Tesla P85 Powered 818C

[kimchikimkim]

Kim3, I read about regen braking in terms of recovering energy but not pure brake performance, like distance to stop vs speed or G-force developed.
In 2WD I speculate brake coordination with the non-driven axle is challenging. Even cumulative effect of regen + friction brakes would be a challenge. Worst case being a panic stop or downhill into turn 5 at Elkhart Lake at 150MPH. I read that F1 has done it but at what cost?
Coordination with ABS?
jim

I do not know nearly enough to comment on this, especially on the driving characteristics of regen but I know that regen is not made to be used to stop extremely fast. I believe this comes down to the huge inductive surge caused by slowing down the rotor and limited space for dc link cap (again not sure but that’s what my intuition thinks). I would assume F1 has a pretty sweet inverter powering the motor and can handle a huge amount of bemf. The bottom line is that I will only be using regen to coast to a stop (as soon as my foot leaves the gas pedal it will be activated). If I want to slow down faster I tap the brakes and rest of the stopping power is all mechanical. As far as traction control goes I will have none (at least until I can afford going AWD).

[kimchikimkim]

Update:

Today I was able to finish pressing the hubs and bearings in the four corners. The bearings feel super smooth and so much tighter than the wrx knuckles with 200k+ miles on them. I cleaned up the garage making more space. I also designed a battery shelf that I will continue to fabricate throughout the week and weekend. I also purchased a ton of new bolts off of mcmastercarr to replace the wrx bolts since a majority were severely overtorqued and the zinc was wearing off one of the many previous owners. I wanted to replate them but my dad told me about hydrogen imbrittlement and it scared me enough to shell out money to just get new fasteners. Here are some pictures of the progress!



DAA39E6E-0B88-4F75-B0AD-B6021DE5C53C.jpg121040BE-F4BB-441B-9F38-F15F9FACC4B6.jpg89426EDA-BD84-4FA1-B670-7B4263F72989.jpg

[Sgt.Gator]

Saw this and thought of our EV 818 folks:

1,500 ft-lb Torque Electric Powered Shelby Cobra Tesla KILLER

[BrettR]

Hi Kimchikimkim, Very Nice,

What sort of range do you think you will get.

Regards
BrettR

[kimchikimkim]

Hi Kimchikimkim, Very Nice,

What sort of range do you think you will get.

Regards
BrettR

I just got another full battery pack from a 2016 Chevy Volt (gen 2 LG Chem based) for only 1k and it had 70k miles. If I can fit it, I'm hoping for 200miles to 250miles.

[DSR-3]

Nice.
That's like double my range on gas, and E85... don't even ask!

[J R Jones]

Tesla love leads to lots of Tesla parts and Valhalla:
https://www.yahoo.com/news/watch-tes...161712290.html

Scandinavian rage = 66lb explosive and overkill.
jim

[kimchikimkim]

Tesla love leads to lots of Tesla parts and Valhalla:
https://www.yahoo.com/news/watch-tes...161712290.html

Scandinavian rage = 66lb explosive and overkill.
jim

I wish he could have let me take the OBC and some of the battery modules before he did that! It really hurts to see someone be so wasteful for views and just blow it up like that.

[dlud]

Actually most of the good stuff was removed from the car before it was blown up. Someone posted on another site all the stuff he was able to pull off the car. He tried to convince the owner to let him have/buy the car but he was insistent on blowing it up. https://www.diyelectriccar.com/threa.../#post-1081208

[Bob_n_Cincy]

Actually most of the good stuff was removed from the car before it was blown up. Someone posted on another site all the stuff he was able to pull off the car. He tried to convince the owner to let him have/buy the car but he was insistent on blowing it up. https://www.diyelectriccar.com/threa.../#post-1081208

If they would have blown up the battery pack. There would be an enviomental mess. Burning Lithium-Ion batteries give off Tixic Fluoride gas. https://www.nature.com/articles/s41598-017-09784-z

[J R Jones]

If they would have blown up the battery pack. There would be an enviomental mess. Burning Lithium-Ion batteries give off Tixic Fluoride gas. https://www.nature.com/articles/s41598-017-09784-z

