yes, said dryers and ranges often used 120VAC timers and controllers, and internally split the two hot legs.
Thank you for this portion of the conversation! I haven’t pulled the trigger on an EV yet and the discussion of the charging plugs has been very useful. I’ve had to replace the wheel bearings, springs/struts and LCAs on my trusty Subaru and since that pretty much ate up my discretionary cash I’m going to have to wait for another year at least before I can think about a used Model S or a new 3. By that time I may have a few more options to look at which is always a bonus! Meantime, I can map out where to put the plug in the garage!
Also, keep in mind that NEC 2017 625.40 mandates that each outlet for EV charging is installed on an individual branch circuit and has “no other outlets” , and I’m not sure how an interlocked circuit with two outlets is categorized.
On a separate note, motortrend has an interesting article on some of the manufacturing and engineering of the model 3, based upon a teardown from Sandy Munro (who appears to be well known / highly regarded in the auto industry for performing these sorts of analyses)
More information is also discussed in this rather lengthy youtube video from Autoline with Sandy Munro.
Some of the highlights:
- Poor mechanical fit and finish, especially on the earlier Model 3 Munro obtained. Panel gaps, tolerances and seals an issue. Lots of laughter regarding Elon’s desire to be able to calibrate a ruler to the panel gaps on the Model 3.
- A lot of issues chalked up to poor manufacturing training (i.e, sound insulation material installed backwards).
- Body is much heavier than it needs to be, doesn’t seem to be optimized well (I’m scratching my head at Munro’s claim about not believing FEA was used), and extra weight/material doesn’t seem to be needed for safety purposes.
- Everything on the skateboard design (suspension, battery, drivetrain) is excellent, and the electronics (battery, circuit card design) is described as being lightyears ahead of everyone else.
- Made a comment that if Tesla managed to just put in a competent/mediocre body design, outsourced manufacturing to someone like Magna, and slapped it on top of their current skateboard, Toyota would be “crapping their pants”.
- Munro doesn’t think the $35,000 Model 3 will see much light of day.
(mech. E here) I interpret it to mean- especially when he talks about “dinky panels all welded together”- is that they seem to have added parts/reinforcements to the body late in the game to address structural deficiencies, rather than take the time to engineer the needed strength into the structure in the first place. e.g. his comment about GM looking at torn-down Honda bodies. The “safety” comment probably means they’re not involved in the crumple zones or passenger cage. I’d wager they’re there to reinforce other body panels/joints/welds which were prone to cracking from fatigue due to body flex.
and I was really surprised to see that control arm. Here we thought it was all composite, and it’s stamped steel with an over-molded plastic insert and quite heavy. sounds like they could have just made it entirely out of stamped steel and not had any weight penalty.
just based on his descriptions, I have to agree with him. they spent a lot of effort on an impressive battery and top-notch electronics, and rushed the design of the rest of the car. like the body fits; that kind of variation tells me their sheet metal panels have insufficient locators (tabs, slots, holes) designed in to the panel interfaces.
The Motor Trend article is a little sketchy given obvious errors like “The 2,170 cylindrical Panasonic cells” (there are not 2,170 cells: there are “2170 cells”, with 2170 being the 21x70mm form factor). So I’m not sure how well they translated all of Munro’s work. Well, it has some nice pictures at least.
The Autoline chat had too much snark for me to make it all the way through but Munro’s findings seemed fair. And pretty positive for Tesla, all things considered–the flaws are mostly things that can be corrected in time, whereas the EV fundamentals (pack design, motor controllers, etc.) are great, as well as basic car fundamentals (suspension, etc.).
Tesla has already made very obvious progress on fit and finish–Munro had a pretty early model and recent ones are nowhere close to as bad as what he reported. At one point he said that he could fit his pinky through some gaps and couldn’t fit his fingernail through others. My 3 is, as said, not perfect but there’s nothing even remotely approaching that bad. And the variations that do exist are more of slight trim misalignment, with nothing that could ever affect door/trunk/frunk closing.
Glad you’re finding it helpful! Feel free to ask any additional questions about charging that you might have. The charging situation is certainly a big part of the EV experience so I don’t mind discussion along those lines even if it isn’t precisely Model 3 specific.
Well, there will be only one outlet on that branch–the one for the charger. The only other load is the A/C, which is hardwired.
