Some questions on restarting a G6 sensor

I’m going to try my first sensor restart soon on my 8L**** transmitter. I have a few questions I wanted to clear up before I attempt it. I’ve gone through the various threads on restarting a sensor and I understand there are various ways to restart a sensor, what worked initially stopped working over time. I gather the removing the transmitter method is the most up to date and attempts have been made to restart the 8L***** successfully.

  1. Stop sensor
  2. Remove transmitter
  3. Wait 30 - 60 mins
  4. Insert transmitter
  5. Start sensor using original sensor QR code

On step 1, do you wait for it to expire automatically then stop or stop it before it dies?
I’m assuming all the steps above are done using the dexcom app. What do I need to do on the pump? Do I need to stop and start the sensor on there as well? Edit - I’m on the TSlim pump

I don’t need to wait for it to expire, nor do I need to stop it. Either way works.
But I have an Omnipod pump.

But I don’t have it linked to a pump. You might say which pump you have and someone will be able to help you specifically with that.

1 Like

On the Tslim

All sensor functions should be done on the pump first and just let the Dexcom app catch up.

All of the steps you listed will work on current transmitters and sensors except I haven’t ever waited an entire hour to restart the sensor. I don’t restart sensors regularly but did last week since there was a mix up with my shipment and I didn’t want to be down to my last sensor. I removed the transmitter and waited 15 minutes (timed it on my phone) then reinserted the transmitter and started it up as usual. You can manually stop the sensor or wait for it to expire on it’s own.

There may be a big difference from the new reading versus a finger stick so in my experience I will calibrate it in sections of no more than 35mg/dl or about 2mmol/l at a time. This is important or you will get calibration errors and those are annoying. Good luck!

1 Like

Ah so I got it the other way around. It should be:

  1. Stop sensor on pump.
  2. Stop sensor on phone
  3. Remove transmitter
  4. Wait 30 - 60 mins
  5. Insert transmitter
  6. Start sensor on pump using original sensor code
  7. Start sensor on phone using original sensor QR code

Is that right?

Nope. :grinning:

If honestly doesn’t matter which device you use first. There are people who do it either way. Personally, I chose to do all starts on the pump, just so there’s a record of it, since I use a 3rd party app, but it doesn’t matter. For Tandem users, I would suggest using the pump for starts just because the “scan the 4 digit code with your camera instead of manually entering it” feature of the app is glitchy. You run the risk of being forced to calibrate twice a day (and potentially miss pump intervention when the sensor is waiting for the calibration) if you use this glitchy feature.

One of the things that DOES matter, though, is that you only enter the startup code into the first device, then choose “no code” for the second device.

Also, there is zero need to wait 60 minutes. People keep “padding” the time, just to be sure. And then the next person who follows their instructions also pads the time, and it’s compounding in their repeated instructions. You only need 15 minutes, but the timing is important. 14 minutes and 59 seconds may not work for you, and it won’t work if you stay timing before you remove the transmitter. Both of which are common mistakes.

There should also be a step where you clean the contacts (the 2 metal polka dots) of both sensor and transmitter. Any gunk that’s accumulated in there can cause sensor errors. I start the timer after I do this very quick step, because it pads the timer for a few seconds to ensure I wait more than the minimum 15 minutes.

  1. Stop sensor on both devices, or wait for the session to expire naturally on both devices.

  2. Remove transmitter using a tear-drop shaped hair clip (the ones you bend to snap shut), guitar pick, or test strip.

  3. Clean the metal contacts on both the sensor and transmitter with an alcohol pad.

  4. Set timer for 15 minutes.

  5. After the timer expires, and not a moment before, reinsert the transmitter and start the sensor on your first device. (Starting on the T:slim X2 is my personal preference)

  6. Enter the startup code. Those of us that do restarts will usually take a photo of the startup code to reference here. I’ve also started taking a picture of the lot number and expiration on the back of the plastic packaging in case there’s a problem and tech support asks for this. You can also choose “no code”, but you will be required to enter 2 calibrations immediately after warmup and an additional 2 calibrations every day for the rest of this session. The benefit of “no code” is that it avoids the hassle of step 8…

  7. Start the session on your second device, if using, using the “no code” option. It will join the session already running.