I did contract development engineering for a popular tool manufacturer, specializing in battery power. I discovered a technical program about Lithium-Ion batteries on PBS and brought it to the department manager. He ignored it.
https://www.pbs.org/video/nova-search-super-battery/

Nonetheless the battery lab had run-away fire incidents four times while I was there. The building evacuated to the parking lot and the fire department did their job. The rate of conflagration and temperatures were unreal.
PBS/Nova covered future battery technology that is safer.
jim

[kimchikimkim]

Update:

I have mounted all the batteries in the car! I'm ordering a bunch of parts to start to wire up the car.

I have changed direction with what onboard charger I am using after picking up a tesla Gen 2 OBC on eBay for $225. This will facilitate 10kw of charging power meaning I can charge the car in about 6 hours off a Level 2 EVSE. I need to figure out how the internal CAN bus works on it but looks like most of the CAN messages are available on open inverter forums.

On the downside, this project is going over budget so I will likely have to start to rethink a lot of design decisions and sell some of the parts I currently have! I still have the full set of WRX spindles, brake calipers as well as a lower control arm and linkages so if anyone Is looking for some I would be more than happy to sell them off.

IMG_2495.jpg

IMG_2514.jpg

[J R Jones]

Kim3, That is a lot of battery. What is the total battery weight? Have you measured corner weight and predicted suspension spring rates?
I assume they can be removed easily for service access?
jim

[mikeinatlanta]

Yeah exactly. That being said I don't think the brakes will be doing much as regen should take care of most of the braking 90% of the time. That's the main reason I went with VHT, it is rated for higher temps than a powered coat at the cost of a surface that isn't nearly as durable. If it starts to give me problems down the road I will fix them but for now, I will keep them as is.

Awesome project. A bit of advice on the VHT. It will only meet the ratings and be tough/chemical resistant if it goes through the heat cycles mentioned on the can. A long bake at lower temp will work as well. IMO better than powder after the heat cycles.

[kimchikimkim]

Awesome project. A bit of advice on the VHT. It will only meet the ratings and be tough/chemical resistant if it goes through the heat cycles mentioned on the can. A long bake at lower temp will work as well. IMO better than powder after the heat cycles.

Yeah I threw them in the oven. I appreciate the reminder though.

[kimchikimkim]

Kim3, That is a lot of battery. What is the total battery weight? Have you measured corner weight and predicted suspension spring rates?
I assume they can be removed easily for service access?
jim

I haven’t even thought about that stuff yet I’m so deep into the electronics that I’m holding all of this stuff for later. Batteries should be about 950ish pounds alone.

Each Chevy spark Ev gen 2 module is approx 53 lb

And for they aren’t super easy to take out but realistically they won’t be removed much

[kimchikimkim]

UPDATE:

I went to the local wrecking yard today and picked up a CCS Charge port from a Chevy BOLT EV. I plan on adding CCS combo support down the road but will use it for J1772 for now. I paid $200 for the port and the HVJB ( a snag considering how expensive the ports alone are from other sellers. I was super tempted to pick up the QAC7000 greenphy module (part K190) that handles all of the PLC for EVSE signaling in DIN 70121 or maybe ISO 15118 (haven't looked into it super deep yet) but passed since I wasnt trying to push my luck with the auto wrecker.

I took apart the HVJB and noticed that they have an opto-isolated HV present circuit using a resistor divider powered by a TI isolator IC. This is super similar to the same voltage sense found in the BMW i3 Fast charging "KLE" module meaning I might be able to use this circuit in a similar sense with the LIM which CAN DBC is already hacked for me! For Christmas, my parents got me a Saleae Logic 8! If I end up going back to the junkyard I will take the K190 PLC module and hook it up via serial to my logic analyzer and see if I can decipher anything.

I also mounted the DC/DC Converter and started plumbing up the front batteries, DC/DC, etc.....

