Fitbit One keeps turning off

[rayersttn]

I am having trouble with my fully charged One shutting off during the day.  I've tried the turning it off and back one three times and that doesn't work.   Been happening for 2-3 weeks. 

[MickiSmith]

No, I have an HP PC. My Fitbit had reset when I started my walk. It shut off during my walk today. First time I had this problem. 

[MickiSmith]

I’m having the same problem. It started today. I did the reset. Went somewhere and it had turned off again. So far today it’s turned off several times. I don’t want to waste my time calling customer service if all they will do is offer me % off a new Fitbit since all you have a wrist worn devices and I don’t want that type of device. Hopefully this is just a bad upgrade and you can back it out. 

[Xanadu5651]

I cut the band off my wrist device and keep it in my pocket......I left my
one off for over six months, recharged it and it appears to work
now.....will keep it off for now though.

[Chrissy1]

I bought a BRAND NEW Fitbit One on eBay to replace my other Fitbit One that I had done the "UPDATE" on.

It has been working great for a number of months now. I did not update it even though the prompt is there.

When is Fitbit going to admit that the update is what the problem is???????

[Bork]

I solved my problem with my Fitbit devices by purchasing an HBand. Nuts to
Fitbit!

[Phin13]

Only problem with that is they are calibrated differently. Not that
anything is 100% accurate, but there is definitely a difference.

[Bork]

Yes, but at least the HBand works!

[MickiSmith]

Which wrist device did you get?

[DebbyP]

I use the Fitbit Flex 2 since my "One" died.   I got used to the wrist band, and I am happy with it.Sent via the Samsung Galaxy S10, an AT&T 5G Evolution capable smartphone

[Bork]

An HBand from Amazon.

[Xanadu5651]

I have an older model Surge....bought two so i'd have a spare. It's still
unused.

[biggerwook]

I have the same problem. Whenever I discover that it has turned off it restarts easily, but obviously data is lost. Apparently the company doesn't care.

[MickiSmith]

I agree.  I think they want people to get rid of their clip on trackers to buy their wrist worn devices.  I don't want one so if this continues or gets worse I'll be looking for something besides Fitbit.  They didn't listen when a ton of people told them they wanted a clip on device.  Mine got better and then I charged it and it turned off again.  They had a bad release but don't want to fix it.

[biggerwook]

I don't want a wrist worn tracker. I have a watch on my wrist. I enjoy being
able to toss it in my pocket and forget about it. But now I have to
constantly check to see if it has turned off. Very irritating.

Guess I will start looking for a different company.

[Kevin70]

I guess there was never a solution to this several year old issue.  Just to summarize the issue.

The charges fitbit One turns "off" in mybpocket for no reason (even if fully charged).

Tapping the button does nothing.  Holding the button for a couple of seconds turns it back on.  It displays the firmware version as it comes on (6.64 in my case).

It will work correctly for a while (from a few minutes to many hours).

When in this "off: state, the fitbit does not record steps and loses time.  If in the off state 5 minutes, the fitbit clock will be behind by 5 minutes.  If in the off state 2 hours, it will be behind by two hours. 

Fitbit should have figured out WHY the device is going into this "off" state.  They should have easily known what state the device is in.  What state loses time, doesn't keep track of steps, and can be brought out of pressing the button for a couple of seconds?  What circumstances would the device normally go into that state?

But Fitbit has given NO useful information on this thread (at least the multiple pages I have read).  Some of the support suggestions are laughable (broken button, seriously).

This is either a firmware issue or a hardware issue.  Fitbit should figure out which it is and fix it if it is a firmware issue (or could have a firmware workaround).  My guess is that this is a hardware issue.  My battery is fine, but maybe there is a hardware error signal or voltage level signal that gets in a state that causes the device to power off.  Maybe there is a thermal sensor in the device we that erroneously reads too high causing the power off (though this device should not run hot at all).  

I am about to give up on Fitbit.  I ruined my original getting it wet.  I have now bought two new or almost new Fitbit One devices on Ebay and they BOTH are failing as described in this thread.

I will NEVER wear a wrist device.  I hate them.  Fitbit has no solution for this issue and no good replacement alternative, so it is time I leave Fitbit.  If only they would release a new Fitbit One device.