  8. After the 2 hour warmup, your first reading WILL be high. Maybe VERY high. Anyone using Control-IQ, or other automated system, should turn it off, so the pump doesn’t deliver a false correction bolus. You will have to calibrate, but do so carefully. BIG jumps may lead to rejected calibrations. I try to use a “rule of 40”. You don’t want to calibrate more than 40mg/dl (2.2 mmol) or 40% of the false high G6 reaching, whichever one is SMALLER. In this instance, it’s nearly always going to be the 40mg/dl (2.2 mmol). Note: the error seems to be less when my BG is close to my lower range limit

  9. If you have to more than 1 calibration because the difference was more than 40mg/dl (2.2 mmol), wait at least 15 minutes (3 data dots) to do the next calibration.


Thanks for the details that was very helpful!:pray:t2:

Regarding this comment, I’m curious whether it makes a difference. When I started my first sensor session last week, the DE got me to enter in the code into my pump and then scan the code in for my phone. This means I must have chosen the use code option.

Is 15 mins generally the time I need to maintain stability in my BG after calibration? I ask because I want to know how long after a calibration I need to wait before I can give myself an adjustment bolus without screwing up the algorithm .

1 Like

You know how it’s been said not to calibrate in the first 24 hours of starting a new sensor. Does it not apply to restarts as well?

Most definitely NOT.

The first day wonkiness has more to do with your body’s reaction to the trauma of the insertion than anything wrong with the programming. You will have to calibrate it, and there’s zero reason to wait. You just have to do so gently.

As to entering the startup code twice… Admittedly, I’ve entered it in both pump and app, too, with no dire consequences. It’s come up quite a few times with Tandem’s tech support, though. Somewhere in their training, they’re taught that it does matter, though I personally have no idea why. Since there’s so much debate about what causes the system to misbehave, I figure it’s just worthwhile to pay attention to this as a possible contributing factor. :woman_shrugging:

You can’t really screw up the algorithm with insulin delivered AFTER a calibration. But if you require multiple calibrations (my first reading can easily come in 100mg/dl higher than reality, that’s too big of a difference and would get rejected if I did it all in one calibration), you CAN screw up the next calibration you’re waiting to do.

In a perfect world, you shouldn’t require a correction while you’re trying to get your newly restarted sensor calibrated back in line. I’ve developed a weird quirk of doing more fingersticks in my two warmup hours without data than I used to do in a single day. I strive REALLY hard to be between 80-100 mg/dl (4.4-5.5 mmol) when the first data point registers… Soley because it makes calibration easier. So my corrections have been done and stabilized before I get to this point.

In an imperfect world where you need the correction bolus… If you’re high, I think that correction is more important than getting your Dexcom pretty and aligned. Go ahead and bolus. Technically the calibration rules still apply (in range and steady), do it’s best to just ignore the data until you can get to a point where you meet the calibration rules. I’ve been known to break them on occasion, though, because I’m impatient. Just know that if you break the rules, your information will be off and require more calibrations later to bully it back in line.

Honestly, it’s all this first day of a restart calibration fussiness that has taught me the right behavior for calibrating and not to fear it. Those who aren’t restarting and haven’t been through this many calibrations just don’t have the experience with it to know the little innuendos. It sounds complicated to write it all out, but the reality is that it becomes intuitive quickly.

It IS intimidating and annoying, though, especially at first. You kind of have to decide which is the better annoyance for you. Would you rather have a BIG annoyance that only lasts a short time before the dex falls back in line, or would you do better with a small but contestant annoyance? If the latter is your preference, then choose the “no code” start up and do the mandatory 2 calibrations a day. The flat and steady calibration rules still apply. Personally, I think it’s easier to wrestle it back in line at the beginning and not have to worry about it again.


You don’t need to do anything on the app. It will catch up and start receiving data on it’s own.

The only time I’ve had to do anything on the app since getting the tslim is when I started a new transmitter. It needed the new transmitter information so I went through the new sensor session and entered the transmitter information and it started giving readings immediately since the pump was already through the warm-up.


