waldguy

The only place that seems convenient might be a corner of the trunk, depending on size or type. A simple battery isolator kit would ensure you never run down your battery too much. I'm sure you've also considered a lithium Ion battery pack, which would be totally independent and portable -- go sit at the picnic table by the lake if needed.

Very good point X-TSub. Some wires may even be copper vs aluminum. Most of us will also know from experience what head scratching and frustration faulty grounds will cause. I tend to use lithium grease on connection I service from time to time, but connections like the one listed here would not likely make that list. Tef Gel looks better, particularly in salty environments, but then the price is dear too (over $75 CAD for 2 oz shipped on Amazon!). I wonder if it's a mouse deterrent....

Thanks for the follow up. I was curious! Yes, that box makes it look like it does something -- why the insulation around the connection when it's just to ground! Glad to hear you're making good progress.

Shouldn't be hard to open. I'd wonder about what appears to be a red wire showing. That's not likely to be grounded.

Pretty sure I used some WD40, and some Kerosene another time. Others say "Goo gone".

I used the dongle, and an app for the CVT temperature. Relearn might be useful in a full fluid replacement, but not flush and fill of 5L or so. I use Android and have the ActiveOBD installed for the data read.

I can't say where your smell is coming from but it's not likely the oil pan gasket. That would just leak to the ground without burning. I know that I had an oil smell and it turned out to be the front crank seal -- the oil would leak and drip onto the exhaust area. It was repaired under warranty and I don't remember the hours of labor.

Wow, that must have been a headache. I thought 2 hours of hammering and prying was bad on my rear wheel bearing. My backing plate was bent a little, but recoverable.

High prices are the mother of invention. One such video on a Subaru:

Could be a case of a faulty connection, some surface corrosion perhaps? Also, sometimes the little prongs need to be bent back out again as well to create a tighter connection.

You're right about the partial CVT fill. I did mine at 50,000 miles and intend to do it again at about 80,000 miles. My drained fluid looked very good. Blackstone is a good idea. Any results yet?

TSB was done yesterday along with some others. Tech at dealership didn't know why it happened either. He just suggested keep turning the key off on off on off on lots of times. He said it was typically more of a problem with the push button start. Apparently with one of the updates, he said, the battery will not be charged at all if the voltage drops below 11.6V. That hasn't been a problem. Just the same, I picked up a cheap load tester for occasional use -- car, boat, truck, etc.

Yep, just be aware that a slidehammer may be insufficient. Prying and hammering with an experienced mechanic took over an hour before mine finally unseated. Yours will hopefully be easier.

Thanks for the ideas. The 2015 model battery charges not with necessarily with long runs, but with periods of deceleration due to the "smart" ECM management for fuel economy, or during night loads. On the night in question, the battery was likely at about 12.25 volts, perhaps just marginal enough not to crack the relays if "stuck" or frozen. In addition I now found out that the plugin source breaker was turned off -- so the engine did not have the heated oil advantage. The 0W20 flows relatively well regardless. Nevertheless, the key turn at first did not engage the starter at all. When it finally did, the engine turned over well enough. So in this case I think I can rule out a faulty heater block cord. How am I addressing this? For the battery, I read and confirmed that a full charge can be force during daylight if I put on load like parking lights. I may also get the battery TSB done (trip to the city and $75). I also gave the battery terminals a slight wiggle and extra tighten. I also confirmed that the relays are indeed the new ones (eg. M09) as specified in the relay update.