DIY Electric Car Forums banner

1 - 20 of 33 Posts

·
Registered
Joined
·
11 Posts
Discussion Starter #1
So much great know-how here, my thanks in advance for allowing me to tap into it! :cool:

As background, I'm a classic gear-head and have built/tuned lots of cars for amateur motorsports (autocross and track), sometimes successfully :) For the past couple years I've been daily driving - and occasionally towing - with a Model X, and have fallen in love with the simple, effortless, and drama-free nature of EV power. For my next performance-car build, I'd like to do an EV.

While the Model 3 Performance does extremely well for what it is, the ability to bring performance to a higher level is fairly limited. What I'd like to do is start with a lightweight tube-framed chassis and body with performance-focused aerodynamics, and modify it to run a high-performance EV power system.

Specifically, I'd like the car to be able to handle 3 full laps at any normal race track (say 6-8 total minutes) without overheating and/or suffering major performance degradation. For those familiar, a Time Attack. It's also gotta be able to do 3 laps again about 3 hours later, where at best a 10-20kw charger will be available between session. The Model 3 in Track mode can do this today, the S/X cannot even complete one lap before drastically falling off.

My budget is probably higher than most DIY hobbyists, but not to where it's a business writeoff. What I've found is there's a huge leap in costs as you go beyond the mass produced stuff. A whole 100Kwh Tesla battery is about $25k, and the Model3 PM perf drive units are like $8k each. So for the big heavy parts, you're under $50k for a 3-motor system with inverters and reduction gears.

The next step up, everything gets like 5x as expensive? $25k motors, $15k+ inverters, and you still have to do the gear reduction yourself? I've asked a couple battery companies for quotes but am afraid at what I'll get back. Feels like it's going to be $250k, or 5x the top-shelf Tesla parts.

My going-in thought is a full 100Kwh worth of the 6.3kwh Tesla modules, and either 2 (1 front 1 rear) or 3 (1 front, 2 rear) Model3 PM rear motors. They've been shown to put out 270kw in the Model Y Performance. One session of 3 laps should draw between 20-40% of the charge, depending on the track. Would run as much AC chiller as I could to try to keep things cool enough for those 6-8 minutes.

I know there's a ton I'm missing here. An area of concern is whether or not the Tesla battery can supply sufficient amperage to run 3 of those PM motors. If not, at some point it makes more sense to stick with 2 motors as there's also a weight savings which helps everything that isn't outright acceleration.

There are probably a ton of other things too! Fill me in! For a budget of $50-75k, are there better options than the Tesla components?
 

·
Registered
Joined
·
5,056 Posts
This sounds like an interesting project. It also sounds like you at least started with the idea of using an Ultima as the base vehicle - is that still the plan? It's really not suitable for the AWD configuration.

Even with an optimal tube frame and minimal bodywork, anything with a full 100 kWh Tesla battery pack, three (or even two) motors and gearboxes, and all the required supporting equipment will not be light.

Is the plan for track-only use, trailered to the track? That makes a lighter car more feasible, by removing some road vehicle equipment requirements and avoiding an onboard charger.

I agree that using any commercially available aftermarket motors, controllers, or gearboxes massively increases the price.
 

·
Registered
Joined
·
11 Posts
Discussion Starter #3
Thanks for the reply brian_! Always enjoy reading your posts.

You are absolutely right, AWD will not be easy. Thankfully I am rather short and can probably live with the front firewall & pedals moved back several inches to allow for a motor to run in its natural orientation. If not, can look to run it inverted and deal with the consequences to cooling system space. In either case there remain the challenges of fitting uprights which can receive drive, and fitting a pushrod-style spring/damper system to make room for the halfshafts. Plus the darn steering rack, fun!

What I hope/believe would make all that worth it, is the ability to better accelerate in inclement weather (of which there is always some) and the ability to perform lots more regen.

