• Please Remember: Members are only permitted to share their own experiences. Members are not qualified to give medical advice. Additionally, everyone manages their health differently. Please be respectful of other people's opinions about their own diabetes management.
  • We seem to be having technical difficulties with new user accounts. If you are trying to register please check your Spam or Junk folder for your confirmation email. If you still haven't received a confirmation email, please reach out to our support inbox: support.forum@diabetes.org.uk

Jugglico is back in the Google app store

Interesting with the update on Jugg? Changed sensor today & as the new one started (the old one still inbedded.) had 2 sets of readings on the app? Had to look up the new one’s serial number on librelink & forget the old one. Which after a finger prick was out (lower by 1mmol. In tune with the new one.) Even my watch had 2 sensors. (Galaxy 4.) Not noticed before. But it is working a treat.
 
You can click on Info and see when a sensor was started (and when it should expire) to see which one to terminate.

It's supported two sensors ever since I started using it.
lol, cheers. Oddly, I’ve just figured out a way to use an analog face watch so the hands don’t obscure the BGs? (At certain times.) Messing about with Juggluco complications and the watch app instead of “G-watch app” for sensor readings? (Posted elsewhere on the before supper topic?)

You_Doodle_2024-12-29T19_00_30Z.jpeg
 
lol, cheers. Oddly, I’ve just figured out a way to use an analog face watch so the hands don’t obscure the BGs? (At certain times.) Messing about with Juggluco complications and the watch app instead of “G-watch app” for sensor readings? (Posted elsewhere on the before supper topic?)

View attachment 33258
For the moment I'm using the sample watchface the developer offers. (It's WearOS 5 compatible, I think.)
Not sure I like it that much since it is a bit confused with the arrow and watch hands. Shame something changed with vibration intensity meaning I need the alarms to make a noise too now. I'm hoping it's either a bug that Google will fix or just a change in API which can be fixed in the app.
 
For the moment I'm using the sample watchface the developer offers. (It's WearOS 5 compatible, I think.)
Not sure I like it that much since it is a bit confused with the arrow and watch hands. Shame something changed with vibration intensity meaning I need the alarms to make a noise too now. I'm hoping it's either a bug that Google will fix or just a change in API which can be fixed in the app.
I have a recent update UI6 on a Galaxy 4. (Not sure if that is a OS thing or just to pretty up the “experience?” I see a couple of changes.) I can mix & match the arrow & BGs on the complications, but have set to what you see above. On my Galaxy the intensity of the vibrations have 2 settings? I’m not sure how loose you like your strap? But I was woken this morning by the watch before the phone chimed in with the rattle on what was a low. My watch seems to alert a fraction before the phone. Yet my phone does rattle first before it’s alarm tone.
 
On my Galaxy the intensity of the vibrations have 2 settings?
Nothing on the Pixel Watch. I managed to work out how to build the WearOS variant of Juggluco and saw some warnings of some deprecated vibration related APIs so I'm guessing it's some change in API, maybe not new but maybe it's only in a recent update that is actually broken this.
 
Nothing on the Pixel Watch. I managed to work out how to build the WearOS variant of Juggluco and saw some warnings of some deprecated vibration related APIs so I'm guessing it's some change in API, maybe not new but maybe it's only in a recent update that is actually broken this.
Sorry to hear that. Any chance of a “roll back? With your watch” All I know about the Galaxy watch is there can be a factory reset. Bit of a pain. I hate it when these upgrades cause niggles at best & worse make it unusable for the user. “Built in (or engineered) obsolescence” flies in the face of not only hard earned cash, but also the environmental impact, too?
 
I don't think I can roll it back. Not sure I care enough to try. It's (to me) more an irritation than a showstopper. And maybe it'll get fixed somehow.
 
