Time keeps on slipping slipping slipping, on my Ford Transit
February 18, 2022 6:52 AM Subscribe
The clock on my 2014 Ford Transit Connect is wrong. It's also cannot be set. It also keeps moving forward, but not in relation to anything I can determine, and very very very slowly.
On October 19, 2021 the battery in my Transit died, so I replaced it. I tried to reset the clock, but when you click "ok" in the hilariously crude gas station ATM style interface it doesn't take. In the few months that have passed in our common timeline, the time on the clock in my Transit has moved forward very slowly, a minute here and a minute there, and the time on the Transit is now October 21, 2021, 5:35 PM. The Transit is surprisingly nimble, but I do not think this is an issue light speed/relativistic physics.
Having a wrong clock, that changes, is a constant source of amusement and frustration and befuddlement. A blinking 12:00, or an unsettable clock that moved forward in lockstep with the rest of objective time would make more sense and be more easily ignored. This is near constantly puzzling.
The rate at which time passes in there is not constant - sometimes a few minutes will slip by while I am driving, sometimes not at all. Sometimes the clock will be at the same time it was when I previously used the car, sometimes it will have moved forward a minute or two. It never moves forward very much. I can not find a common thread to when it moves.
The fact that it is still October in the Transit leads me to think there is some OTHER battery in there somewhere. But I guess maybe there is non-volatile memory?
I am looking for hypotheses. Something to make this make sense so I can stop thinking about it. A practical fix would be great but is almost too much to hope for.
On October 19, 2021 the battery in my Transit died, so I replaced it. I tried to reset the clock, but when you click "ok" in the hilariously crude gas station ATM style interface it doesn't take. In the few months that have passed in our common timeline, the time on the clock in my Transit has moved forward very slowly, a minute here and a minute there, and the time on the Transit is now October 21, 2021, 5:35 PM. The Transit is surprisingly nimble, but I do not think this is an issue light speed/relativistic physics.
Having a wrong clock, that changes, is a constant source of amusement and frustration and befuddlement. A blinking 12:00, or an unsettable clock that moved forward in lockstep with the rest of objective time would make more sense and be more easily ignored. This is near constantly puzzling.
The rate at which time passes in there is not constant - sometimes a few minutes will slip by while I am driving, sometimes not at all. Sometimes the clock will be at the same time it was when I previously used the car, sometimes it will have moved forward a minute or two. It never moves forward very much. I can not find a common thread to when it moves.
The fact that it is still October in the Transit leads me to think there is some OTHER battery in there somewhere. But I guess maybe there is non-volatile memory?
I am looking for hypotheses. Something to make this make sense so I can stop thinking about it. A practical fix would be great but is almost too much to hope for.
Wow. This is a fun puzzle. I know nothing about cars. But, my very naive guess is that there's probably some bad wiring, such as ungrounded inputs to a chip or ground loops and also radio frequency interference or power supply issues that are causing problems with the clock-signal that drives the display clock such that it's missing most of the ticks in an intermittent way. (Could also be a failing oscillator, which is probably a one dollar replacement and a real pain in the neck to get to.) Is it easy to physically get into the thing and stick a volt meter or oscilloscope on it, if you have such tools or a friend with them?
(You could also tape a cheap LCD clock over it, or put opaque tape over it and stick a clock elsewhere, I guess. Or embrace your new life as a time-traveler.)
[edit: does it ever go faster than real-time?]
posted by eotvos at 7:42 AM on February 18, 2022 [1 favorite]
(You could also tape a cheap LCD clock over it, or put opaque tape over it and stick a clock elsewhere, I guess. Or embrace your new life as a time-traveler.)
[edit: does it ever go faster than real-time?]
posted by eotvos at 7:42 AM on February 18, 2022 [1 favorite]
Digital clocks usually base their timekeeping on a high speed oscillator, often locked to a 32768Hz quartz crystal, followed by a counter circuit that turns an incoming 32768Hz pulse train into a 1Hz pulse train, followed by more counters that accumulate seconds, minutes and hours. Modern implementations usually incorporate all of the above into a tiny low-power microcontroller that can also do things like drive displays and read buttons.
An oscillator is essentially an inverter (if an inverter's input is low, the output goes high and vice versa) with its input hooked up to its own output via some kind of delay, so that a short amount of time after the output goes low, the input will see that low, which makes the output go high, which after a short delay will make the input see that high, which makes the output go low again and so on.
From your description, I would guess that the master oscillator is not running properly at present, perhaps because it's been damaged by a power spike caused as the new battery was being connected, or perhaps because it's become locked up in a metastable condition where the output is neither low enough to be properly recognized as a low nor high enough to be properly recognized as a high and there's consequently not enough activity to make the quartz crystal resonate.