The intended use of the car is the One Lap of America. It's a series of time-attacks (plus skidpad, autox, and drag-racing) held at different tracks over several days. I ran it last year in a Viper ACR and had a blast. There were two teams that year in Teslas (one a P100D, the other a M3P) who I spent a lot of time talking to. Bought a C8 Corvette to run it again this year, but the event was cancelled due to Covid. Sold the C8 yesterday (compared to an EV it's not that great!) so now have some budget, garage space, and thanks to Covid - time! available to get started on something like this.

And right again - yes an Ultima RS. I love how they look, and one should be just barely streetable enough for this use (the event has been done before in ICE Ultimas). A lot of their natural shortcomings (noise, heat) should be mitigated going electric. Some have been built with supercharged LS engines and AC at under 2100 pounds; I'm hopeful I could get a 100kwh (~1000lbs.) 3-motor (~600lbs.) one at or under 3000 pounds, or ~900 pounds heavier. The hope is that over time as battery tech improves, the car gets lighter, with longer range, and more power available to the motors.

One Lap has 500-600 mile transits between tracks each day where the car must be driven, it cannot be trailered. As I won't be able to use the Supercharger network, will need some kind of "power trailer". Have some thoughts on that but it could be a whole separate thread.
 

·
Registered
Joined
·
5,056 Posts
It looks like you have a good handle on the many issues. ;)

Yes, a battery or generator trailer is probably better in a separate thread, as long as it is clear here that the car itself doesn't need much range.
 

·
Registered
Joined
·
522 Posts
To go faster than a Tesla for the price of a Tesla involves starting with a Tesla. You're not going to see revolutionary kW-per-Kg improvements in batteries year over year...More like decade by decade.

I gotta wonder how much faster this Cobra is around a curvy track than a bone-stock Model 3 Performance with 1,000 pounds of interior and HVAC removed...and that's without trying to get more power out of the Tesla stuff (which people are doing with custom controllers and what not):


Another reality to consider is that, in two years, when you can get a Taycan for $75k, do you still want to be tweaking the power trailer integration or working out the bugs in the suspension of your custom rig...?

I, too, spend much money just to spend time building something fun...if that's a big part of your goal, have at it!
 

·
Registered
Joined
·
11 Posts
Discussion Starter #6
Have followed the Cobra for some time, great to see. Some of the design positions I hold are based in part on his experiences - using the PM motors instead of induction motors, and using water-cooled batteries, even if they don't offer the same "stiffness". Thankfully my use case of 6-8 minutes of high output is only about 1/3 of the Cobra's 20+ minute stints. I do want at least 2x the power though.

The Taycan is great in lots of ways and I have some friends who performed certification testing for Porsche in the Turbo S. For my use case the absence of a robust high-speed charging network would be a problem, and likely still necessitate some kind of trailer. Plus in a few years, it will still be well over 5000 pounds and look like a 911's overweight cousin :)

If the Model S Plaid or Roadster were available today, and particularly if they were available in some kind of hardcore track spec, I might reconsider my position. Not going to pull the trigger one way or another for a month or so, though I suspect we probably won't hear any big product announcements from Tesla until next year. The OE options will always be better thought through and engineered than what I could do at home, but they'll also always be a lot heavier and more compromised. If it weren't for the surplus of time these days (and for the foreseeable future) to do big projects at home, I'd be more inclined to shy away from the DIY approach
 

·
Registered
Joined
·
136 Posts
i'm a big advocate of ditching range but keeping power, and factoring in some dc fast charging into the build.
100 miles of range isn't so bad if you can fast charge in a pinch, it just won't road trip
but 100 miles of range worth of batteries is far cheaper and far lighter than like 300 miles of range worth

keep the weight low and the power high with an awd configuration, and you'll make an absolutely vicious track weapon
 

·
Registered
Joined
·
11 Posts
Discussion Starter #8
I’d be open to other batteries - especially if the overall package were smaller and lighter! And could be kept cool. It feels like the LG Chem stuff has potential but one would need to go to great lengths with all manner of custom cooling plates to approach the OE level of cooling the Tesla batteries have. And by then its size and weight advantages are minimized. With the three motors there is room to use a 400v battery than can provide like 2400 amps. It’d still need at least 50kwh usable to get through a session also.
I put in a request to the EV Drive company for something like what they built for the Palatov car, but have a feeling it’s gonna be a 6 figure deal, haven’t heard back yet.

