transmitter battery drain.

[SJoWie80]

Your Environment

• Uploader Version Number: 0.7.1
• Your Android Phone Model Name (e.g. Samsung J1 Ace): samsung galaxy s8
• Android Version Number (e.g. use 4.4 for 4.4.4): 9
• Network Connection at the time the issue occurred (Wi-Fi, Mobile): wifi

Brief Explanation of Issue

i have the guardian 3 transmitter. after about 4 days the battery of the transmitter is empty.
i allready got a replacement transmitter from medtronic but the problem stays.

i even changed the phone but it still keeps draining the battery.

any tips or tricks?

[Pogman]

Keep the pump and sensor transmitter very very close without them being blocked from each other by the body. Keep the phone away from the pump + transmitter.

The transmitter uses a lot more battery life when it loses signal with the pump and it's a very low power radio signal, it doesn't take much to cause a lost sensor signal situation.

[pheppy]

Your Environment

Uploader Version Number: 0.7.1
Your Android Phone Model Name (e.g. Samsung J1 Ace): Sony Xperia Z
Android Version Number (e.g. use 4.4 for 4.4.4): 8.0
Network Connection at the time the issue occurred (Wi-Fi, Mobile): wifi

I have exactly the same issue. Never had any problem with the previous transmitter. But the battery of my guardian 3 transmitter is also empty after about 4 days. I also got a replacement transmitter from Medtronic, but no improvement.

I know the connection between pump and sensor transmitter is delicate but the connection interruptions are fewer than ever before (hardly any). Could the connection between CNL and pump (at night) interfere with connection between transmitter and pump?

[Pogman]

At night suggests that the pump and sensor are not "hearing" each other and this is often due to a person rolling around in their sleep with the pump ending up on the opposite side of the sensor or pump or sensor being slept on top of and signal being blocked by the body.

Look through the log in the uploader for "pump lost sensor" messages. The amount of time it's in this lost state will have a large effect.

Also a tip: the transmitter charger is poor at determining when full, it's worth doing a double dip and putting it through a charge cycle twice.

FYI the uploader stays out of comms for a considerable period when sensor communication is expected and is unlikely to be an issue here.

[pheppy]

I've checked for "pump lost sensor" messages: none.

There is no indication during any part of the day about communication problems between transmitter and pump. Double and Tripple Dipping is already our standard for many months. So much, we considered the risk that double dipping could cause the battery to get worse.

BTW when the replacement transmitter did not last more than 4 days the first time we called Medtronics. They adviced us to leave the transmitter for 8 hours in the charger (without double dipping). Of course this did not have any impact. I also checked the charger type number, because we could have mixed the gardian 3 charger with the similar looking previous charger.

It's mistic....
I might turn off nightscout for a week to see if it makes a difference. Although we can't really do without.

[Pogman]

Can you enable "extended errors" in the debug menu and look back through the logs again for STATS info. There is one posted each day at midnight and ideally it's from 24/7 use but maybe I can see if there is any other points to look at so post a few days worth here.

The uploader only comms with the pump and only effects the pump battery, there is nothing involved with the transmitter. The only life loss I've ever seen is from signal issues, there is a second signal issue type that goes under "old sgv received", these indicate that no sensor sgv was received at the expected time and the sensor will sent again usually within a minute. Have a look to see if you can see these in the log. While not a lost sensor a lot of these means the transmitter is doing extra work.

[pheppy]

When starting the last sensor I turned off nightschout. Unfortunately not a very hopeful but clear finding. Without nightscout the sensor manages 7 days easily (day 4 - 80% batery). With nightscout we'd hardly got 4 days. I'll make a definite posting in 3 days. But it seems very clear. I'm clueless about the mechanism that causes the sensor to drain while nightscout is running.

[Pogman]

@pheppy it's important to get the information I've requested in the previous post above on stats while the uploader is in use.

[claesg]

We are running the uploader with a poll interval of 10 minutes. That helps the sender to last 7 days. We usually have 20-30% after a full week. When using the default 5 minutes poll interval it did not last a full week.

[Pogman]

@claesg can you try adjusting these timing parameters as shown and run the uploader at the default 5 minute interval and report back? Thanks.

Find them in the "Advanced System Settings" menu.

[claesg]

We just started a new sensor and I have set the parameters according to your instructions. I have also enabled "extended errors" in debug.

[claesg]

The sensor was placed in a "bad place". We get that sometimes and it uses a lot of power and only lasts for a few days. Right now it has been running for 2 days and 7 hours and the transmitter is at 40% and i dont think this time is a good example. Anyways, I got the stats from midnight where we have a full 24 hours. I think it will tell something about the connection between nightscout and the 670G pump.

[Pogman]

It can be difficult to draw any conclusion from one piece of data and without controlling the user setup and environment. But keep them coming and we can see if anythings up.

The CgmOld value here suggests that there were 13 occasions where the pump did not receive a sgv on time and that could be due to sensor-pump being out of range. The CgmLost at 6 means there were 7 recoveries within the 120 second period when the uploader does a recheck.

[claesg]

Here is one more. At SAGE 4d6h and transmitter power at 47% I think this one will do the full 7 days.

[Pogman]

Looking good. Now to see how it does over a few sensor lifetimes.

[claesg]

This time we have SAGE at 2d1h and transmitter power at 53%. It will not make 7 days this time.

[Pogman]

I could have you adjust settings again but I'm weary of "tilting at windmills" with these tweaks.
Where exactly is the sensor placed for this session? For myself I use my lower leg and this causes a fair amount of transmitter battery strain with lost sensor happening several times each night and can have the transmitter battery at 0% over it's life. If it's any help the transmitter battery charge reading is kinda poor and even at 0% I see it run 2 extra days at this value.

Try changing the values as suggested previously from 120 seconds to 150 seconds.

[claesg]

The T1D is a teenager and the sensor switches from left to right side of the body (buttocks). The pump stays on the right side. It has to be like this since the it is very important (for the T1D) in which pocket the phone is...
Right now the pump and sensor are on different sides. During night this does not matter much since the pump is laying loose in the bed.
I have adjusted the values to 150.

Just as a note: This way of placing the pump and sensor seems to work with a pulltime of 10 sec.

[claesg]

Yesterday something very strange happened. In just 6 hours the transmitter battery went from 87% to 27% and refused to take a calibration. We had to replace the sensor. The midnight report is below. I will now go back to 10 min polltime and 120 sec timing parameters to see if we can get back to 7 days runtime.

[nicolaswebb]

I know this is an older issue, but I wanted to share that we had some real challenges with this also (3 separate transmitters, none making it longer than 3-4 days), and it was due to some 2.4 Ghz overlap (which seems to be partly addressed in the comments here). In our case it was a younger kid, and we still had a wireless video/audio camera in the room for monitoring, and removing that solved the problem.