Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[KarenRei]

As I understand it you don't want the supply voltage too much higher than the pack voltage, you want to be able to smoothly ramp the voltage as the pack charges, so I'm not sure if simply switching SC modules in parallel with the pack to charge it is really an ideal solution

1) One can add an arbitrary amount of resistance to the circuit to choose an arbitrary voltage
2) One can deliberately offbalance bricks so that they're not uniform increments of ~3-4,2V; given the number of bricks, you can create pretty much any arbitrary voltage
3) A couple volt overvoltage is not that significant.

Plus, as the SC packs are run down their voltage will drop

Hence the entire point of switching - to change the number of bricks that are in series, and which ones.

However, rather than each pair of stalls and their corresponding cabinet being separate systems from the next pair of stalls, etc, it would probably make a lot of sense ... a centralized and large battery bank

Agreed on this.

 

[jhm]

I waded through some TSLAQ on twitter to get any nuggets of what their attack plan will be after ER.

- A fair amount of Fraud implications - whatever
- A few pointing to Accounts Receivable not going down after the lengthy explanation in Q3 of why it was higher in Q3 due to ending quarter on a Weekend.
- A few pointing to 2019 Capex spending that was lower than forecast 2019 Capex (and instead of pointing to efficiencies, stating that they couldn’t afford the Capex they wanted because they don’t have the cash and had to manage Capex down to match only what was available)
- Another couple citing Q4 interest income based on yada, yada benchmark shows their actual cash position to be shockingly low

One of my favorite parts of this forum is having the knowledgeable people here dissect and obliterate short arguments so please have at it with the above bullsugar.

Now I need a shower.

Thank you for your service. Now clean up.

 

[KarenRei]

I use my SEP IRA as well. It's great to trade with no thought of tax.

I trade with Real-IRA. No taxes, but the penalty for late deliveries is pretty severe.

Aww, nobody laughed at my joke. Better cancel that book of Irish paramilitary puns I was working on...

 

[Vitold]

What do you mean? Took 2 years to get the Model 3 out after revealing...and probably 4 years for Semi/Roadster.

Tesla loves to generate buzz. Announcing the car doesn't mean production will start for any car company. I mean how many concept EVs have we seen from all the legacy automakers that are still a no show? It's kind of how the industry works.

In fact, the reason why people even talk about "future competition" is because all these legacy car makers showed off their concepts. Tesla can't just sit on the sideline and show nothing.

That worked before because Model 3 was substantially different than S so there was little to worry about ‘Osborne effect. However I think that many potential Model 3 owners in US and to lesser extent in Europe, would postpone their purchase until Y arrives. That is why I cannot understand why Tesla is showing Y in March if they do not plan on producing it for another year.

 

[HG Wells]

Really makes you wonder why he still has a job. What other job can you fail so miserably and consistently at without being fired?

PRESIDENT ?

 

[ggies07]

PRESIDENT ?

****in A

 

[tinm]

Quick New Mexico update:

The bill passed 4-3 in favor!

Onwards to the next committee!

we still have a long ways to go but today, glory!

 

[Bet TSLA]

Thought re: supercharger expansion. We've noticed that it's been slow for the past couple quarters

Supercharger expansion has been proceeding at a record pace the past two quarters. So what are you talking about? I guess it could be considered low on a percentage basis since the base number has gotten so big, but supercharge.info shows >80 in the permit process and >40 under construction worldwide. Numbers this high have been rare in the past.

 

[Carl Raymond]

Tesla is a slot machine. It could move +/- $20 either way tomorrow, depending on what news comes out.

Correction. It could move +/- $20 either way tomorrow, regardless of what news comes out.

 

[Carl Raymond]

A couple of ways I believe:

• Tesla won't be making some key components, such as the drive train, in China.
• They'll be keeping their cell making R&D results in Nevada as well.
• As China's economy matures, so might its legal system.
• But most importantly, they'll compete the "Tesla way": not by building a patent portfolio, but by being a fast moving target.

Additionally, much that is secret is in the form of source code. Source code will never leave the development environment. Only executable code (compiled and incomprehensible, except to a computer) will go near GF3.
It’s the same reason that having MS Office on your desktop does not make you a threat to Microsoft.

 

[Singuy]

That worked before because Model 3 was substantially different than S so there was little to worry about ‘Osborne effect. However I think that many potential Model 3 owners in US and to lesser extent in Europe, would postpone their purchase until Y arrives. That is why I cannot understand why Tesla is showing Y in March if they do not plan on producing it for another year.

It's a car, not a phone. For people with leases up, they would like to buy what is available and not pre-order something that may come in a year or two. People only did that for the Model 3 because there's no other alternative(since the S is almost 2x the price). Will there be massive amount of reservations? Absolutely(suspected to be mostly from Model 3 owners who want a second car). There are also die hard SUV or bust people out there who will order. But the majority of the people wants what is available and sensible to their needs. A compact SUV really doesn't give you all that much...utility over a conventional car. It's totally marketing.

 

[RobStark]

I don't see why they couldn't source their metal for GF3 from China. It seems unlikely that there is zero capacity for high end steel / aluminum in China. After all a lot of premium products are built there. Though it might require a capacity ramp or something like that at a foundry. They may be importing their steel / aluminum this year, but I imagine within the next 1-2 years after GF3 start of production they'll have local suppliers for it.

US uses a lot of premium steel. That it imports from Europe.