Hi folks,

I'm thinking of doing an experiment to compare the Libre 2 with the Dexcom One+ while at the same time comparing the official apps with some 3rd party apps. The idea is to wear both sensors at the same time and do a mess of finger-prick readings with several meters testing the same drop of blood after meals, then create plots of the readings to compare the CGMs and the calibration options. The plot below illustrates the kind of thing I'm thinking of - that's Libre 2 data exported from the LibreView site plotted alongside readings from four meters.

If I were to install Juggluco and xDrip, activate the Libre 2 with the official app, connect the sensor to Juggluco, link Juggluco to xDrip and calibrate that data, would that allow me to get the LibreView report and CSV data, 1 minute and 15 minute 'raw' readings from Juggluco and also calibrated readings from xDrip? I assume that if I connected the sensor directly to xDrip I would get much the same thing but would prevent me from getting the 'official' data export and reports from LibreView? I'm trying to figure out the best sensor and app combination to get a good idea of how high I spike after meals. As you can see from the plot the uncalibrated Libre 2 doesn't appear to be great for this purpose.
Dinner_Dec8th2024_Libre 2_4_meter plot.png
 
Hi folks,

I'm thinking of doing an experiment to compare the Libre 2 with the Dexcom One+ while at the same time comparing the official apps with some 3rd party apps. The idea is to wear both sensors at the same time and do a mess of finger-prick readings with several meters testing the same drop of blood after meals, then create plots of the readings to compare the CGMs and the calibration options. The plot below illustrates the kind of thing I'm thinking of - that's Libre 2 data exported from the LibreView site plotted alongside readings from four meters.

If I were to install Juggluco and xDrip, activate the Libre 2 with the official app, connect the sensor to Juggluco, link Juggluco to xDrip and calibrate that data, would that allow me to get the LibreView report and CSV data, 1 minute and 15 minute 'raw' readings from Juggluco and also calibrated readings from xDrip? I assume that if I connected the sensor directly to xDrip I would get much the same thing but would prevent me from getting the 'official' data export and reports from LibreView? I'm trying to figure out the best sensor and app combination to get a good idea of how high I spike after meals. As you can see from the plot the uncalibrated Libre 2 doesn't appear to be great for this purpose.
View attachment 33454
Hello, just picking through your post regarding what your goal is. Personally, I’ve never had any major discrepancies with my Libres & meter? However you should see data on your Libreview account if you start the sensor with the app. From memory, it updates to Libreview every 20minutes. But I haven’t logged into the account in a long time. If you are calibrating any third party app? I would recomend doing it whilst you have been on a level read out for at least an hour, in range. Not while BGs are changing. There can be a bit of lag with what your blood says & the interstitial fluid with rapid changes. Especially with Juggluco reading every minute. (Just to add in.) depending on your phone? Be carful of adding xDrip into the mix. I’ve known it to be quite “hungry” as an app, causing BT drop outs resulting in multiple restarts & connections? Wishing you well with what you are doing.
 
Last edited:
Hello, just picking through your post regarding what your goal is. Personally, I’ve never had any major discrepancies with my Libres & meter? However you should see data on your Libreview account if you start the sensor with the app. From memory, it updates to Libreview every 20minutes. But I haven’t logged into the account in a long time. If you are calibrating any third party app? I would recomend doing it whilst you have been on a level read out for at least an hour, in range. Not while BGs are changing. There can be a bit of lag with what your blood says & the interstitial fluid with rapid changes. Especially with Juggluco reading every minute. (Just to add in.) depending on your phone? Be carful of adding xDrip into the mix. I’ve known it to be quite “hungry” as an app, causing BT drop outs resulting in multiple restarts & connections? Wishing you well with what you are doing.
Many thanks. For me almost all the Libre2s have read a bit high and it seems to vary a bit from sensor to sensor, some reading higher than others, compared to the Contour Next. The Libre seems to read higher compared to that meter at higher BG concentrations. Aside from trying to solve that problem I'm also trying to find the best/most affordable meter or meters to compare/calibrate the CGM data with. Based on this study and two others - Link - The Contour Next appears to be about the best around but also has expensive test strips. That meter is 'System 6' in Fig 2, where readings are compared to two different lab-grade testing machines as reference points. I've been testing with several meters at a time, testing the same drop of blood, using the Contour Next as a reference point. The plots below illustrate that some meters tend to read higher or lower on average at a given BG concentration, so depending on the meter the Libre 2 may appear to compare well or not so well. It may also have just been unlucky with the sensors I've used or perhaps some factor such as the depth of the layer of fat under the skin where I placed the sensor is causing the automatic calibration to read high. I'm not sure. I've just had my meds reduced for a second time and they're likely to be reduced again, and I'm trying to find a better way to analyse things like changes in meds than waiting months for a HbA1c test and being unable to distinguish between changes in body weight, diet, exercise changes and meds.
C Next above 8 cutoff at 13.pngC Next Difference absolute above 8.png
 