I am waiting on the HV traction cable from IGUS. I asked for a sponsorship for some 2/0 chainflex cable which is super high quality and shielded so I'm waiting on a response. With so much current going through these wires shielding is a must for signal integrity of low voltage harnesses! It would be close to $450 in HV cabling alone so if I can get it for free it would really help the budget.

On a side note, if I'm not updating as much it's because of a busy week in school coming up.

IMG_2566.jpgIMG_2561.jpgIMG_2560.jpgIMG_2569.jpgIMG_2574.jpg

[kimchikimkim]

UPDATE:

I got the sponsorship through the Igus young engineers support program. They sent me over $500 in HV traction cable and I am super thankful for it.

I also landed a sponsorship with Alphacool in Germany for radiators, low-power d5 pumps, and really nice delrin/glass reservoirs. I will be using 10x 420mm aluminum radiators from them to cool everything in the car. It should be overkill which is definitely not a bad thing considering the AC induction motors' tendency for the rotor to get pretty toasty due to rotor currents!

I worked on the drive unit contactor box and have a pretty good idea of how it's going to look (photo below). This box will contain only stuff related to driving the motor. Sadly the wrecking yard where I bought the non economized GV200 contactors decided to use an impact wrench on them and the underside of the stud threads is busted so I need to buy new contactors. This time I'm just getting GV200MA's new since I don't want to deal with creating my own coil economizer and transient expression system.

I am using a Rincon Power 400a service disconnect which is going to reside in the center column between the seats. This will be the main HV junction box that each set of batteries will lead up to. Since I have 3 "sets" of 96s groups of batteries I will have 3 leads going into it. On the output side, I will have a lead for the charging box, and another for other high voltage systems (such as PTC heater, AC compressor, DC/DC Converter, etc...)

That's all for now.

On a side note, I decided to take a little bit of a break and make some money by buying a Nissan leaf 2011 with 75k miles on craigslist. I paid $550 for it and tore the whole thing down today. I plan on keeping parts but pick n pull will give me $475 for the shell so it's basically a no-brainer. It's pretty interesting but the battery pack but terribly designed (no active thermal management). No clue what the Nissan engineers were thinking but clearly not much....... I'm going to test the battery pack's capacity (I'm praying that it is good). There is a boatload of parts that I will use to upgrade the first electric car that I built in 2016. I plan on using the motor, inverter, and DC/DC converter to make my Locost a serious rocketship. If the battery pack is good I will use that too (and I will design a forced-air cooling system for it as well). I've gotten a little too sucked into the 818 project so stepping away for a few days has been really good for me! ​

Attachment 160102
Attachment 160103

[kimchikimkim]

UPDATE

I got my hands on a Prius electric power steering column at pick-n-pull (I also got an inverter from a gen 3). If anyone is looking for cheap inverters the Prius gen 3 inverter is absolutely beans cheap for what it is. It was $51 for the inverter and $42 for the steering column + control module that goes with it. The column uses the same spline as the Subaru rack I got off my donor, Perfect! I just cut up the console spaceframe up to get the epas motor to fit. I plan on adding CAN control on the column but I need to use my logic analyzer to look into how the control module communicates with the rest of the car. I've been asking all of my friends but none of them have a Prius with the same year as the donor I got it out of. Just another tangent to go down....

I got the Prius inverter working with an open inverter custom logic board to generate PWM for the IGBT drivers. This inverter will have both of its motor outputs parallel for more current handling since I'm not going to use the synergy hybrid drive. This stack will go in the first car since the leaf inverter is far too big to fit.

IMG_2539.jpg
IMG_2772.jpg

[Ajzride]

You probably don’t need to control the EPAS unit, limp mode should be about right for an 818. We use them in classic mustangs all the time in limp mode and it’s perfect. A 65 mustang is about the same weight and track width as an 818.

[J R Jones]

You probably don’t need to control the EPAS unit, limp mode should be about right for an 818. We use them in classic mustangs all the time in limp mode and it’s perfect. A 65 mustang is about the same weight and track width as an 818.

