticker on the side of your car and get a cheaper price.

there is a company called Power Japan Plus who produce a "dual carbon battery" that looks very promising indeed but they are very much in the r&d stage i believe.i would talk to these people or Panasonic and see what they say.
PJP is testing their batteries in a racing car as they have very good thermal stability.

stay away from RC stuff unless its top of the range, there is a company that make very good cells but i forget their name maybee someone here knows?
 

·
Registered
Joined
·
4 Posts
Discussion Starter #20
Thanks for all the great suggestions / ideas. I have some homework to do.

Building a battery sounds like a good approach. Panasonic batteries look good for engineering into a custom pack.

The modular 18650-prismatic battery sounds like a good approach.

Buying 2 Tesla packs & putting them in series sounds like a workable approach, assuming a) can find them b) can fit them. The density of them is better than the ARM config I came up with. It looks like the only way to acquire the tesla batteries is out of salvage vehicles. Not a bad approach, to start tinkering.

YASA-750 looks very nice, with the integrated controllers.
 
1 - 20 of 25 Posts
Top