Car will need some kind of range extending trailer in any case.
 

·
Registered
Joined
·
136 Posts
Chrysler pacifica hybrid batteries perhaps, or chevy volt batteries
both of those have very good power density, although not great energy density, and the volt cells at least have cooling essentially built in
 

·
Registered
Joined
·
1,007 Posts
We can build power dense battery modules for around $700 per kWh. That’s way more expensive than salvage packs, but no OEM batteries are designed for high sustained power, not even Tesla.
 

·
Registered
Joined
·
136 Posts
people are getting incredible performance out of Chevy Volt packs in terms of power density, you can absolutely smash amps out of them and they don't even blink
also the pacifica hybrid packs are about as good, ev west uses them in their 675ph pike's peak ev bmw
 

·
Registered
Joined
·
5,056 Posts
We can build power dense battery modules for around $700 per kWh. That’s way more expensive than salvage packs, but no OEM batteries are designed for high sustained power, not even Tesla.
A challenge in battery configuration is always finding the right size, shape, capacity, and capability of modules to suit the vehicle. I don't want details to fill up this thread if the builder (j-rho) is not interested in them, but is there a website or document with more information about specifically what configurations can be built? For instance, in a stack of pouch cells the cell dimensions constrain two dimensions of the module - is there a choice of cell sizes? The balance between capacity and physical pack size (for a given cell) is determined by how many cells are used in parallel - is that selectable by the customer? Power density is limited by (among other factors) the cooling system - are modules available with a choice of interleaved fluid-filled fins, or set up for chill plates on one side or two? I don't see any of this on the EV Drive website; the Battery Modules product page just says "Product listing coming soon…".
 

·
Registered
Joined
·
5,056 Posts
I’d be open to other batteries - especially if the overall package were smaller and lighter! And could be kept cool. It feels like the LG Chem stuff has potential but one would need to go to great lengths with all manner of custom cooling plates to approach the OE level of cooling the Tesla batteries have. And by then its size and weight advantages are minimized.
Almost everyone using Tesla (any model) and Chevrolet Volt modules uses the OEM cooling system, because it is internal to the modules. This is certainly an advantage of these modules, both for performance and for ease of installation.

It appears that almost no one using the modules from any OEM EV with a chill plate system (including the LG Chem modules from the Chrysler Pacifica) uses the OEM chill plates. That makes sense in the case of a single chill plate for a large under-floor pack which doesn't fit in the conversion vehicle, but the Pacifica has a relatively compact pack, and it doesn't seem like anyone has even looked to see if it is one big chill plate (which might still fit some projects) or perhaps two 3-module plates. It might be worth having a look.
 

·
Registered
Joined
·
1,007 Posts
A challenge in battery configuration is always finding the right size, shape, capacity, and capability of modules to suit the vehicle. I don't want details to fill up this thread if the builder (j-rho) is not interested in them, but is there a website or document with more information about specifically what configurations can be built? For instance, in a stack of pouch cells the cell dimensions constrain two dimensions of the module - is there a choice of cell sizes? The balance between capacity and physical pack size (for a given cell) is determined by how many cells are used in parallel - is that selectable by the customer? Power density is limited by (among other factors) the cooling system - are modules available with a choice of interleaved fluid-filled fins, or set up for chill plates on one side or two? I don't see any of this on the EV Drive website; the Battery Modules product page just says "Product listing coming soon…".
I don't want to turn this thread into an advertisement for our modules, but I'll do one post answering your questions since they could be relevant to OP. Let me know if you want me to delete these, OP.