625.42 does say “Where an automatic load management system is used, the maximum electric vehicle supply equipment load on a service or feeder shall be the maximum load permitted by the automatic load management system.” I would think the relay system counts as an automatic load management system.
Also, it’s a little strange that they have special rules for an outlet that just happens to be used for an EV charger. What if I just asked for a generic 6-50 outlet in my garage?
Nice try, but the NEC definition of “outlet” doesn’t distinguish between a receptacle and something hardwired. 
Good to know. I guess “outlet” can include “receptacle”, but refers to any place where a circuit supplies current to a load.
I think I’ve come up with an even better solution. Well, hacky in a way but more doable both in cost and safety.
The original suggestion by the electrician that came over was to install a switch that allowed me to choose charging or A/C. I’ll go ahead and get that option, with the only requirement that they use a fairly ordinary switch type like this.
I then build a system that mechanically flips the switch based on the thermostat input. I know of the Switchmate, which is a little motorized cover that you put over a light switch so that you can control it over Bluetooth. I probably wouldn’t use that, but instead something more homebrew and perhaps 3D printed. I’d like it to still be hardwired so as not to be subject to 2.4 GHz interference. I know I can do something like this with an Arduino and RC servos. Not a difficult project.
This way, the electrical stuff is all very standard, while still behaving the way I want with regards to giving the A/C preference over charging.
Should be able to use an automatic transfer switch with manual crossover. In fact, it would simply be a standard ATS wired backwards. This is completely safe and electrically valid because alternating current has no polarity. The only “difference” is the terminals where the actual power for the ATS itself comes from. (so a very slight wiring difference)
That reduces the complexity of your arduino project to
-
"connect thermostat signal to input on arduino, using something to isolate the arduino from the 24 V AC it uses. "
-
If signal present, send signal to ATS (using a solid state relay controlled by the arduino, those are about 10 bucks)
-
Wait for it to switch.
-
Close second relay wired across the thermostat control line, allowing the thermostat signal to reach the A/C unit.
So about 5 lines of code.
You would need an optical isolator that will convert the AC 24 volt signal to something your arduino can detect.
Note that since it’s AC it makes the algorithm need a couple more lines of code, since a “low” state (thermostat signal not present) can’t be detected instantly because the AC waveform might be low or negative.
You would need two small relays like these. An automatic transfer switch that has at least a 30 amp rating. There are some for RV use, like this one, available for reasonable prices. (industrial ones are 1-2 grand)
Just make sure you stick everything into a metal box on the wall to contain the fire if something goes wrong! Also, have a separate, *inner *metal box that has the actual high voltage, keep the outer stuff just your hobby project worth of a breadboard, wires, etc all low voltage!
Oh, get a 240 volt power wall-wart to power the arduino and have a 240 volt outlet inside the inner box that takes power from the available 240 volts lines in there.
In reverse :
-
Detect “no signal” for at least 1/30th of a second for the thermostat signal.
-
Open relay to thermostat control line.
-
Send signal to ATS
-
Wait for ATS to switch
Go to start. (don’t literally use the goto command, of course, just have 2 tiny functions, one for each case)
Back to a conversation for the non-engineer types… Dr. S, have you noticed features in the car that still need an OTA update? Like a couple months ago, things like the radio and intermittent wipers were still being released. Any notable functions that you are still waiting for?
It would be interesting if the 2nd motor was an after-market option.
No big missing functionality, but there are a couple items on my OTA wish list.
One minor one: a stop time for charging. It supports a start time, which is very useful, but no stop time. I’d like to limit charging to off-peak hours only.
I’d like to see more work done on the voice commands. As I mentioned before, it works great for music and navigation. But clearly there’s a lot of potential for more, like switching screens on the display (trip meter, phonebook, etc.), general phone integration like dictating texts, and so on.
Incidentally, the electricians I spoke to sounded fairly enthusiastic about my proposal–their main caveat being that the components needed to be UL certified in order to be permitted. Not a tall order but I’ll need to do some more shopping for another relay.
There’s no Arduino needed for the relay version. It’s literally just the thermostat wire connected directly to the relay coil. The relay is wired in an NC state and opens when the thermostat is on. This is exactly how it works already with the relay inside the A/C unit, just in reverse.
The Arduino idea comes from the manual switch version. This is more Rube Goldberg but has the advantage that there’s almost no complexity in the 220v electric parts of it and can be done easily by normal electricians. The complicated stuff doesn’t need any special permitting or installation because it’s just mechanically flipping a switch.