Many thanks. For me almost all the Libre2s have read a bit high and it seems to vary a bit from sensor to sensor, some reading higher than others, compared to the Contour Next. The Libre seems to read higher compared to that meter at higher BG concentrations. Aside from trying to solve that problem I'm also trying to find the best/most affordable meter or meters to compare/calibrate the CGM data with. Based on this study and two others - Link - The Contour Next appears to be about the best around but also has expensive test strips. That meter is 'System 6' in Fig 2, where readings are compared to two different lab-grade testing machines as reference points. I've been testing with several meters at a time, testing the same drop of blood, using the Contour Next as a reference point. The plots below illustrate that some meters tend to read higher or lower on average at a given BG concentration, so depending on the meter the Libre 2 may appear to compare well or not so well. It may also have just been unlucky with the sensors I've used or perhaps some factor such as the depth of the layer of fat under the skin where I placed the sensor is causing the automatic calibration to read high. I'm not sure. I've just had my meds reduced for a second time and they're likely to be reduced again, and I'm trying to find a better way to analyse things like changes in meds than waiting months for a HbA1c test and being unable to distinguish between changes in body weight, diet, exercise changes and meds.
View attachment 33455View attachment 33456
What meds are you taking. Are any a form of “diuretic?” Are you suitably hydrated? (Which from what I’ve read, could affect the concentration in interstitial fluid?)
 
What meds are you taking. Are any a form of “diuretic?” Are you suitably hydrated? (Which from what I’ve read, could affect the concentration in interstitial fluid?)
I'm on no meds that have a diuretic effect though I believe have observed inaccurate readings from the Libre 2 when I've woken up dehydrated (after too much alcohol the night before). Generally though I'm well hydrated.

I'm seeing two problems with accuracy with the Libre 2. The first is calibration - how accurate it is at my fasting level, which seems to vary from sensor to sensor for me and sometimes from day to day, especially in the day or two after the sensor is applied. The second is the bias at higher BG concentrations. It's known that the Libre sensors read higher as BG concentrations rise. The image below is from some study on Libre sensor accuracy. I haven't found the study itself yet - I got the image off the FreestyleLibre forum on Reddit. The smaller dotted line represents the sensor bias while the thicker dotted line at a 45 degree angle represents objective accuracy. The two lines cross over each other at around 4 mmol/L so that's the point at which the sensor should be closest to objective accuracy (though generally not in my case it seems). At around the 8 mmol/L mark the bias starts to become significant enough to cause problems for me. I'm trying to find some way to 'post-calibrate' the sensor data, if that might be possible, to get a reasonably good idea of how high my BG goes after meals. If this bias effect is consistent from sensor to sensor and if I can work out some kind of formula to adjust the data, perhaps I could figure out a way to use a record of finger-prick readings to calibrate the data and after eating the same meal many times get a better idea of how high my BG typically goes on average after eating that meal. As things stand I'm having to do a fair bit of guesswork every time I look at a graph after eating to try to figure out how high my BG might have actually gone.
Bias Diagram.png
 
Back
Top