RESOLVED: Steps count increase while charging Fitbit Versa

[AlessFitbit]

Fitbit Update: 9/09

 Hi, Versa Fam! 

Thanks so much for your patience while our team continues to investigate and work on this issue. We're still working on this and I hope to have more information for you all very soon. 

As soon as there is a fix in place, I will make sure to announce that here. Please stay tuned and subscribed to this thread for updates. 

---

Fitbit Update: 7/25

Hey, everyone!

I know many of you are wondering what the status on this bug is and I'm here to reassure you this is still being worked on by our team. For anyone still getting ghost steps on their Versa's, we are definitely working towards a fix. I don't have a timeline for when the fix will be released, so I appreciate everyone's patience for the time being. I recommend if possible, placing your Versa on a different flat surface without any vibrations to prevent the extra steps. 

---

Fitbit Update: 6/21

Hi, everyone!

Sorry to hear that some of you are accumulating unwanted steps while charging your Versa's. Thank you for taking the time to report this issue here on the forums. Our team is aware of this affecting some of you and is working towards a fix on a future firmware release.

Thanks for your patience for the time being. 

[MarcoGFitbit]

Hello everyone, I hope you're doing well. 

I would like to start by thanking you for all the feedback and insights you've provided about the Fitbit Versa. I understand you're experiencing some issues with it and I would like to apologize for this situation. Please be sure that our team is working hard to address all your concerns, which is why this thread has been pinned in the Versa board. Additionally, please note that our team has started releasing a new firmware update in hopes to resolve many of these issues you're mentioning. However, this new update has only been released to 10% of users so our team can monitor its efficiency. Keep an eye on this thread for future updates about this new firmware version. 

Thanks again for your patience and understanding. Feel free to continue sharing your experiences, all the information provided is helpful for us. 

Happy stepping.  

[lenny4d]

@Nu2_4rum"So is beef full of antibiotics." - yes and it is still beef 🙂 The two-electrode pseudo-EKG cannot replace a full 10+ electrode professional EKG but it works pretty well. Two electrode designs are commonly used in portable EKG equipment, you can find all of it on the web.

Wether that is useful to have in a watch is another matter. Would it not be much easier to have a kind of 'health-station' (just like there are small weather stations), with - as a suggestion - proper SPO2, proper HR, proper EKG (well, not 10 electrodes obviously), maybe even Blood-pressure, the whole home-health-pack. Interfaced via bluetooth to the fitbit ecosystem (or any ecosystem for that matter, but I am happy to give the favour to fitbit). Better quality, for those that need it, without bloating the wrist-based hardware... What do you think?

"This surprises me..... I could be very wrong but keeping it in a pocket sounds a bit of a rarity."

There are several threads on this on the fitbit community, in fact in various professions one is not allowed to use a watch during work and thus have to keep it in a pocket. It is much less a rarity than you think. As a sidenote your comment is exactly what people complain about wrt to fitbit, it could be an out-of-context carbon copy of the fitbit helpdesk. Think about it 🙂

[slellis128]

Has anyone who keeps an eye on this post received the firmware update that was recently released to 10% of Versa users? If yes, has it fixed the issue with the ghost steps?

[RJayR]

It would help to know what version of the firmware is the "new" one. Mine's on 32.10.15

[Exactcopies]

A previous Fitbit Moderator post said the beta version released to “10% of Versa Users” is 32.10.19.

James R. Gordon

[Glenmcd]

---

@RJayR wrote:

It would help to know what version of the firmware is the "new" one. Mine's on 32.10.15

---

Is this what you are looking for?

[Exactcopies]

A previous Fitbit Moderator post said the beta version released to “10% of Versa Users” is 32.10.19.

[lenny4d]

According to the new firmware release threads in the community there are many problems with it and it is back into the lab. So doubt we will see it anytime soon, even less that solves some of the spurious effect. That being said, I STRONLY recommend that those people that seem to have ghoststeps while charging to try using a USB battery a/o their PC USB port instead of the wall-charger and see if the problem disappears (and report here). If it does, well then you know it's something to with the charger and maybe some **ahem**ty charging voltage that interferes with the electronics. If it doesn't, now then that would be really really interesting. For the record, I don't have any ghoststeps, nor when charging, driving, sleeping etc.

[Nu2_4rum]

OK, here’s something else to try:

Suppose the GPS coordinates from your phone fluctuate due to poor signal or atmospheric effects or whatever.

Try turning off GPS on your phone at night. See if the watch still records ghost steps.

Either that will fix the problem or it will not. Either way, it will tell us something new.

Sent from Mail for Windows 10

[lenny4d]

@Nu2_4rumJust for the record : GPS doesn't really work indoors when you are asleep, it just doesn't produce anything (unless you are camping out that is). Secondly, the GPS (coonected or build-in as for the Ionic) is only used in exercise mode, and thirdly (and as mentioned before) steps are independent of GPS.

But of course if you suspect with good reason that somewhere there is a bug that steps can be recorded due to (absent/weird?) GPS signal even when not in exercise mode and even when steps supposedly independent of gps in the firstplace, yeah look forward to the result.... 🙂 For the record, we're talking about 1,000s of steps, not just the odd 10-100.

[Nu2_4rum]

You seem very negative about the possibility of GPS being a relevant factor in regard to night-time ghost steps. You may well be right. But it is still useful to rule this out by experiment to back up the logical conclusion. (In ancient times, philosophers believed they could discern truth by thought alone. The big steps in modern science came from those who the philosophers looked down on, the ones who got their hands dirty doing experiments, which was considered to be beneath the dignity and social class of a philosopher.)