Just because you use it doesn't mean you manufacture it.

 

[mongo]

As I understand it you don't want the supply voltage too much higher than the pack voltage, you want to be able to smoothly ramp the voltage as the pack charges, so I'm not sure if simply switching SC modules in parallel with the pack to charge it is really an ideal solution. Plus, as the SC packs are run down their voltage will drop, and so you'll probably need DC-to-DC conversion in heavy usage scenarios in order to provide the necessary voltage anyways.

However, rather than each pair of stalls and their corresponding cabinet being separate systems from the next pair of stalls, etc, it would probably make a lot of sense to have each stall go to a DC to DC converter which is then connected to a centralized and large battery bank (which in turn would connect to a large efficient AC to DC interface, and perhaps DC to DC from solar power). The bigger DC to DC conversion should be easier/cheaper/more efficient than lots of small AC to DC, and while it might have been a cost saver initially to adapt S/X chargers for the purpose, purpose built monolithic (or at least, modular but with larger modules) power systems are probably more cost effective now at today's scale.

Each of the 16 modules in a powerpack has its own DC-DC converter. Assuming the voltage bus to the inverter is in the 400V range, this would allow a powerpack to charge a vehicle directly. Each inverter plus 4 power packs could charge one car at 200kW from grid, pack, or a combination. With some extra contactors, 4 PP and an inverter could do 2 cars at 100kW.

1) One can add an arbitrary amount of resistance to the circuit to choose an arbitrary voltage

That is a huge lossy heat source. No reason to go linear when DC-DC is already there (or can be added). Just 0.01 ohm of resistance to adjust the voltage 3V at 300Amps would waste almost a kilowatt, now scale that to a 400V pack charging a 350V pack, 50 Volt drop * 300A = 15kW to dissipate.

2) One can deliberately offbalance bricks so that they're not uniform increments of ~3-4,2V; given the number of bricks,

That takes extra balancing HW, plus contractors to stack the proper bricks.

3) A couple volt overvoltage is not that significant.

Yes it is when you are talking a lithium pack, equivalent series resistance is very low (look at drop vs current for a 1,500A launch, or cell arrangement: 96s86p) , it's in the tens of milli-ohms. Even at 0.1 ohm (which would be 150V drop on launch ignoring cable loss) 1 volt would be a 10A current change or 3.5kW rate difference on a 350V pack.

 

[MartinAustin]

That worked before because Model 3 was substantially different than S so there was little to worry about ‘Osborne effect. However I think that many potential Model 3 owners in US and to lesser extent in Europe, would postpone their purchase until Y arrives. That is why I cannot understand why Tesla is showing Y in March if they do not plan on producing it for another year.

The plan for Y is to reveal it in Q2 (approximately) or maybe end of March 2019. Then low volume production in early 2020 rising to volume production by end of 2020.

That is a pretty aggressive schedule IMO!!! Not only that, they're going to be producing in scale in two continents in 2020!!! I don't think they are moving slowly.

Model 3 was unveiled March 31st 2016 and deliveries began before August 2017. A gap of about 16 months.

 

[Yodi]

Bloomberg shows 33k M3 production for January. Interesting, any thoughts?

 

[HG Wells]

Bloomberg shows 33k M3 production for January. Interesting, any thoughts?

I dont think thats right. It's only 7,616 cars per week.

 

[Hock1]

I’ve seen this before, but don’t follow. Why would it matter how many shares they’ve lent to shorts? Said shorts will lose money as the stock rises, but that has zero effect on GS.

It's complicated. Suffice it to say GS lends as many shares as they possibly can because they get paid to; they certainly cannot have an analyst who recommends that their clients buy a company (Tesla) whose shares they have lent out, (perhaps in excess of the amount that are actually available to lend)..... Like I said, it's complicated.

 

[MP3Mike]

The plan for Y is to reveal it in Q2 (approximately) or maybe end of March 2019. Then low volume production in early 2020 rising to volume production by end of 2020.

That is a pretty aggressive schedule IMO!!! Not only that, they're going to be producing in scale in two continents in 2020!!! I don't think they are moving slowly.

And actually they will be building two new factories in parallel. The GF3, for Model 3, in China, as well as a new building at GF1, for Model Y. Both are supposed to be producing cars around the end of this year.

I'm curious as to what will be done at GF1 for Model Y... My thoughts, assuming there is capacity, is that they will do the stamping in Fremont and use the return trips of the Semis that brought Model 3 battery packs/motors to Fremont to take the stamped parts to GF1. (Big advantage there is that you make both directions of the GF1<->Fremont trip productive. (I'm assuming that they aren't already hauling raw materials to GF1 which could be wrong.) And it takes the large lead time for stamping machines off the table.

We know that they can build a GA line in about a month, but I would think body and painting would take longer to build.

 

[MP3Mike]

Bloomberg shows 33k M3 production for January. Interesting, any thoughts?

I dont think thats right. It's only 7,616 cars per week.

Of course it isn't correct, that is 7,700 per week, and Tesla has said that they are only planning to take Fremont to 7,000 per week.

Also, last our source at GF1 has reported is that they are only producing ~6,000 Model 3 packs per week. So unless there was a lot of downtime in December for them to build up a stock of packs that can't be producing at that level yet.

 

[Wnuk]

Really makes you wonder why he still has a job. What other job can you fail so miserably and consistently at without being fired?

Meteorologists?