A, My 1965 SCCA GT350 Shelby B Production spec weight was 2700lb + or - a tolerance. Based on Kim3's battery weight comments I predict his 818 will weigh more.
I had my first drive by wire steering experience in a 2010 Acura TL SH AWD. It was wonderful until I pushed it to understeer. Setting up for a high speed left hander onto a freeway entrance ramp the steering felt routinely great, but the front tires did not, they were sliding on the wet pavement. If you prefer genuine feedback, drive by wire does not provide it.
jim

[kimchikimkim]

UPDATE

Got the Tesla gen 2 OBC (onboard charger) mounted in the vehicle! I probably spent 4 hours drawing up ideas for where to mount the charger but I'm pretty happy with what I ended up with. A big part of its location is the fact that the gen 2 charger is not weatherproofed (as it sits under the rear seats in a model s. I don't have room under seats and I didn't have enough money to get a waterproof enclosure so it basically had to be mounted in the interior of the car.

The Igus cable came in the mail today. Super high quality and the RAAL 2004 looks super nice brand new! Both SKUs of cable I got are fully shielded.

Starting to fab up the emergency HV disconnect mount. I'm making this super overbuilt as the last thing I want is dangling HV.

IMG_2785.jpg
IMG_2788.jpg
IMG_2789.jpg
IMG_2787.jpg

[kimchikimkim]

You probably don’t need to control the EPAS unit, limp mode should be about right for an 818. We use them in classic mustangs all the time in limp mode and it’s perfect. A 65 mustang is about the same weight and track width as an 818.

I'm pretty sure I don't have one with failsafe when CAN isn't present. Also, for my application I want to be able to adjust the column assist strength.

[kimchikimkim]

UPDATE

I've been working on the car steadily since the last post.

I had to completely redesign the BMS system due to a major oversight. This has been a HUGE roadblock in the project so far as I have had to learn a lot to just get the boards ordered. I've had to design my own BMS based on the ZEVA EVMS V3 system. I finalized the Gerbers and sent them off to get partially populated at JLCPCB. I have so much soldering ahead of me it's no joke lol. I am by no means an electrical engineer so it took a while to understand how to use EAGLE EDA and modify the boards for my purpose.

Because I will have a total of 288 cells to monitor, I had to change the microcontrollers to ones with larger flash capacities and redo the embedded code architecture to read all of these cells on the bus. The way the BMS works is it uses LTC6802-2 ICs to do the cell management, then it uses an ATMEGA32M1 in my case to take the isolated SPI and throw it on the CAN Bus via a Texas instruments CAN transceiver! Pretty simple from an electronics standpoint but the code behind it is far from simple.

I decided to move to CAD for a lot of the high voltage subsystems since I'm routing cabling. I'm using EMI mitigation guidelines to keep logic-level signals from picking up on interference from the traction battery cables which is a huge headache.

I'm also thinking about getting another 3d printer since my Prusa is running 24/7 and I need to print a boatload more parts.


too many projects too little time.

21331213231312.PNG3213231231312.PNG

[kimchikimkim]

UPDATE

It's been a while since I made an update since I had a lot of reading and learning to do. I am working on making a custom BMS solution since I am working with a total of 288 cells. A few weeks ago I started by taking the reference designs of the ZEVA BMS system and tweaking the boards to work with components that are still in production and work with my use case. I also rewrote firmware to increase the amount of BMS modules supported by the system. Just to be safe, I increased the flash size of all microcontrollers using pin-compatible ATmega64m1. I'm not a programmer by any means so it took weeks to understand both how the IDE works and how to use the chips. Since some of the parts were not stocked by the PCB/PnP manufacturer, I had to hand solder a ton of microcontrollers including the LTC6802-2 BMS IC. Finally have some good news to share after all that pain-the all 12 BMS slave units are spitting out respective CAN frames! I will be working more on software integration down the road as I need to make a thermal management circuit board to control the battery heater over LIN as well as integrating some sort of Economizer for my flow indexer. Lots more to come, but I am slowing down a lot since this is very new stuff for me. I've barely touched the car since the last update since I have been on my computer so much but I did a bit of work today on the battery thermal management system (almost done plumbing up the water loops).