The concern I have is the possibility of jitter in the GPS coordinates that might be interpreted as motion. Jitter is a common problem in electronic systems, particularly digital ones. The GPS coordinates are quantised to integer values. As you move around, those values change. But the exact point at which the change occurs is fuzzy. It depends on noise (not in the audible sense). So if the phone is stationary, the GPS coordinates can still change.

Engineers often handle jitter by means of a hysteresis function, ie they discount small positional changes, only accepting the new coordinates when they have changed from the previously-accepted ones by more than some threshold value. So if that is being done with GPS coordinates before they get to the watch, then fine, there may be no jitter to worry about. However I do not know if such jitter-removal is done by all devices that receive GPS coordinates. Perhaps someone who knows more about such devices can chip in here. And the following may be of interest:

https://softsolder.com/2009/06/23/gps-position-jitter/
https://softsolder.com/2010/08/15/gps-position-jitter-into-the-drink/

I have never used the Exercise app on the watch but nevertheless I find the watch both counts my steps and tells me how far I have walked. And before you tell me that the distance walked is computed from the step length I supposedly should have entered, consider this:

A moderator posted a link to a web page about how the watch measures stride length. Turns out there is no need to enter a length. It figures it out for itself from actual data (which includes GPS). Quite clever. And indeed necessary because stride length is not a fixed quantity but varies with velocity. So it looks to me like the use of the GPS is not conditional on using the Exercise app on the watch.

So when, during the night, the watch detects steps, It occurred to me that it might actually be noticing GPS location changes, and adjusting the step count based on an assumption of step length. Just a thought. I could be wrong. But I am willing to be wrong. And finding out whether I am or not, will help eliminate one possible source of the ghost steps. So to anyone reading this, please consider the experiment I suggested, I disable GPS on your phone at night, and see if that has any impact on ghost steps.

Sent from Mail for Windows 10

[Nu2_4rum]

“But it is a very hard claim that the code is broken and it is code-bugs that cause this. My take is it is a 'singularity' in the counter algorithm making it susceptible to micro-vibrations.”

So that susceptibility would be a bug, then. In other words the code is broken. Can’t have it both ways. A bug’s a bug for a’ that (to plagiarise Rabbie Burns).

Sent from Mail for Windows 10

[RJayR]

---

@Nu2_4rum wrote:
OK, here’s something else to try:

Suppose the GPS coordinates from your phone fluctuate due to poor signal or atmospheric effects or whatever.

---

When I've seen the Ghost steps, there've been, for instance, 24000 steps for a total distance of 0 miles. So it's registering steps, but not distance. 

[Nu2_4rum]

OK, that means GPS can be ruled out as a factor. It is easy to claim that it can be ruled out without trying it, but it is far better to have the evidence. Thanks for providing some evidence. This narrows down the potential cause.

I am wondering now if we have feedback from people who have looked for ghost steps while charging from an idle PC or a USB power bank. If those do not cause ghost steps, that would be very interesting. On the other hand, if ghost steps still occur, that would be interesting too. It tells us what to look at, or what to ignore, as a cause.

Sent from Mail for Windows 10

[slellis128]

I tried charging from every conceivable place... my computer (think pad), my husband iPad, every wall outlet in my house, wired in USB outlets, multiple hotels in 8 different states. Ghost steps no matter what.

I also got ghost steps with GPS off and on. That made no difference either.

Mine also gave thousands of steps with 0 miles.

I tried 4 units before demanding a refund. Every one of them did the same thing.

[Nu2_4rum]

Thanks. Good evidence.

Question: Do the ghost steps occur only while charging? Or do they occur while out and about too?

Sent from Mail for Windows 10

[slellis128]

All 4 of them counted fine if I was out and about and kept moving, I think...

But the ghost steps happened when I sat still. I watched it climb by over 5000 steps while I sat unmoving in a recliner one night. And the steps added up faster than humanly possible unless maybe you are an Olympic sprinter!

And no fans or any source of vibration by the way.

[Glenmcd]

---

@slellis128 wrote:

All 4 of them counted fine if I was out and about and kept moving, I think...

But the ghost steps happened when I sat still. I watched it climb by over 5000 steps while I sat unmoving in a recliner one night. And the steps added up faster than humanly possible unless maybe you are an Olympic sprinter!

Lets say in this example that step count was 1000 before and 2000 after.  But did you actually see a step count in the 1001 to 1999 range at all? If so, how many different values did you see in this range, and over what time period?  In the digital world, half a second is much, much longer than a nanosecond.  And data corruption typically occurs in < a nanosecond.

[slellis128]

I sat and literally watched the steps go up. The steps jumped up in differing increments. The number of steps ranged from maybe 15 or 20 to 100 or so each time it went up. It went up every few seconds. I remember telling my husband "This stupid thing thinks I walked 1000 steps in the past minute!" It was less than 5 minutes or so when I stopped watching and it had accumulated more than 5000 steps.

One of the times I was time on with support and reported to them the numbers of steps as they increased in real time. I dont know if their developers ever got that info though. They simply sent me another replacement... which also did the same thing. *Sigh*

[Glenmcd]

Interesting.  You mention increments that are all a whole multiple of five (5), but this isn't typical of data corruption either.  More typical values in a similar range would be 127, 128, 255, 1023, 1024 etc.  Numbers that are interesting at the binary level rather than decimal.  You have done well to glean what you have @slellis128