View Single Post
Old 07-07-2015, 10:37 PM   #11 (permalink)
Gandalph
EcoModding Lurker
 
Join Date: Jul 2015
Location: Natick, MA
Posts: 8

Orangemobile - '11 Honda Fit Sport
90 day: 41.42 mpg (US)
Thanks: 0
Thanked 0 Times in 0 Posts
Quote:
Originally Posted by kafer65 View Post
On the solar 12v assist angle, how can you safely apply this to a modern car? I've seen photos of car fires from dash PVs plugged into the power plugs on some cars. I have a Volkswagen solar panel that has the car plug (and it makes plenty of voltage in full sun) but I'm apprehensive about using it in my Mazda. I know that there is less draw down when modules go into sleep mode on my old Jetta and if you interact with the car (like unlocking it) the modules wake up. Some cars shut down the power plugs in sleep mode and some don't. If I start measuring voltage on the plugs I may wake up the modules and nullify my findings? You've attracted ecomodder heavyweights here and being an EE graduate I wonder if anyone had some ideas on the matter?
There are unfortunately quite a few ways to accidentally dissipate power and start a fire in an electrical system (and I've done a few of them myself). Here's why routing through the main plug would maybe not be such a good idea:

-I'm not sure how that panel is set up. If it has its own integrated charge controller, it's OK to connect it directly to the car's electrical system because the controller will be regulating the output voltage that the devices in the car see. If not, then the voltage from the panel might not be a good plan to pump directly into the car's electrical system, as the car's other accessories could be exposed to voltages higher than what they were designed to be exposed to (the panel might put out higher than the approx. 14V that the electrical system will be running at when the battery is at full charge). This could cause some of them to fail in spectacular fashion, especially the battery. Charging at too high a voltage can cause a lead-acid battery to produce hydrogen gas instead of lead metal in its redox chambers. Hydrogen can explode and catch fire (a-la Hindenburg).

-The accessory plug usually doesn't have a very high rating for how much current it can take (the one in my car is rated to 10 amperes, which you could easily go over with a big enough panel). Just like wiring in a house, if you push more current though a wire than it's meant to support, it will start heating up beyond its design tolerances and could catch fire. I'd assume that this is protected by a fuse on most cars, but it doesn't pay to take a chance that you might cause a fire.

So the way that I was going to handle it is with a cheapo solar charge controller that's already meant to be used with 12V lead acid batteries. This will ensure that the other devices in the system will not be exposed to potentially dangerous voltages from the panel, as the charge controller will step the voltage down to a level appropriate for the car's electrical system.

Most of these controllers also do Maximum Power Point Tracking (MPPT), which changes the working voltage and current from the panels to ensure that they are always operating under a load that maximizes transfer of power from the panels to the controller (and getting you the most juice that you can get out of a given amount of sunlight with a given set of panels).

Such a controller that can easily handle the currents that we'll be working with on a single panel installation can be found on Amazon for around 30-40 dollars (look for "Solar charge controller" and pick one that has high ratings, that's what I was going to do).

As for where this will go, I was planning to mount it under the hood somewhere. These devices are pretty efficient, but they are not 100% efficient and will therefore generate some heat. Having that inside the cabin of the car is asking for the heat coming off the charge controller to start a fire.

As for wiring, I was planning to wire the controller directly to the battery terminals to ensure that there is a minimum amount of wire that could potentially heat and lose useful power/start a fire.

Some controllers can also be configured to dump power into a load when the battery reaches full charge (which would be useful to for instance dump excess current into the blower motor to run the fans when your car is parked in the sun on a hot day, or run a small block heater when it's cold in the winter). I still have some experimenting to do to see whether the blower will run on any amount of current that it's given, or if it's designed to run only on the few currents that the car does when setting the dial from inside the car.

There are a few things that I haven't worked out:

-How will the charge controller react if I swap out the battery for a bank of supercaps? Those don't react quite the same way to being charged as a battery does, so the controller might have difficulty working with them. This will require some experimentation to prove safety and reliability.

-If the car decides to turn the alternator on at the same time as the charge controller is on, they may fight each other (the power supplies may be at different voltages, leading to uncontrolled flow of current between the supplies). This is bad news (and a potential source of a lot of heating and fire), and we need to prevent this from happening. I was thinking to either:

Option 1:
Set up protection diodes to prevent current from flowing backward into either supply.

Advantages:
-Easy set up
-Cheap power diodes are readily available from distributors like Digikey

Disadvantages:
-Diodes always have a voltage drop to turn on, and this means a current draw will cause them to dissipate power (read: heat will be generated and energy lost). Because the car's electrical system is low voltage DC, it draws a lot of current, leading to significant power loss in the diodes. Metal-semiconductor junction diodes have lower voltage drop and may be a good plan if we decide to go this route.

Option 2:
Set up a system where there are two separate electrical systems for the car, one that runs from a battery supplied by the panel, and one that runs from a battery supplied by the alternator. The two systems are isolated by a DPST relay that selects between the two sources (maybe selectable from inside the cabin, maybe selected automatically by whichever one has more stored charge).

Advantages:
-Keeps more power in the system, electromechanical relays don't have much voltage drop across their contacts and don't dissipate much power as a result.

Disadvantages:
-I need two batteries, this is heavier and also requires spending more money on a second battery.
-Electromechanical relays that can handle the kind of current that a car's electrical system uses might be pretty expensive.
-Electromechanical relays are not as reliable as solid state diodes. They will eventually stop working correctly and will need replacement.

Option 3:
Use Some Power FET's as switches to cut off supply to the battery from each device as required.

Advantages:
-Power FET's are readily available (found over 100 on Digikey this morning that can handle my alternator's max output), cheap (less than $5 per), and offer low on resistance up to very high currents (one I looked at had 1.2 mOhm on resistance at 100 amperes).

Disadvantages:
-Because of the body diode, they can only effectively block current flowing in one direction, and the body diode typically has really bad forward voltage drop, so I don't want to configure it to go through the body diode under normal conditions. This means that I need a total of 4 transistors and to put them in sets of two with the body diodes back-to-back. But doubling the cost and complexity of a setup that's already not particularly long or complex is not a huge deal.

EDIT: Added discussion of option 3 for safety blocking.


Last edited by Gandalph; 07-08-2015 at 06:50 AM..
  Reply With Quote