Investor Engineering Discussions

[GSP]

The mfr date shown is 12/20, which is a prototype from last year.

But considering where the battery is, wouldn't the weight be split evenly between the front axle and int axle - since the trailer weight would only go directly onto the rear and int axles?

Edit: So just a long winded to guess that the semi weighs 25,200 lbs?

Note that GWAR is Gross Axle Weight Rating. The rating is the max weight the axle can support, so it includes empty weight plus cargo and trailer. I don’t think knowing the GWAR provides any insight into what the empty weight is.

GSP

 

[Discoducky]

Been watching some Munro Live vid's on the R1S/T suspension (they've gone quite in-depth) on YT and I've got a suspicion Sandy is most likely contracting with Tesla on CTs suspension.

Who better than Sandy? He talks about his lifelong Jeep 4x4 experience, he loves his 4 motor R1T and is hopelessly in love with giga casting. Here's to hoping!

 

[mongo]

this. I’m looking into FIY Solar and storage and you can get 5kWH server rack battery systems from China with shipping included for under $2000 vs the Powerwall is $12,000 for 11.5kWH. From reputable companies many in DIY community go to, and you can even pay for third party inspections to verify stats prior to shipping too.

China competition for tesla in energy is huge. However, I do think tesla is transitioning to an lfp Powerwall. But yah, the software advantages are mitigated because it’s not as emotional as a car purchase.

So much talk about lanthrop. It’s great and mega packs will do well, but it’s ramping and it won’t have the margins or financial impact we want for near term.

I still think tesla energy is huge because much of the world will be as protectionist against China as possible (hence Enphase success and potential), but it’s not a silver bullet.

PW is now 13.5 kWh so around $5k in rack batteries?
How much additional for full system with 7kW continuous 10kW peak utility approved grid tie sine wave inverter/ charger that supports islanding with 200A automatic transfer switch?

 

[MC3OZ]

I don't want to clutter up the main thread with speculation about the 9000T press and Model 3 Highland, but I'm happy to clutter up this thread.

I would investigate casting a frame for the roof, and the Model 3 roof seems big enough to require a 9000T press.

Ideally the cast frame would work for glass, (LR AWD) Performance and Metal (SR).

This would be in addition to using Model Y front and rare castings and a structural battery pack.

The sides probably remain as stamped as very little advantage in casting them,

The cast roof might need some steel reinforcement, but that would be internal where fit/finish are less of an issue.

For the metal roof version stamped panels would be fitted to the frame before painting and some additional reinforcement might be needed.

Access to the inside of the vehicle is still straight forward before the seats on the structural battery pack are installed.

What a metal roof would hopefully achieve is, lower cost and lower weight, both important for the SR.

The roof casting would also hopefully save a lot of body shop work and make supporting a dual metal/glass roof option easier.

I'm also sure there are arguments against this suggestion, fell free to make them, speculating here is part of the reason this thread exists.

 

[mongo]

I don't want to clutter up the main thread with speculation about the 9000T press and Model 3 Highland, but I'm happy to clutter up this thread.

I would investigate casting a frame for the roof, and the Model 3 roof seems big enough to requires a 9000T press.

Ideally the cast from would work for glass, (LR AWD) Performance and Metal (SR).

This would be in addition to using Model Y front and rare castings and a structural battery pack.

The sides probably remain as stamped as very little advantage in casting them,

The cast roof might need some steel reinforcement, but that would be internal where fit/finish are less of an issue.

For the metal roof version stamped panels would be fitted to the frame before painting and some additional reinforcement might be needed.

Access to the inside of the vehicle is still straight forward before the seats on the structural battery pack are installed.

What a metal roof would hopefully achieve is, lower cost and lower weight, both important for the SR.

The roof casting would also hopefully save a lot of body shop work and make supporting a dual metal/glass roof option easier.

I'm also sure there are arguments against the suggestion, fell free to make them, speculating here is part of the reason this thread exists.

A roof frame is an interesting case because it is dimensionally large, but not large in terms of material crossectional area. So it would need a large press, but not necessarily a high tonnage one. I'm guessing it would need a bunch of post cast trimming due to flow path runners (which boosts tonnage) because of the narrow final form .

 

[Oil4AsphaultOnly]

Since all cast parts have been aluminium alloys, I doubt it would be for any parts involved in the passenger cell - which have been made of stamped boron-steel for strength.

My guess is that this 9000t press is for a cybertruck variant for the Asian/European/Australian market. What better product to dethrone the Toyota Hilux than a BEV pickup truck?

 

[MC3OZ]

My guess is that this 9000t press is for a cybertruck variant for the Asian/European/Australian market. What better product to dethrone the Toyota Hilux than a BEV pickup truck?

That is the other possibility, but it doesn't explain the timing.

Cybertruck production would need at least 20-30% of the Austin factory floor space, 4680 cell production and a bunch of tother stuff, It is very likely to be a major factory construction project, which could not fly under the radar.

A single (admittedly large) press for Model 3 Highland production needs less floor space, may save, some space/robots in the body shop area, and overall needs less floor space, fewer new processes and fewer changes to the existing build process. That also fits the timing neatly.

