Charging directly from an off grid solar electric system, no inverters involved.

Charging directly from an off grid solar electric system, no inverters involved.

Charging directly from an off grid solar electric system, no inverters involved. I would like to find a way to charge my tesla three in the future by simply going DC from the panels directly into the car battery. So far no one at Tesla will give me any clues.

DonS | 14 June 2016

Your solar system would need to impersonate a supercharger to be able to charge with DC. For AC charging, the signaling protocol is the same as the J1772 standard, so you would need to sniff the bus during DC to figure out the protocol for DC charging.
Although going direct DC to DC would be more efficient, it looks to be non-trivial to figure out, and potentially expensive if there is a mistake. Given the growing number of items that generate and use DC power, I think DC some connection methods will be standardized within the next decade.

topher | 14 June 2016

I asked Elon Musk about this but got no reply. The voltage is likely to be the biggest issue.

ir | 14 June 2016

I don't think solar panels output 400V. If you cannot overpower the voltage of the pack, the power will run out of the battery and into the panels. Poof! Your panels explode and your house burns down or you get electrocuted.

The reason you can use 100V - 240V AC is that the onboard chargers have capacitors that build up 400V and then dump it into the battery in a charging pulse. But DC charging bypasses the onboard chargers so you're on your own to make it work.

You would need at least some kind of DC-DC converter to get the right voltage and it needs to talk the supercharger protocol because the battery power needs decline as it gets full to avoid exploding.

In the end a pure DC solution would probably cost more than a regular solar DC to AC inverter and probably set your car on fire anyways.

bb0tin | 14 June 2016

What is your off-grid system?
Do you have AC wiring in your house and a DC-AC inverter?
Do you have DC wiring in your house and a DC-DC converter?

tgretz | 15 June 2016

The power coming in from the solar panels is also known as "dirty" power. It is not a clean wave and the inverter is what converts the solar DC into clean (usable) AC power.
Even with my 10kw (40 panels) of solar going directly to the car it would take 8 1/2 hrs for an 85kw pack to charge if the sun is steadily producing 10,000 watts per hour....never happens.....

PV_Dave @US-PA | 15 June 2016

Charging from a DC power source is SuperCharging. Much like the CHAdeMO adapter emulates a SuperCharger even though it's not one, you'd need to do the same thing, and it'd probably void your warranty even if you figured out how to do it correctly.

This is not a trivial undertaking, or one with a high probability of success as a hobby project.

bb0tin | 15 June 2016

You said "The power coming in from the solar panels is also known as "dirty" power. It is not a clean wave and the inverter is what converts the solar DC into clean (usable) AC power."
DC stands for direct current. It does not have a wave form, clean or otherwise.

lec_maj | 15 June 2016

Perhaps with the Tesla battery, anyone want to design this together? Ping me on twitter or otherwise. I am looking to solar up + battery off grid.

Badbot | 15 June 2016

I think DC for the tesla is 24v.
DC has no wave. as long as the voltage is above the pack voltage it will charge.
Just keep it under 30v and you should be good

bb0tin | 15 June 2016

You think wrongly.

ir | 16 June 2016

@Badbot: DC for Tesla is 400V

SUN 2 DRV | 16 June 2016

Tyler: You're trying to jump straight to a solution, without telling us what problem you're trying to solve. It's likely to be a lot more practical to use the typical Solar->Inverter->Charger->Battery architecture which already exists and is virtually Plug-n-Play vs your proposed Solar->??->Battery approach.

So what problem do you want to solve regarding the typical AC interconnect approach?

lolachampcar | 16 June 2016

Efficiency is likely one. Why tolerate the DC-AC-DC losses if you do not need to.

Start with the open source chademo project to get coms to activate the DC switch in the Tesla.

Use a MPPT tracker circuit together with a boost (or buck depending on how many panels you want to string together) to pull power from your panels. Use the MS coms to tell you when to taper current or simply charge only to the level where your panels are the limiting factor.

Easy in theory but too much of a pita for 20% in efficiency.