IMG_3296.jpgIMG_3292.jpgIMG_3300.jpg

[kimchikimkim]

UPDATE

Working on wiring and completely redesigned the High voltage junction box.

I built up the custom driveshafts that adapt the Telsa CV cups to STI spindles I got the cv shafts made by dutchman axles and they cost $630 shipped for the pair (they fit perfectly!)

IMG_3460 (1).jpg

Here are some pictures of the work I have done.

IMG_3437.jpg

IMG_3466.jpg

I'm trying to finish the car within the next few months but as I'm learning on the job I'm not sure if that's feasible.

[Bob_n_Cincy]

kimchikimkim
Don't hesitate to ask any questions. In 2010 I did a power distribution box for a port container mover. I also designed BMS hardware and software for using LTC6802 chip.

pdb.jpg bms.jpg

[kimchikimkim]

kimchikimkim
Don't hesitate to ask any questions. In 2010 I did a power distribution box for a port container mover. I also designed BMS hardware and software for using LTC6802 chip.

pdb.jpg bms.jpg

That BMS looks awesome. The discrete bleed resistor network on your custom BMS is insanely beefy where your parallel groups really unbalanced? How did you deal with isolation fault detection on your build too?

[Bob_n_Cincy]

That BMS looks awesome. The discrete bleed resistor network on your custom BMS is insanely beefy where your parallel groups really unbalanced? How did you deal with isolation fault detection on your build too?

At the time we were running 96 700AH thundersky (winston) batteries. It needed a lot of balancing power.

I design an isolation fault detection circuit board that would measure the voltage Pack+ and chassis ground and also measure between Pack- and chassis ground. I had a resistor array on the board to adjust the sensitivity of the circuit.
Dealing with isolation faults was a big problem with our trucks. In southern California, it was not a problem. But running our trucks in Michigan winters there was salt spray that would get into the battery packs. The salt would dry on the cells and leave invisible conductive tracks.

[kimchikimkim]

UPDATE

Working on the cooling system right now. I had to buy new seats since the NRG ones I got were too wide to center the steering wheel properly. Since I am using a Prius EPAS column, I needed to redo the mounting system for the WRX column. I'm pretty happy how it ended up but since I am using the stock AT Subaru pedal box (that is really wide compared to the manual one, I needed to hack it up to make the u-joint fit. Even with this mod I still can't center the wheel perfectly in line with the driver seat space so I ended up getting some Kirkey Seats. Although they were expensive they were much thinner in x y and z so I'm definitely going to enjoy the extra legroom as well as the ability to offset them inwards to be centered with the steering wheel.

After the cooling system is tested in a vacuum I will proceed to fill both the power electronics loop (with G48) and the battery loop (deionized dexcool). I'm hoping to be able to fire up the motor then and get my VIN! A bunch of features that I want on the car are being skipped right now in the interest of time such as the battery conditioner logic as well as a majority of the CAN bus system including the BMS integration with an ECU that I will have to design down the road.

IMG_3544.jpgIMG_3558.jpgIMG_3569.jpgIMG_3570.jpg

[Dave 53]

That BMS looks awesome. The discrete bleed resistor network on your custom BMS is insanely beefy where your parallel groups really unbalanced? How did you deal with isolation fault detection on your build too?

If the parallel groups are unbalanced, try a series counter phase rheostat coupler. Isolation fault detection is best delt with by installing a class 3 micro incapacitator linear inspection gauge.

[Bob_n_Cincy]

If the parallel groups are unbalanced, try a series counter phase rheostat coupler. Isolation fault detection is best delt with by installing a class 3 micro incapacitator linear inspection gauge.

Hey Dave, you have been watching too many Turbo Encabulator videos.

[kimchikimkim]

UPDATE

The motor is finally spinning. I cleaned up the harness I made and plasma cut out a temporary key switch to gokart around.IMG_3664.jpg

[kimchikimkim]

UPDATE

I got a VIN! Luckily the CHP didn't care much for body panels and just riveted on the VIN no problem.

