best way to build a 2170 or 4680 pack 4 a first gen Roadster?

[MitchJi]

Hi.

Trying to get a good deal on a Tesla Roadster with a solid pack. I hope to find a good deal on a Roadster witH a bad battery or PEM. The OEM cells are older than their design life. One possibility is to buy Tesla's 3.0 upgrade for 29k plus 5k to get a rebuilt PEM. In addition to the high cost, I think that building a pack from either structural slabs of 4680's or pieces of 2170 modules could be done for much less than 30k. I think that doing that pretty much means dumping he PEM, which is a pain but not only does doing that make it possible to us the most cost effective cells available to today but it will make it much simpler,to replace either individual cells or the entire pack in the future.

Main reasons for choosing the 4680 structural packs is that I believe that dividing the pack into self-contained chunks will be relatively easy and on 6-18 month's I think they will be the least expensive.

Taking apart a structural battery pack will be inherently challenging. The whole point is that there is structural adhesive holding all the cells together so that they act as a solid structure. Take a look at the Model 3 modules and how hard they are to take apart without damage. Now imagine that there are no modules and the whole pack is bonded together.

Thanks! I think a sawzall or a bandsaw might work. Depends on if the cells are touching. If not switch to 2170's.


I can't do this project myself, (health issues). I'm trying to come up with a plan that EV-West or some close friends can implement for about 10k-30k that will be better than the 3.0 package.

NOTE:
Posted by @X.l.r.8 on another forum:

background so I understand the desire to build concept. Head over to open inverter. That’s where we are doing exactly this. The car itself makes little difference as you will see. Using a model Y battery may not be the ideal once you read through the battery threads. The simp bms controls pretty much all BMB’s so you can use what you like. The Prius/leaf and Tesla model s and 3 have been pretty much decoded and have can codes for pretty much all the components, even DC charging. Not sure why you would subject this to a roadster, but then I had a V8 in a 924 CGT so who am I to judge.

I don’t know what the saturation point is for a roadster motor is, but the fact they are hand wound for the sport and the temperature it climbs to on 200 mile drives, showed it’s pretty close to max. If you could push the max voltage rather than increasing the amps, then find a way to get some better cooling. Again, at that point I think a M3P motor would be a much better option. Built in inverter, throw a Gen 2 48A charger, a volt 12v DC converter and you could have quite a package. Again stopping it may be problematic. The weak point of the roadster is the brakes, it was a primary reason I went to the Porsche platform for my model S performance motor.

To your point the 3.0 gives more range, that’s it, the 3.0 does not charge quicker so the range is hindered by the increased charging time, something model 3 owners fail to understand. If you are charging for 6 hours for a 2.0, the 3.0 will take 9 hours. So unless you can reach your destination in one go, the 3.0 is basically bragging rights. The modern pack will need cooling? And that’s controlled via the PEM. There goes the HVAC, I’m not seeing any improvement, it’s no longer an integrated package. You may improve range, but that’s one metric, knowing the roadster is one of the most efficient vehicles, period, you are going to have to do something amazing to ‘improve’ on the original in all areas, making it go faster is easy.

Would the MP3 motor be easier to control than the OEM motor? How difficult to mount? The most fun would obviously be either one new MS carbon wrapped motor that weigh 30 lbs. and can put over 400hp or the complete dual motor rear drive system with torque vectoring.

[P.S.Mangelsdorf]

Main reasons for choosing the 4680 structural packs is that I believe that dividing the pack into self-contained chunks will be relatively easy and on 6-18 month's I think they will be the least expensive.

As of right now, the 4680 cells do not exist outside of Tesla's laboratories. Based on Tesla and Musk's recent comments (last day or so), they seem to be worried about whether they can make them a reality.

I can't do this project myself, (health issues). I'm trying to come up with a plan that EV-West or some close friends can implement for about 10k-30k that will be better than the 3.0 package.

NOTE:
Posted by @X.l.r.8 on another forum:

Head over to open inverter.

