Hi folks!
I've had my TI84 CE for just over a month so far, and in that time the internal clock has on 5 separate occasions "lost" a day! It keeps time correctly, but the date is one day behind reality. None of my previous TI84 models did this. How is this even possible? Has anyone else encountered this? I'm wondering whether it's even worth claiming on the Warranty (although it's a relatively minor fault so probably not worth the hassle).

I must say that I haven't experienced this, but it sounds more like an OS error than a hardware problem to me. Do you turn it on at least once every few days? I'm setting the clock on mine properly, and I'll keep an eye on it to see if it is losing days.

Thanks. I haven't managed to pin down when this seemingly-random fault occurs, but it's just done it again. We're past midnight here, I know it said June 13 yesterday, because I edited it from the previous instance, and now it says 0:53:32 on June 13 again.

is it only misreporting date within one hour of midnight? Sorry if you explained before - i may have missed it reading the forum on a mobile. If it is within an hour of midnight, could it be just some summer saving mode it does not take into account properly?

I set my clock yesterday as well, after reading this post. It now reads the 13th. Going to reset it again, and see if it fails again, to confirm the issue on my calc.

Edit:

I set my 84+ and CSE calculator clocks as well, so we'll see if any of them suffer this failure, or just the CE.

I'd gander a guess that it's something to do with a 24h clock. Switch it to a 12h clock and see if that helps with anything.

You can speed up testing by changing the clock to 12h then changing the time to 11:59p and seeing if the date changes when it rings as 12:00a. Repeat process with 24h clock; 23:59 and it the date doesn't change when the clock changes to 00:00 then that's likely the pinpoint of the bug. For the sake of being thorough, while still in 24h mode move the clock forward to 11:59 and watch what happens when the clock changes to 12:00.

This is what I would do when faced with this situation.

Thanks for all the posts, guys. I tried all the suggestions in your post, comicIDIOT, and the date didn't misbehave at all (and neither had it done so Sunday night, when I didn't change anything). But although I appreciate the attempt to analyse the issue scientifically, fact is it hasn't been happening predictably anyway. See the OP: 6 times in about a month. Intermittent faults are always the weirdest and toughest to track down.

For what it's worth, after I set my clock correctly on Saturday, at 12:05ish on Monday the clock said June 14th, despite it actually being a few minutes into the 15th. Like OldMathTeacher, I was in 12-hour mode, so this is an eminently replicable bug.

I actually wasn't using it in 12-h mode (except briefly today when running through comicIDIOT's diagnostic suggestions). I have some programs that read and use the built-in date, so I am likely to notice if this goes astray. And the 84+SE and CSE never did so in all the time I've had them: only the CE OS seems to have this unreliability.
It's not a big deal really, except for the mystery: how can this sometimes (but not regularly) happen? In a way, I'm glad it's not just me: perhaps this increases the likelihood that's a fault that will get ironed out in a future OS release.

I wanted to bring this back up because it seems that the problems that you guys are having is when the calculator is around midnight when it is switching over to the next day. If this is so, then it could be that the calculator was misprogrammed to switch the date at a time other than midnight.

No, I'm referring to instances where it loses a whole day; the calendar doesn't advance at all (sometimes), so without user intervention it will remain one day behind, until the next time it loses a day.
I've never known it to advance at a time other than [its own] midnight.

Posted July 14th:
I'm now looking at the theory that if the calculator is in use (maybe has to be running a program) when it passes through midnight, the clock keeps time, but the calendar forgets to advance. If so, that would account for why it only happens some days and not others.
Any related observations welcome.

Edited August 1st (by the real-world calendar):
No, that wasn't it; I have sufficient evidence that the glitch can occur when the 84CE is not in use, and can fail to occur even though it is in use (and certainly vice versa, as established previously). I have detected no pattern to this intermittent fault at all.