When I changed my sensor today (not at expiration), it seemed as if the transmitter was still working as I moved it to the new sensor, and I wondered if I could just skip the warmup. Unlike some here. I don’t seem to get a lot of crazy readings early in the life of a new sensor. Does anyone just skip the warmup at a sensor change and wait until you are forced to do it?
I usually only ever change sensors when the session shuts down, so I’ve never tried it. I don’t see why it couldn’t work. But the thing is, the reason for the warm-up is presumably that there’s enough variability between sensors that the algorithm needs to be recalibrated for each new one. It might not be a problem for a given sensor, but my inclination would be strongly against skipping that step just because I want the thing to have the best shot at being as accurate as it can be. I would be inclined to skip warm-up when I’m rolling over into a new session on the same sensor, but that’s not possible. AFAIK.
If I’m understanding you, you change sensors and instead of “stop sensor” and then “start sensor” commands you just physically move the piggyback transmitter to a new sensor and never tell the Dex system that you’ve done that. I’ve never tried that but that moved transmitter still has a 7-day clock ticking and when it reaches zero, it will need a new “start sensor” command which entails a 2-hour warm-up. Am I missing something? You’ll still need to give it a 2-hour warm-up blackout period and will have not gained any advantage.
Out of curiosity, how long is the Dexcom warm-up when just restarting with the same sensor after the 7 day lifespan? With Medtronic it takes about 15 minutes when restarting “in place”.
The 2 hour warm-up when connecting a transmitter to a new sensor is definitely not optional with Medtronic. The internal logic inside the transmitter detects either when it is no longer connected to a sensor or when connected to a new one. Either way, the transmitter always resets itself for a new sensor and begins a 2 hour initialization whenever it is connected to a sensor. It doesn’t matter whether the sensor is old or new, the Medtronic transmitter will always go through its initialization process as the first thing it does when connected to a sensor.
In the case of the Medtronic CGM, the initialization is a different process from the calibration.
With calibration the device is attempting to derive a function which maps the internal signal from the sensor (ISIG in the case of Medtronic) to the value returned by your BG meter.
During the Medtronic transmitter initialization, the transmitter really does not know yet what the correct value of the internal signal from the sensor is. During initialization it appears to be going through a process similar to when a digital electronic scale “zeros” itself. How it manages to determine what “zero” happens to be I have no idea, but if you track the ISIGs from the sensor/transmitter during the Medtronic initialization, they oscillate up and down by smaller and smaller amounts until they eventually hone in on a nominally “stable” result.
Perhaps Dexcom is completely different. Certainly they are promising to move forward in a direction which will be different in the future if not now. But I expect the G4/G5 are probably doing much the same thing during a two hour “warm-up”. The transmitter is probably figuring out how to reliably read the internal value returned by the sensor so that it can then proceed to do a calibration.
No, Terry, you’re right, I’m not gaining much, just delaying the warmup until I’m forced to take it.
I regularly “skip” the Dexcom warmup period… sorta. I keep both, the Vibe receiver and my Dexcom G4/Share receiver calibrated - usually carry both, but not always. Rather than stop/start them both at the same time, I have them staggered - with each one having 7-day sessions that start/end on a different day. Since the Vibe uses the older algorithm, it tends to get a bit more “confused” when I change sensors, but don’t “tell” it, so I schedule actual sensor changes on the Vibe’s schedule. The G4/Share receiver sometimes doesn’t notice a sensor change at all - or at worst, goes to ??? for up to an hour before going on its merry way. Accuracy does not appear at all impaired by my (bad) habit.
@irrational_John, the Dexcom goes through the same two-hour warmup at “expiration” whether you actually deploy a new sensor or keep the old one. In changing a sensor between expirations, it doesn’t seem to know it has been moved from the old sensor to the new one, perhaps based on how quickly the move is made.
It’s the same 2-hour interval, because as far as Dexcom is concerned you come to the end of the session, you put in a new sensor. We almost all of us “cheat” of course, but Dexcom doesn’t officially know about that. But a question back to you: [quote=“irrational_John, post:4, topic:50819”]
With Medtronic it takes about 15 minutes when restarting “in place”.
[/quote]
Sounds like Medtronic does officially make provision for rolling over to a new session with the same sensor then? And if so, what’s the 15 minutes for?
On edit: skip that last question. You’re talking about the difference between “calibration” and “initialization” right? With Dex there’s just the one process.
Oh we know about your devious Dexcom-deceiving ways, @Thas, you wily scalawag. I’m hoping to do the same when I’m eligible for a new Dexcom receiver, but that’s not for another year and change.
no, i always do the warm up,.
When you run the pump and the receiver do you keep one without alarms? I ran both for about two weeks and now only use the receiver because I like how I can push one button and its cleared.
I sometimes reboot the same sensor and it bypasses the warm-up - or the warm-up is short. I have never inserted a new sensor and it not require the usual warm up time. I noticed the new Dexcom sensors (past yr) don’t stay on as long so I am more likely to change it at the end of the 7d period. I actually lost my transmitter 2x - once was recently - fell right off! Tracked it down at my Dr.'s office. Thankful because my new insurance charges me a bunch for a new transmitter! Forgot to say…I use it w/ Omnipod…so less possibilities to “trick” it, it seems.