Badbot | 16 June 2016

"ir | June 16, 2016

@Badbot: DC for Tesla is 400V"

I was remembering the following,
16 modules, each with 6 groups of 74 cells = 444 cells, producing 25 volts.

I knew the inverter delivered 400v AC. I'm not sure where I missed that all the 25v packs were put in series.

bryan.whitton | 16 June 2016

You would need to build a unique DC voltage charge controller. They are common for the standard 48V batteries typically used in stationary applications( You would need a charge controller suitable for the battery voltage for the Tesla.

My question is why? Just get an inverter(or microinverters) and tie to the grid. That can be your battery at least in the near future. At some point in the future when you have solar you will need batteries for self use. In the meantime the technology just isn't there.

WHitchings | 17 June 2016

The OP is just trying to avoid DC-AC-DC conversion, that's all; I've often wondered the same thing. If you put in a Tesla Powerwall to store the day's solar energy, you're looking at DC -> AC -> DC -> AC.

The short of the story is that AC's primary purpose (thanks, Nicola) is that it runs vast distances with relatively little loss. It's worth the conversion of DC -> AC to be able to place solar panels away from their use or storage media.

lolachampcar | 17 June 2016

I thought I read that the latest Sunnyboy could handle 400V packs?!?!

ir | 17 June 2016

@badbot: The submodules are in series with only the main contractor (there is no way to target charging to 1 submodules). Check out the battery pack teardown and analysis:

Post #7:
However, the BMS has no authority, per se, over what is actually done with the power.

The supercharger just feeds in high voltage high current DC directly to the pack. The BMS monitors the pack to help tell the supercharger how much power to feed.

However, from a technical standpoint, with access to the HV endpoint connection it is basically just one big HV battery with no more granularity than that available to the charger. The charger cant choose which cells or modules to charge. It can only charge them all at once.

From what I can tell the BMS is only capable of monitoring the modules. They do not seem to have the ability to charge or discharge individual modules or module sections.

I could be wrong, but I do not see any connections on the BMS that would facilitate such functionality. They seem to be watch-only.

ir | 17 June 2016

contractor -> contactor

ir | 17 June 2016

It is this inability to target individual battery cells / submodules that makes charging the traction battery so risky.

As the pack approaches full, some submodules could be at 100% while others need a bit more power. If you do not carefully taper the charge current, you could "blow" one of the full modules while trying to squeeze power into another.

The supercharger control protocol allows the car's computer to throttle back power when the various BMS modules report their modules are nearing 100%. You would need some kind of DC charger between your solar panels and Tesla to wind-up and carefully deliver a 400V pulse at the right amperage for the right duration according to the car's requests.

topher | 17 June 2016

"I don't think solar panels output 400V."

Solar cells (The little 4" square pieces) produce a voltage difference of 0.5 volts. In order to get higher voltages you need to put many of them in series. But there isn't a limit, 800 cells is 400 volts. Solar panels don't generally have 800 cells, nor do they arrange them all in series. But you can wire _panels_ in series to increase their voltage. 33 Panels at 12 VDC apiece, gives 400 Volts.

Thank you kindly.

PV_Dave @US-PA | 17 June 2016

+1 @topher: It's all a question of how you wire them, and what your inverter can handle. If all my home's PV panels were in series, they'd produce over 3,000V open circuit, and 2,750V at peak power production.

SUN 2 DRV | 17 June 2016

Whitchings: Yes that's apparently true... but he still hasn't said why... specifically what pain he's trying to remove...

Is he simply allergic to AC waves? :-)
Is it a cost issue with what he perceives as unnecessary hardware?
Is it a space issue?
Is it really an efficiency issue to him? Is it worth the cost?
Is grid backup not possible or is it some arbitrary commitment to self sufficiency?
Does he want to do it once for himself or does he plan to sell this solution?
And if he plans to sell it, what benefits (from list above) are important to his potential customers?

Understanding what he REALLY wants to accomplish goes a long way towards identifying a workable answer.

topher | 21 June 2016


The reason I asked EM, was efficiency and simplicity (thus cost) for solar systems.

Thank you kindly.