EV West is not going to use Open Inverter control. They have several control systems they use including some they have codeveloped. That does come with a cost premium.

If you want to build this with some friends, Open Inverter is the right place to be. If you want a shop to do it, get in contact with that shop and start discussing options with them. From what I understand, EV West is pretty good at guiding clients for in house builds.

NOTE:
Posted by @X.l.r.8 on another forum:

background so I understand the desire to build concept. Head over to open inverter. That’s where we are doing exactly this. The car itself makes little difference as you will see. Using a model Y battery may not be the ideal once you read through the battery threads. The simp bms controls pretty much all BMB’s so you can use what you like. The Prius/leaf and Tesla model s and 3 have been pretty much decoded and have can codes for pretty much all the components, even DC charging. Not sure why you would subject this to a roadster, but then I had a V8 in a 924 CGT so who am I to judge.

I don’t know what the saturation point is for a roadster motor is, but the fact they are hand wound for the sport and the temperature it climbs to on 200 mile drives, showed it’s pretty close to max. If you could push the max voltage rather than increasing the amps, then find a way to get some better cooling. Again, at that point I think a M3P motor would be a much better option. Built in inverter, throw a Gen 2 48A charger, a volt 12v DC converter and you could have quite a package. Again stopping it may be problematic. The weak point of the roadster is the brakes, it was a primary reason I went to the Porsche platform for my model S performance motor.

To your point the 3.0 gives more range, that’s it, the 3.0 does not charge quicker so the range is hindered by the increased charging time, something model 3 owners fail to understand. If you are charging for 6 hours for a 2.0, the 3.0 will take 9 hours. So unless you can reach your destination in one go, the 3.0 is basically bragging rights. The modern pack will need cooling? And that’s controlled via the PEM. There goes the HVAC, I’m not seeing any improvement, it’s no longer an integrated package. You may improve range, but that’s one metric, knowing the roadster is one of the most efficient vehicles, period, you are going to have to do something amazing to ‘improve’ on the original in all areas, making it go faster is easy.

I'm not very familiar with what is in the roadster, but this post makes some good points. You need to think about what range you want, how you want to charge it, and how much of the original car are you willing to take out to facilitate upgrades.

If you want a range upgrade, Tesla batteries are probably a good bet, but not cheap. The S & X are made in several bricks (16 I think?) that are easy to reconfigure. The 3 & Y are in 4 long modules that really need to fit under the car and from what I understand are a pain to reconfigure.

If you want to upgrade charging, Chademo is available open source, and Damien is in the alpha stage of testing CCS. These options are not compatible with Tesla's charging stations (at least not in the US). I know Tesla said they will open to other manufacturers next year but a) thats most likely in Europe where they already use CCS for their stations b) even if it wasn't, based on how they treat rebuilt Teslas in the states, I think we can say with a high degree of confidence they would lose their mind over a conversion/modified/upgraded car using their chargers.

Motor-wise, the Model 3 open source solution is in process. The Model S & X open source solution is functional and successful. The new Plaid motors appear to use Model 3 inverters, so hopefully those will be easy to crack however, I expect the prices for one out of a wreck, at least for the next 5 years, to be utterly insane.

for about 10k-30k that will be better than the 3.0 package.

Something that I have learned the hard way over the last 2 years: Your budget needs to include all of the small stuff that seems trivial, but adds up. $100 here and there for grade 8 nuts and bolts adds up fast. Cables at $5+ a foot adds up fast. Metal for battery boxes, motors mounts, etc adds up fast. Hell, welding wire and gas adds up fast.

That's not to say it can't be done on that budget. It totally can, but you should go into it knowing that it will require shopping around, having patience, planning well, and taking every effort to not screw things up or break them.

[MitchJi]

thanks!

RE the 4680's Tesla believes that they will probably be producing enough cells in austin by the end of 2021 to launch the MY in texas and germany. they said on the ER on July 26 that they have a contingency plan to use 2170's if they miss that goal.

I'm highly confident that by July 2022 that they will be producing enough MY's with 4680's that used packs be available in junk yards.