I've been really quite since I'm diving really deep into embedded systems coding. I decided to redo the BMS system and instead leveraging chevy volt gen 1 BMS modules (since the CAN network is completely reversed engineered). I bought most of the LTC6802-2 ICs on eBay and lets just say that some of them have some pretty wack mem register values when reading cell voltages. I have 3 months before college starts so realistically I don't think I can finish with how much more I have to learn in the coding department. I'm now using a TMS570 MCU with 4 CAN peripherals. Its a really cool chip that has dedicated ECC memory for each CAN message box as well as super robust watchdog/hardware level clock protection. CCS and HALcogen is still super confusing to me so don't expect any sort of serious leap forwards anytime soon haha.

IMG_3841.jpgIMG_3839.jpg

[FFRWRX]

You're going to need one of these for your car:

Screen Shot 2022-06-10 at 5.10.49 PM.jpg

https://www.etsy.com/ca/listing/1235...-2&frs=1&sts=1

[fletch]

UPDATE

I got a VIN!

OK. I'm jealous.
But seriously, congratulations. That's an amazing accomplishment, especially since you're building an electric 818. Well done! And good luck at college. I'm sure you will do just fine!

[kimchikimkim]

Update

Well I couldn't say no to a deal on a Chevy BOLT 2020 battery pack and basically I'm starting from scratch. I scored a full chevy bolt battery pack for $3.5k with only 6K miles! This car was literally brand new. I pulled out all of the battery modules in the car now and I sold some of them on craigslist. I was actually able to offset the cost of the new battery and pocket $500 in profits on top of that by selling only 2/3 of the cars original pack! I tore the battery down to module level and am designing a new thermal management system. This battery has a volumetric energy density and a gravimetric energy weight density that is 50% better than the volt pack so it will save around 600lbs off the entire car while giving me 10kwh extra of range (I'm looking to closer to 280miles now abt 66kwh in total!) They are a pain in the *** to package though since they require me designing a custom cold plate and a bunch of other custom bracketry/boxes to house them. I've spent the past few days modeling everything in Solidworks I really want to do this right so I'm doing flow sim on the coldplate to make sure that I am able to dissipate more than enough heat given a typical delta T of coldplate/cell. This is a ton of work but I as nearly everything I'm doing is teaching me more and more about engineering, I am more than willing to go the long route and end up saving money in the end.

IMG_3918.jpg

[kimchikimkim]

UPDATE

The car is mostly reworked with the new battery setup. Since I knew exactly what was going into the car and I recently bought a cnc plasma cutting table I headed straight to CAD for everything. I'm pretty happy how it turned out! I just took it around the block today for a drive and it was absurd how much power this thing had. The whole time I was barely touching the throttle

68264161650__CC9473D5-BA70-499C-A4E4-2E87CEF9C19B.jpg

[kimchikimkim]

Its been a while since I updated.

I have decided to make my own ECU for the car which has taken me a while to design and layout but here is a board shot of it in EDA.

I sent off the Gerbers yesterday and will be developing firmware for it as soon as it comes in. It will be inc charge of every can bus network on the car and I plan on using it to even control DCFC through the implementation of CHAdeMO and hopefully, the CCS LIM now that the CAN bus messages have been decoded (thank you Damien Macguire)

some quick specs of it:

µC: NXP ARM CORTEX-M7 MIMXRT1062 @ 600mhz Core

Input Vdc: 5v-40v

5v 3a Switch Mode PSU

2x CAN 2.0B+ 1x CANFD Interfaces

8x isolated 12v Digital Inputs

8x Smart High-side driver outputs

1x Low-side driver

2x TE Connectivity 23 Pin AMP SEAL connectors

2x 5v PWM

2x 12v PWM

Screenshot_2022-09-23_122022.png

[kimchikimkim]

The PCBs arrived and I assembled 2 up. Found that I didn't set up some trace DRCs correctly and one of my nets was linked incorrectly but that's life. A botch wire and some soldering later confirms that the switch mode power supply I designed works flawlessly!

I have a lot of coding now ahead of me!

IMG_4769.jpg