I did your shopping for you. This one meets your requirements. UL listed, 50 amp rating, normally closed, solid state. (you want solid state for reasons explained below). You’ll need 2 of these, one for each AC hot wire.
So you just need a normally closed solid state relay. Upon 24 volt AC signal from the thermostat, it opens the power to the EV charger.
Done. And solid state relay switching is faster than the mechanical contactors inside the AC unit.
So when the AC unit thermostat sends a signal to the air conditioner, before the mechanical contactors can close, the load to the car has already been shed.
In the other direction, the power to the charger comes back up instantly, but turning off an electric motor doesn’t have the surge of current that startup has, and the car’s electronics probably won’t resume charging upon restoring power instantly, there’s probably a delay. Enough time for the contactors inside the HVAC to open.
Update : that relay isn’t in stock at Mouser, this one is : https://www.mouser.com/ProductDetail/Crydom/A2490E-B?qs=sGAEpiMZZMtq49AUx5G377%2F6WBu4BNUZOoooLq91NWE%3D
A bit more expensive, but it’s what they have.
Went on a mediumish trip this weekend.
Mostly freeway driving, but had a varied mix between long 70 mph stretches, medium traffic going at 55-65 mph, and occasional stop-and-go (say, 0-25 mph).
Gotta say, Autopilot is really nice. It does a great job on slow traffic and low-density traffic. It’s actually the medium case where I didn’t care to use it: at those speeds, there’s people constantly changing lanes and merging into traffic. Good drivers will try to stay out of blind spots, be on the lookout for turn signals or more subtle signals for lane changes, cars on an on-ramp and about to merge in, etc. Autopilot doesn’t pick up on any of that. It will take action if someone tries to change lanes into you, but it’s obviously better to avoid that situation in the first place.
For stop and go, though, it’s fantastic. It’s obvious that it reacts to speed changes in front of you basically instantly–before I could perceive a difference, at least. It could easily pick up on a car slowing down just by coasting. It was just aggressive enough with the brakes to be comfortable.
And of course it does great on empty roads. There are some minor flaws–for instance, it does basically fine when two lanes merge into one, but there’s a brief moment when it decides there’s only one extra-wide lane and tries to center itself in it. You end up with a subtle shift in that direction, when it would have been better to follow the line on your side of the merge. Not a huge deal, and probably not something a passenger would notice, but it could use some refinement.
Total trip stats were 292.4 mi, 71 kW-h consumed, and 244 W-h/mi. I probably could have just made it on one charge, but for the sake of battery life if nothing else I decided to try out a Supercharger to give it a bit of a top-off. There were four slots at the station (a relatively small one), and the first stall I tried wasn’t working well–only charging at 17 kW (still 15x higher than at home, but they can do better). Some guy there said the next stall over was also having problems. But he was just leaving, so I pulled into his stall. Maintained about 70-80 kW, and spent about 15 min putting an extra 80 miles to give me some margin. Worked out well in the end, though it was a bit concerning that two out of four stalls weren’t working right. There was a guy coming in just after me that said he worked for Tesla and he’d be reporting the problem.
244 W-h/mi is pretty good, and about equivalent to the ratings. The battery is about 75 kW-h, which makes for a range of 306 mi to depletion. Obviously it’s not great to go down that low if possible. Actually, the second half of my trip was much better than the first–I was around 265 W-h/mi going out, and if the trip average was 244 then I must have been somewhere in the 220s coming back. There was more traffic coming back, which probably (somewhat ironically) helped, since it meant I was rarely going as fast as 70.
A coworker just picked up her Model 3. 1/3 of our local team is now a Model 3 owner :).
Apparently, the delivery process didn’t go quite as smoothly for her as for me. There was some kind of big scratch on it–I didn’t get the details other than that they said they’d repair it right there, so it must have been a relatively shallow one in the clearcoat if they could buff it out there. Still, they should have caught that beforehand.
There were also apparently some problems with the head unit rebooting or something. Not quite sure what happened there, but she did ultimately take the car home (after 4 hours) and I think it drove fine after that.
This was also at the Fremont location so I’m not sure what went wrong, but clearly they have some problems with delivering a consistent experience. Of course, that’s still liable to be less time than at a traditional dealer, and no point where they shoved her in a cramped room with an associate so they could upsell her on floor mats. Nevertheless, they have work to do.