Filled up at a different gas station Wednesday night, yesterday morning, engine was really pulsing when I first stated it, only 5K miles on it now, but it cleared up in 30 seconds and was good the rest of the day. Figured I got a little bit of water in the gas that would settle over night. Naturally the fuel input is at the very bottom of the tank.
Last fall, went up north in our motorhome, got gas up there with a lot of rain, that was dumb, put some in both the front and rear tanks, also dumb. When we took off, didn't get very far, engine stalled in no-mans land. Fortunately could pull to the side of the road, had a tool box, and clear gas filters, added an electric fuel pump, and a flash light. Could crawl under, disconnected the fuel line, had my wife switch on the fuel pump until I got clear gas. We made it home okay, but then the rear tank, this I could put in a container. Got out about three gallons of water.
Relay contacts are only used by the radiator fan, blower motor uses a PWM power MOSFET transistor, wow, sure love the speed on this thing.
To me, this sounds like another lack of power on reset problem, BCM will go crazy without a proper microcontroller reset. Sure was the problem with my thermostat heater, but with component tolerances and a ton of computers in this thing, not sure what would go wrong. With my old Cruze, was the radio first, but then the ECU, car wouldn't even start 40 miles from home. But removed the key, opened the door, second time doing this it started. In this thing was the ignition relay, darn thing switches everything on. In my Elantra, was a loose negative battery cable, been perfect ever since. But did check both ign 1 and 2 relays, have precision test equipment for this, they were good.
Never had computer problems before when we didn't have computers, slide rule worked perfectly and nobody could hack it. Never had computer car problems before when before they added computers. Motorola came out with the first processor in around 1970, was a one bit unit, claimed a lot easier to write software than designing hardware. First ones used a separate port for the instruction codes, another for the data. But with a 4 or 8 bit marketing got involved, cost 2 cents extra for a pin, oh my God this is adding 16 cents to the cost!
So multiplexing was used for both ports, just invented the crash, been this way every since. In the 80's, cost 16 cents for a byte, again marketing, only used a two digit year date code, looked what this caused in Y2K. Been told a million times to use a 89 cent microcontroller by marketing, sure, but you only gave me a 4 mm square to put this in. As we know today, hardware cost next to nothing, but watch out for that software that has to be updated every five minutes. $$$$$! Ha, getting carried away again.
Radios use to be radios, now they are computers, When a real blower motor switch was used with resistors, didn't have these kinds of problems, sure do today.
Another little problem was putting in high impedance CMOS technology with less than a volt noise immunity into a vehicle loaded with motors and solenoids that have the Tesla effect, generate extremely high voltage that can mess up this logic. Layout was more of an art than a science, just invented EMI. Took them long enough to get rid of those spark plug wires, another battle between engineers and bean counters, 40KV!