Our modules use 21700s, so one dimension is always about 76mm. Apart from that, the small size of the individual cells gives a lot of dimensional flexibility, and our method is such that we can do different configurations fairly easily. We usually embed BMS in each module, and so in the past we have typically done 12S modules to fit that, but we're developing a 14S BMS so we'll be doing more of those in the future. We have done anwhere from 30Ah to 144Ah but there aren't too many limitations there, apart from the modules becoming too big to be managable. We have a high power cell that we usually use but can use energy cells as well if really needed. As far as cooling goes, we pull the heat out to aluminum surfaces such that it can be cooled via liquid coldplates or air fins or just by conduction to the battery box.

Sorry about our website, we've been working on a bunch of projects the last couple of years and have been slow getting the standard products out there. Two projects that have used this module design have been the Palatov/Cascadia racecar that did Pikes Peak last year. That one had a 79 kWh and dynoed 1300 hp at the wheels. Also the Webasto Mustang Lithium from SEMA last year had them, though it has not been tested to full power yet due to a motor issue.
 

·
Registered
Joined
·
11 Posts
Discussion Starter #15
Please type away Hollie Maea! I read your posts in the Palatov Pikes Peak thread, which lead to me filling out an interest for on your website. The price is a bit more reasonable than I'd anticipated. Is there a way I can contact you or your company for more info?
 

·
Registered
Joined
·
1,007 Posts
Please type away Hollie Maea! I read your posts in the Palatov Pikes Peak thread, which lead to me filling out an interest for on your website. The price is a bit more reasonable than I'd anticipated. Is there a way I can contact you or your company for more info?
You can email me directly at PAULL AT EVDRIVE DOT COM
 

·
Registered
Joined
·
7 Posts
Someone correct me if I'm wrong, but you might be able to keep the LS in addition to the electric motors? Essentially your setup looks like this:
  • Normal EV drive setup, likely 2 Tesla rear performance motors, one at each end of the car
  • LS mounted using stock mounts, intake, ECU, everything, likely tune for torque vs power
  • Instead of a transaxle hook up a generator to the LS. Take the power from the generator and pour it into the motors.
  • Have a small 100+ C battery (a few kWh) to absorb the generator output differences and to give maximum power to the motors when needed, and use the LS to recharge these batteries and power the car in every other situation.
With this setup, instead of having 1000lb of batteries or more you now have a 500lb LS (with fuel tank, intake, etc), 100lb of batteries, and a 200lb generator. Plus you can refuel in minutes vs hours.
 

·
Registered
Joined
·
136 Posts
that's a serial hybrid setup, if the engine isn't connected to the wheels
and at that point you're much better off with an engine designed to do that, like a chevy volt engine
but there's a lot of extra software and controlling needed for that kind of setup
 

·
Registered
Joined
·
7 Posts
Well I suggested the LS because there are versions designed for torque, it fits in the Ultima chassis without modification, and the Chevy Volt engine isn't powerful enough for what the OP is suggesting. Of course it could work with any sufficiently powerful engine.
 

·
Registered
Joined
·
5,056 Posts
Our modules use 21700s, so one dimension is always about 76mm. Apart from that, the small size of the individual cells gives a lot of dimensional flexibility, and our method is such that we can do different configurations fairly easily.
I guessed that you were using pouch cells based on the website images - apparently I guessed incorrectly, because I didn't notice the high parallel cell count indicating the small cells.

As explained above, that means that the module thickness is constrained, with the other dimensions being relatively flexible, although to avoid every customer getting a custom product there would need to be some standard sizes.

We usually embed BMS in each module, and so in the past we have typically done 12S modules to fit that, but we're developing a 14S BMS so we'll be doing more of those in the future.
I think it would be nice to have the option of fewer than 12 cell groups in series, to allow for more modules (for the same total capacity and pack voltage) for more flexibility in packaging without making every module uniquely shaped to fit its mounting location.

As far as cooling goes, we pull the heat out to aluminum surfaces such that it can be cooled via liquid coldplates or air fins or just by conduction to the battery box.
The cooling design sounds like the same approach as Rivian.
 
1 - 20 of 33 Posts
Top