As a result, the counter circuit will be seeing what looks like random noise rather than a nice clean 32768Hz pulse train, and its output will be a divided-down version of that same noise rather than a nice clean 1Hz pulse train. And if the hilariously crude gas station ATM style interface is being implemented by a microcontroller that's also now being clocked off random noise instead of a nice consistent high speed pulse train, the amount of time it takes to respond to button presses is also going to be wildly variable.
If the oscillator has stopped because electronics have actually been damaged, you're unlikely to be able to fix it. But if it's just stuck in a metastable you might be able to unstick it by depriving it of power, then leaving it deprived until all its own internal power filters have had time to discharge completely (give it a whole minute to be safe), then powering it up as cleanly as you can so it resets itself properly on the way up. Best way to do that would be to find the fuse that supplies it and pull that, rather than fiddling about with heavy duty battery leads that will almost always cause fairly long-lasting noise bursts as you undo and redo their connections.
posted by flabdablet at 8:07 AM on February 18, 2022 [9 favorites]
An oscillator is essentially an inverter (if an inverter's input is low, the output goes high and vice versa) with its input hooked up to its own output via some kind of delay, so that a short amount of time after the output goes low, the input will see that low, which makes the output go high, which after a short delay will make the input see that high, which makes the output go low again and so on.
From your description, I would guess that the master oscillator is not running properly at present, perhaps because it's been damaged by a power spike caused as the new battery was being connected, or perhaps because it's become locked up in a metastable condition where the output is neither low enough to be properly recognized as a low nor high enough to be properly recognized as a high and there's consequently not enough activity to make the quartz crystal resonate.
As a result, the counter circuit will be seeing what looks like random noise rather than a nice clean 32768Hz pulse train, and its output will be a divided-down version of that same noise rather than a nice clean 1Hz pulse train. And if the hilariously crude gas station ATM style interface is being implemented by a microcontroller that's also now being clocked off random noise instead of a nice consistent high speed pulse train, the amount of time it takes to respond to button presses is also going to be wildly variable.
If the oscillator has stopped because electronics have actually been damaged, you're unlikely to be able to fix it. But if it's just stuck in a metastable you might be able to unstick it by depriving it of power, then leaving it deprived until all its own internal power filters have had time to discharge completely (give it a whole minute to be safe), then powering it up as cleanly as you can so it resets itself properly on the way up. Best way to do that would be to find the fuse that supplies it and pull that, rather than fiddling about with heavy duty battery leads that will almost always cause fairly long-lasting noise bursts as you undo and redo their connections.
posted by flabdablet at 8:07 AM on February 18, 2022 [9 favorites]
i have a 2016 transit connect as a work vehicle and it's infotainment/electric system has been an absolute garbage fire. i can't speak directly to your issue, but mine has degraded steadily, starting from a lil bit of static coming out of the speakers. then the buttons stopped working intermittently, the bluetooth stopped working, the entire head unit completely failing intermittently, then wild full-volume blasts of random static from the speakers from which there was no remedy, since the thing can't even be turned on or off anymore since the buttons and knob don't work. finally it just completely died, no static blasts anymore, no nothing. every once in a while the screen will boot up randomly and chide me that 911 Assist requires a phone to be connected. now i just pretend it doesn't exist and carry a bluetooth speaker with me to listen to tunes when i'm driving.
the backup camera still performs flawlessly, though.
posted by glonous keming at 2:42 PM on February 18, 2022 [1 favorite]
the backup camera still performs flawlessly, though.
posted by glonous keming at 2:42 PM on February 18, 2022 [1 favorite]
Besides the fuse thing I'd try cleaning the hell out of the battery terminals and clamps. Lead oxide is a crappy conductor and all sorts of weird things can happen when you combine a unexpectedly high resistive circuit with the very low draw of the clock circuit. I have to do this with my spouse's car approximately annually and while ampain in the but I'm always glad it isn't the fuel pump or starter or whatever.
posted by Mitheral at 3:11 PM on February 18, 2022 [1 favorite]
posted by Mitheral at 3:11 PM on February 18, 2022 [1 favorite]
According to this forum post disconnecting the battery and leaving it off for 5-10 minutes to completely drain the car may reset the clock.
Another posted suggested removing the NEGATIVE terminal for 2 minutes.
posted by kschang at 4:23 PM on February 18, 2022 [1 favorite]
Another posted suggested removing the NEGATIVE terminal for 2 minutes.
posted by kschang at 4:23 PM on February 18, 2022 [1 favorite]
This thread is closed to new comments.
posted by fritley at 7:04 AM on February 18, 2022 [1 favorite]