G6 sensor 24 hour break in period

@bkh @Kaelan
I’ve noticed the rewrite as well, even with sensor errors. I thought that I was imagining it!


The G6 phone app filters the preceding 3 data points to smooth the curve using what appears to be a 4 point moving average. This is quite noticable to those of us using Sugarmate or Xdrip, where the data points are not filtered.


To catch them in the act, in the Dexcom app you need to see a trend in the graph, and the single most recent dot shows a change in the trend, such as a reversal from falling to rising. Then, if 5 minutes later the next dot is falls on the previous trend, the trend-reversal dot from 5 minutes earlier will no longer be there in the graph. Instead it will have been replaced by a dot that is on the trend, neatly between the most recent dot and the one from 10 minutes prior. The smoothing to remove outliers may be more extensive than this, but perhaps it may be easiest to find examples of them removing singleton outliers from the tail of the graph.

When I ascribed motivations to Dexcom for doing this smoothing, I just made that up and have no actual knowledge on that point. But the smoothing I have observed, and the difference between the graph in the Dex app and the graph in the LOOP app during periods of jitter is clear.


I have observed this, too, and am learning to be more cautious about taking any action when I see a weird dropping/rising arrow like that.


I guess that brings me full circle back to my first question…the sensor I am on right now started out much jumpier than typical (for me) on Day 1 but smoothed out and seems fine after that first day. But does that excessive jumpiness provide anything to extrapolate to Day 10 or beyond? I sometimes crave a predictable end date for a sensor, mostly to plan ahead for sensor restarts or replacements.

Has anybody correlated Day 1 jumpiness with G6 sensor longevity? Might be too subjective to correlate.


@John58 My currently running sensor displayed excessive jitter for more than 48 hours prior to finally settling down.

Its 10 days finishes in another 6, I’ll let you know how it performs when I restart it this coming Sunday.


Here is my questionably jumpy sensor on Day 8…bouncing along within a tight range. Accuracy is not a problem, hoping we can bounce along to Day 10.


This is Day 6 of my current sensor. Not sure what it’s trying to tell me but the BG has been within 10-15% when I’ve spot checked. It’s been jumpy like this since the start.

Will it last another 36 hours? I sure hope so, I have plans tomorrow night that will go off the rails if my CGM is inoperable.

That does look unstable. I’d make sure it is well glued down, and maybe even use some overtape since you’re trying to reduce the chance possibility of failure for the next 36 hours.

1 Like

I most often see this behavior when a) it is a newly installed sensor . ‘Usually’ settles down and b) when I sleep on my side where sensor is located (compression low) (I don’t start off sleeping on my side where sensor is located but migrate to that side during the night)

1 Like

Yes these sensors often are jumpy like that at the start. This sensor seemed to get worse/ jumpier around Day 5/6. It ended up dying about 12 hours after that picture was taken. Death by sensor error drop outs.

Dexcom is good about sending G6 replacements for sensors that don’t last 10 days. I had an easier time of it with G5 sensors, mainly easier time looking ahead and preplanning sensor restarts or replacements in consideration of my weekly schedule. Have not been able to consistently schedule G6 sensors due to the occasional defective sensor that dies before my preferred date.