The press not being for Tesla is another strong possibility.

 

[willow_hiller]

Tesla Solar Inverter Notes
I did some more research today.

Tesla launched this product two years ago in Jan 2021, as indicated by an Electrek article I found and a few other sources. Tesla's warranty document revision 1.0 is dated 7 Jan 2021 too. It looks like tesla.com has had this for a long time Tesla Solar Inverter | Tesla.

So, why are we suddenly hearing about this again? It appears that for some reason Tesla recently decided to publish this whitepaper. Why they're doing it now instead of two years ago, I have no idea. The whitepaper is dated just "2022" on the document itself, but its URL address is "https://...blah blah blah.../Tesla-Solar-Inverter-Architecture-White-Paper-NA-EN-12212022" indicating that it was published on 21 Dec 2022, which seems likely considering that this whitepaper hasn't been discovered by the online Tesla community until now.

Tesla claims their inverter has reduced the all-in LCOE for home solar by 3-8% relative to competing systems. That isn't a massive difference overall, but if much of that cost improvement feeds into gross margin then it has been making a substantial difference for the profitability of Tesla Solar. The gross margins were thin to begin with, so a 5 or 6% boost is big and might double the profitability. If so, this has clearly been a materially contributing factor in Tesla's aggressively low prices for solar projects in the US and their price-match guarantee.

I don't have enough expertise in this area to evaluate how Tesla's inverter stacks up against other string inverters or how much their expertise and scale in power electronics really translates to cost and performance improvements as Tesla claims. It's one thing for Tesla to say that switching design architecture from having electronics on the roof under each panel to string inverters simplifies and saves money these days, but it's another to say that Tesla's string inverter solution is better than other string inverters. For any engineers who want to try tackling such a comparison, here is the official technical spec sheet for the inverter and its integrated emergency shutdown device, dated 2 Feb 2022. We can take further discussion to the engineering thread.

Yep, Tesla unveiled and began installing their in-house inverter a couple of years ago. I believe this white paper was written this year because this is the first time that Tesla is planning on selling their inverter to certified third-party installers, instead of installing it directly.

I've seen lots of examples of Tesla Inverter installations on Reddit over the last two years, and from what I can gather the power electronics are rock solid. Most SolarEdge or Delta inverters tend to have a steep bathtub curve of failure rates (on my solar panel system with a SolarEdge inverter, the first one died within the first two months, and now the second has lasted a few years). The only common failure point I've seen on the Tesla Inverter is the cooling system, where production gets throttled due to overheating.

As with many things that Tesla vertically integrates, the advantages are primarily in manufacturing costs, ease of assembly, and ease of maintenance. With MPPTs instead of optimizers, all of the equipment on the roof is solid-state, so 99% of repairs will not involve climbing onto the roof, and all of the employee training / additional insurance that entails.

Oddly enough, I've seen a recent uptick in the number of SolarEdge inverters installed on Tesla systems in the last few months. But those may just be areas of the country that are presently serviced by Tesla Certified Installers, and this change in sales-strategy will enable them to roll out their in-house inverter in those regions too.

 

[Gigapress]

I believe this white paper was written this year because this is the first time that Tesla is planning on selling their inverter to certified third-party installers, instead of installing it directly.

That's what I was wondering but I wasn't sure. The website has the white paper link under the heading for third-party installers. What do you think is the significance of this change, and why would they wait until now to sell to other installers?

 

[MP3Mike]

why would they wait until now to sell to other installers?

Maybe they didn't have enough chips available to make more inverters than they were using themselves until now.

 

[willow_hiller]

What do you think is the significance of this change, and why would they wait until now to sell to other installers?

In the short-term it may be about providing consistency across the US for Tesla customers. Even if you request an installation from Tesla, if you live in a part of the country where they don't have first-party installers, they'll connect you with a certified third-party installer. But from what I've seen, the certified third-parties can be much more expensive and install inconsistent equipment that requires more warranty repairs. So giving certified third-party installers access to a more cost-effective and easier to install system helps solve that.

In the long-term, if they choose to expand sales beyond the certified third-parties, it might mean expanding their reach. At the moment, I think Tesla only installs solar panels in the US, and that's probably because of the time and expense required to be certified by the local authorities. But if they can transition from an installer to a manufacturer, it would be much easier to expand internationally.

 

[Buckminster]

Titanium: The Metal of the Future
Already being used by Tesla. Could they use more? Increase in cost per vehicle wouldn't be that much.

 

[mongo]

To be honest, if North America had chosen CCS2 instead of CCS1 things would probably be different. It still isn't a great design but it is better than CCS1.

Because three phase?

 

[MP3Mike]

Because three phase?

Better because of three phase support which really doesn't apply to CCS, but more importantly, like the NACS and MCS connectors, it doesn't have any moving parts on the connector to break. (Which is a common failure on the CCS Type 1 connectors.)

 

[MC3OZ]

I've started thinking about applying the high amount of regen braking in the Semi to track racing.