[Ashleynanette]

Fitbit won’t fix it they don’t make the one anymore and they do it to force
you to buy their new products

[scaler123]

They can't support everything forever. I got a charge 2, thought I would
hate it, but it's actually great. The band's come off easily so in theory
you could pop it in your pocket like a One. You'd lose the HR monitor
though.

[Kevin70]

People were reporting this issue in 2016 (this thread started in April 2016).  The Fitbit was discontinued in 2017.  Fitbit absolutely should have investigated this issue and provided some sort of update on it.  They suggested connecting the device to a computer and doing a reset and that CLEARLY didn't work based on the many responses in this thread.  I fully expect that doing a reset as suggested is just a "catch all" suggestion for any non-simple issues with the Fitbit One.  Either Fitbit did investigate the issue and there is no fix for the issue outside of hardware replacement and thus Fitbit didn't say anything about it, or they never bothered to really root-cause the issue.  They should have done more.

My concern with devices like the Charge 2 is that they are meant to be wrist devices.  Do they actually suggest you can remove the bands and put it in your pocket?  The movement associated with walking for a wrist based device is going to be different than that of a device in you pocket or clipped to your pocket/belt.  I have tested the Fitbit One and it keeps very accurate step count.  I don't know whether putting a wrist device in your pocket is going to give such an accurate step count.

One other item associated with this issue in case Fitbit actually decides to look into the problem (which I know seems very unlikely at this point).  At one point when the issue was particularly bad last week (the device would "turn off" every minute or less), I noticed that the step count kept going back to a previous value.  The device would "turn off".  I would hold the button until it came back on.  The step count would display a value (such as 4812).  I would watch the step count increase on the screen for a handful of seconds (to say 4900).  The device would "turn off" again fairly quicky (not just the screen blank, but the problematic state where it "turns off" and doesn't keep track of step).  I would then hold the button and power it back on.  The display would show (for example) 4812 steps again.

Why is this important (from a debug of the issue standpoint)?  I would guess that the Fitbit One firmware keeps track of steps in some variable in its memory.  They probably periodically (once a minute, once every 2 minutes, or whatever) flushes that memory value into the device's flash memory.  My guess is that when the device incorrectly "turns off", it doesn't flush the memory step count to its flash memory step count.  When you power back on, it probably pulls the step count out of its flash memory, so you lose some amount of steps which hadn't been "flushed" to flash memory yet.  So I expect whatever the issue is, it is a power down event that does NOT do the memory flush/write-back.  Fitbit's engineers should know what causes the One to go into the "off" state.  Whatever those reasons are, some may flush the memory contents to flash and some may not.  This issue appears to NOT do so.  

I think one very possible reason for the issue is an erroneous voltage loss signal.  The battery is fine and charged.  However, there may be a signal from the battery that can get a glitch to indicate a loss of power (or maybe the voltage itself really has an issue and goes down for some reason even though the battery has power shutting the device off).  The shutoff may be controlled totally by hardware which would likely mean this can't be fixed.  However, if this is a signal that the firmware monitors and shuts down the device, Fitbit One could likely do a firmware work-around.  If the firmware sees the signal for which they take action to shut off, they could read it multiple times to confirm power is really lost before shutting down the device.  If it just a signal issue that last for a very small amount of time, such a change likely could fix the issue.  It would essentially be filtering out erroneous transitions of the signal.  Of course, that is just guessing on how these things work.  Only Fitbit knows how the device works.  I just wish they would have provided something useful to this thread.  The devices are out of warranty at this point.  If it is a hardware issue and can't be fixed, just tell us that.  Also, it would be good to know if this is a "degenerative" issue that will get worse as devices/batteries get older.  I don't want to buy another device off of Ebay (even a new one) and see this issue.  Unfortunately, Ebay devices sell for more than the device originally sold for.  

[Xanadu5651]

They didn't fix the issue when they were still manufacturing the One....
why should they now!

[biggerwook]

I don't understand why Fitbit doesn't understand that not everybody wants a
good tracker in a crappy watch. I have a watch. I don't need another one. I
need a credible, accurate tracker that I can put in my pocket or clip on my
bely and forget about it. The Fitbit One meets that need, but Fitbit decided
to that product and couple it with a crappy watch wristband. Sure wish they
would wake up and either fix the One or come out with Two that works.