Worse than that - I also send the data via Share to my phone and watch! I occasionally leave the receiver behind, so I do have alarms on both Vibe and receiver; however, the Vibe alarms are so quiet and the vibration is not terribly annoying to me. I agree that the receiver, being easier to clear, is preferable.
Recently, I’ve decided to put somewhat more aggressive limits on the alerts from my watch, so that I can react to numbers that are heading out of range earlier. I leave the receiver on the wider range, so that I don’t have to remember to change the settings at night – sleep is less likely to be interrupted that way unnecessarily 
It’s not actually enforcing a 15 minute warm-up; rather, MedT waits for the next 5 minute-interval to prompt for a calibration, then waits 5 more minutes to internally “synchronize” your calibration with the sensor reading, then finally a third 5-minute period to display a sensor glucose reading.
Performing a calibration immediately BEFORE stopping/staring a sensor circumvents this problem and you don’t lose data – just bad sure to do this before forced via a “sensor end” message
I am not sure what you are referring to by “lose data” so I am unsure whether it does or does not match my own experience.
What I think I know is that whenever you do a “regular” calibration, the Medtronic (pump) will always attempt to incorporate/extrapolate the new calibration factor with the previous calibrations. This process takes about 15-20 minutes to occur depending on where the pump is in its “every 5-minutes” synchronization with the transmitter because it takes at least 3 of the 5-minute periods.
I believe the only two exceptions to this are when you use either Link To Sensor > Reconnect Old Sensor or Link To Sensor > New Sensor. Both of these operations appear to do the same thing as far as calibration goes. That is, they start a “15 minute” calibration process which ignores any previous calibrations. The difference between them seems to only be whether or not the Sensor Age timer is reset to 0 or not — not surprisingly, Reconnect does not reset and New does.
The only connection I have seen between losing previous calibration data and the Sensor End alert/alarm is that the pump always forces one to use Link To Sensor > New Sensor after the end. But one always starts from scratch, ignoring any previous calibrations, whenever you use Link To Sensor > New Sensor. That’s just how it works.
Sometimes this is an advantage. Often my sensor performance changes drastically during the first 12 to 24 hours. If this happens then using Link To Sensor > Reconnect Old Sensor to calibrate essentially instructs the pump to ignore the previous calibrations and just work with the BG and ISIG it is given for that calibration. This immediately “snaps” the Sensor Glucose (SG) to the Blood Glucose (BG) you enter for the calibration rather than waiting for the pump to gradually grind it’s way to what appears to be a new ISIG equilibrium point.
What I mean is that you don’t get the gap in data. If you calibrate the “old” sensor before selecting START NEW SENSOR, that calibration is applied to the first measurement of the new sensor. Otherwise, it takes a sensor reading 1-5 minutes later, realizes you’ve never calibrated the new sensor and you get a METER BG NOW message. Then you need to wait even longer before a glucose reading appears.
I think I am seriously misunderstanding what you might be saying. Are you talking about actually moving the transmitter from an existing sensor to a newly inserted, new sensor? Or just about doing Link To Sensor > New Sensor for an existing sensor/transmitter pair where the transmitter has already completed the 2 hour initialization process?
If you are saying that there is a way to perform Link To Sensor > New Sensor without a gap in readings, then I disagree. There is always a gap because using the New Sensor process instructs the pump to ignore any previous calibrations.
Existing sensor. Yes, starting a new sensor should discard previous calibrations. However, when you calibrate, the value is held in a queue until it can be correlated to an ISIG reading (every 5 minutes). While previous calibrations are purged, as you stated, the queue is not.
That’s my experience with Revel/Sofsensor and 530G/Enlite anyway. I’m no longer using the MedT sensor, so my recollection of the actual menu items is a bit hazy. But I’m confident that it worked.
Oooooooooooooooooooooooooh! Now I think I see where we were not connecting.
I was thinking you were saying to perform a calibration, wait for it to complete, and then perform New Sensor. However, what I think you are saying is to perform a calibration and then perform New Sensor before the previous calibration has completed.
It never, ever occurred to me to try something like that. To my midwestern Wisconsin raised mind it sounds somewhat, well, inappropriate. But, what the heck, maybe it’s time to try living on the wild side, eh? Ya and hoo. 
Well, I tried out what I think was your suggestion using Link To Sensor > Reconnect Old Sensor. I didn’t use New Sensor because I didn’t want to reset the Sensor Age at the moment. (I’m easily confused. 
)
What I tried did not eliminate the “Warm-Up” gap. It did seem to reduce the gap from 15 minutes (3 ISIGs) down to only 10 minutes (2 ISIGs). It also eliminated the meter BG NOW alert/alarm. But there was still a gap, probably because the pump needs at least 3 ISIG values to generate its “first” calibration factor. 
For whatever it’s worth … which admittedly probably ain’t much … below is the play-by-play as logged in the CareLink Data Table.
- 20:06:00 143 [27.12nA]
- 20:11:00 142 [27.4nA]
- 20:13:00 Calibration BG: 140 mg/dL (Linked)
- 20:14:00 Sensor Start: RECONNECT OLD
- 20:16:00 — [27.42nA] Waiting for calibration effect
- 20:21:00 — [27.26nA] Waiting for calibration effect
- 20:26:00 140 [27nA] Calibration Factor: 5.154