Specifically if a Tesla Roadster and/or Plaid+ Model S could have a peak regen rate of 300-500 kW. Not only does this save the brake pads, it captures a lot of energy on the track, extending range.

My impression was the Roadster was originally 2 x 100 kW 18650 packs - 200 kWh.

Getting something close to that might be possible by making 4680 cells a bit taller say 46 x 120....

If a Roadster (perhaps with front and rear castings) has (for example) a (structural) 46120 pack with 160kWh-180 kWh and say 350 kW of regen, it may be able to slay any ICE vehicle on a track including over longer durations.

A larger pack size is needed to maximise regen, the Plaid powertrain does the rest.

For track racing even Supercharger V4 probably doesn't charge fast enough to complete. The best bet is avoiding the need to charge. perhaps eventually followed by battery swapping.

The only place battery swaps might make sense is track racing.

IMO the high purchase price for the vehicles will partially offset the additional R&D costs, this also allows Model S and perhaps Model X to move to 4680 cells and for higher numbers to be built if that is desirable,.

Obviously this is a lower priority than Gen3 vehicles, but the Roadster will provide a strong clue..

 

[JRP3]

I've started thinking about applying the high amount of regen braking in the Semi to track racing.

What high amount of regen braking in the Semi? It's probably proportionally lower than the regen in the Performance Model 3 if his numbers are correct. PM3 100kW regen 80kWh pack, Semi 700kW regen 900kWh pack.

If a Roadster (perhaps with front and rear castings) has (for example) a (structural) 46120 pack with 160kWh-180 kWh and say 350 kW of regen, it may be able to slay any ICE vehicle on a track including over longer durations.

A larger pack size is needed to maximise regen, the Plaid powertrain does the rest.

For track racing even Supercharger V4 probably doesn't charge fast enough to complete. The best bet is avoiding the need to charge. perhaps eventually followed by battery swapping.

The only place battery swaps might make sense is track racing.

Structural pack means no swapping and Tesla isn't making a race car.

 

[MC3OZ]

What high amount of regen braking in the Semi? It's probably proportionally lower than the regen in the Performance Model 3 if his numbers are correct. PM3 100kW regen 80kWh pack, Semi 700kW regen 900kWh pack.

Structural pack means no swapping and Tesla isn't making a race car.

My takeaway was that the Plaid drive allowed higher regen, a consideration is where else would that come in handy.

The aim would be a proportionally high rate of regen. i.e. better than a PM3.

Cars already have "track mode", so some ability to track race is handy.

I agree Tesla isn't doing battery swapping, but that is a path for EV race teams to follow. e.g. Modify a Roadster so that the pack could be swapped more easily. No need to change any other part of the car, except perhaps reducing weight, improving brakes and aero.

 

[Oil4AsphaultOnly]

I've started thinking about applying the high amount of regen braking in the Semi to track racing.

Specifically if a Tesla Roadster and/or Plaid+ Model S could have a peak regen rate of 300-500 kW. Not only does this save the brake pads, it captures a lot of energy on the track, extending range.

My impression was the Roadster was originally 2 x 100 kW 18650 packs - 200 kWh.

Getting something close to that might be possible by making 4680 cells a bit taller say 46 x 120....

If a Roadster (perhaps with front and rear castings) has (for example) a (structural) 46120 pack with 160kWh-180 kWh and say 350 kW of regen, it may be able to slay any ICE vehicle on a track including over longer durations.

A larger pack size is needed to maximise regen, the Plaid powertrain does the rest.

For track racing even Supercharger V4 probably doesn't charge fast enough to complete. The best bet is avoiding the need to charge. perhaps eventually followed by battery swapping.

The only place battery swaps might make sense is track racing.

IMO the high purchase price for the vehicles will partially offset the additional R&D costs, this also allows Model S and perhaps Model X to move to 4680 cells and for higher numbers to be built if that is desirable,.

Obviously this is a lower priority than Gen3 vehicles, but the Roadster will provide a strong clue..

You might want to check out the regen capabilities of the Formula E Gen3 battery packs - 600kW!

Those packs aren't designed to last 1M miles, but I think exemplify what you were thinking about for track use.

Edit: So yes, eliminating the need to use actual friction brakes would be ideal in racing.

 

[JRP3]

My takeaway was that the Plaid drive allowed higher regen, a consideration is where else would that come in handy.

The limit for regen is likely still cell level charge acceptance rate not the drive motors.

 

[mongo]

Right, but those center modules seem to be close to double-wide (6 "ribs" on top, as opposed to 3), so I'im wondering if we are seeing the exterior enclosures the modules are in, and the center "can" holds 2 internal cell modules...?

Otherwise, it seems odd to have one module larger than the other two in each row.

The other thing is that it appears there's some metal framing in "front" of the forward-most visible module on the left side of the vehicle:

View attachment 900312

That looks angled "inwards" and looks like it would preclude a full row there...

@JRP3
The wider center modules are due to the main frame rails. While it may be possible that the center block contains two copies of the outer module's innards, that might be over optomization/ an unneeded constraint.
That cross framing looks like it is above the module level and below the rear stand off attachment points.