Thanks. This is good to know. In my last calibration, I moved dinner out by an hour fearing that if I bolused too close shortly after my calibration, it will screw the algorithm up.

Interesting, from this it seems the transmitter is automatically able to sync up the information you entered into the pump with the phone. I wonder why it doesn’t relay the transmitter ID across as well.

Is it common for the sensor reading to be out on the last day? I’ve got less than 24 hours left on this sensor. A few hours ago it was still relatively accurate. And all of a sudden, the readings went off (about a 4.2mmol/75mg gap).

One thing that happened was my sugars were stubborn all afternoon so it has been sitting around the 11 mmol/196mg mark for about 3 hours. While this is out of range, its not terribly high. But not sure whether this is the cause of the problem because the inaccuracy happened after this 3 hours when my sugars were starting to come down. Sensor reading was lower than my meter reading so it’s definitely not a projection of 20 mins later.

I can’t say I’ve noticed this, no.

The Dexcom algorithm sort of smooths everything out, though. It uses the previous two values to help figure out where to plot the next graph point. If your sugar is changing rapidly, the graph much be a little off, but catch up in a few plot points.

Most commonly, inaccuracies are just due to dehydration. Could you have forgotten to drink your water? You’ve got to be hydrated the you to have movement in the interstitial fluid the Dexcom measures.

It’s possible you could be nearing the end of your sensor life, too. I do find i need more calibrations in the end to bully it into line. I did my 2nd restart yesterday, putting me at day 21, and now I’m in the phase where I’ll fingerstick at least once a day just to check if it’s drifting. Until you get those post-prandial figured out, you won’t get the crazy long sensor life. It’s still worth trying the restart, though.

You’re still doing fingersticks anyway, right? I’d go ahead and just do the “no code” option for this first sensor at least. It will require two calibrations a day, but it won’t fight you on them. And then you’ll be able to really see how long your sensor will last, Sir we know if end-of-life is an issue.

1 Like

I don’t think I was dehydrated at the time, I’m been making a conscious effort to drink more water. But maybe still not enough, I don’t know. :woman_shrugging: Just seem kind of weird that all of a sudden it was out so much.

But the sensor reading ending up realigning again a few hours later. It wasn’t exact, but close enough. I’m restarting today so we will see how it goes. Thanks.

Entering transmitter id on your pump or app or receiver is like giving it the password to talk to YOUR transmitter, but not mine!!

1 Like

Just an update and I restarted my sensor yesterday and it restarted after the 2 hour warm up. The initial reading that came up was not too far off about 2mmol (36 points). But then during my meal when my blood sugar started to rise, the rising sugar had more discrepancy about 85-95 points off.

I calibrated once in the middle of the night when it was about 45 points off. It aligned the numbers however this morning when I woke up, its been about 20-36 points higher than my meter. :roll_eyes: I guess its still considered “within range” but not ideal. I’m going to see how it behaves when I eat again. Hopefully it will sort itself out by then. Still in the first 24 hour window.

I’ve developed that weird quirk of more fingersticks in that 2 hours than in a whole day too lol…

I also strive to be around 80-100 when it restarts because it likes to start so high at first.


@tedos, you’re a few hours ahead of me, as I also engage in the 1st sensor everrrr restart. Assuming that you followed Robyn_H well listed 9) step procedure, and all went well. Any last minute precautions?
I am very unclear as what to do with the 4 digits sensor code that I never did enter for this first sensor install, 9.80 days ago.

Typically you should always enter the code on first use of sensor. You should be prompted for it during start, or you choose no code.

For restart, I usually choose no code, since in my opinion the sensor wire is “worn down”, so I want to be prompted for calibrations to keep it more accurate.
Others just use same code on restart.

1 Like

@MM1 thank-you.
As we strive to get maximum mileage on a 10day product, exceeding manufacturer’s recommended design guidelines(!!), is it now akin to drinking milk that has exceeded the best before date? > Might taste OK for 12 hours, maybe 48hrs… …but that milk is gonna turn!! :smiley:
Will the sensor keep sensing, and transmitting it’s data until the numbers are so out of whack with the finger pick BG calibration, that it is time for that sensor to be changed?