My husband and I are interested in doing this for our daughter's transmitters, does anyone have any extra dead ones sitting around that they are willing to part with? We could paypal you for postage/compensation.
anyone try this yet i have 2 dead ones lying around wondering does it work?
When the transmitter is packaged, it is inside a magnetized case which apparently keeps the battery disconnected from the circuit to maintain long shelf life. Dexcom suspected that I had a malfunctioning transmitter and replaced it under warrantee. Since the battery still had at least 1 month of juice left in it, I packed the OLD one in the case. Several months later I tried using the OLD transmitter and found the battery DEAD.
Anyone know how the Magnet Switch works? Apparently it is a 1 way switch, activating the circuit but not able to turn it off again.
Just a guess here but I think maybe it has just enough pull from the magnet to keep battery from making a connection so the battery life don't start or lack of better words turn it on
Its a fairly strong magnet. Strong enough to pull a large paperclip 1" across a business card. Stronger than many magnets on my fridge.
I'm certain the transmitter is on for testing before it is packaged, just trying to understand how THEY can turn it off but we can't.
I think the unit is actually on but running in low current mode. It just doesn't activate the transmit function when it detects the magnet. They say that it can't be put back into sleep mode so it could simply be a part of the program. Once it wakes up the software won't let it go back to sleep. It's so hard to decap the transmitters that it's unlikely that I'll spend any time reverse engineering the hardware itself. The thought did occur to me that I could design and build my own transmitter that has a small lipo cell inside so we could just recharge it every month, or a little screw like compartment to replace the cells.
I have great difficulty knowing the list cost for a battery is $500+. Electronics should only require replacement d/t failure, not dead batteries.
For what reason - waterproof integrity or profit? I do want to see Dexcom continue to exist as they do have a good product but need to mention that I did delay my purchase for over a year d/t total annual cost.
I would be willing to sacrifice a slight increase in size and initial cost to incorporate 'wireless (inductive) charging', 'motion charging' ie shake to charge flashlight or other means introduced by non winding watches decades ago.
There's a reed switch inside.
http://en.wikipedia.org/wiki/Reed_switch
You will hear a little "click" when you hold the transmitter
right near your ear and approach a magnet over it.
To check the sleep mode, use a multimeter on the two contacts
and measure the output voltage. This should be 0.02 Volts. Now
approach a magnet, wait a minute and measure again.
Ha, you are absolutely correct. That big black thing on the tip of the bottom board is in fact a reed switch. This is useful info in case anyone has a spare sensor they want to put back to sleep.
Very good and very hard to hear. I assume that the 0.02 volts is in sleep mode. On my digital meter I read 0.17 volts on my operational XMTR, not put to sleep.
Next sensor change I need to remember to cut a couple holes in the old packing case and see that if it does indeed drop down to 0.02 which still indicates some potential battery usage.
That still has me considering why the old one was working OK, put to sleep 6 months ago and now is DEAD, but then again any battery self drains/discharges over time even without a parasitic draw.
You mean 0.017V That's normal.
NO: That's why I asked if 0.02 was the sleep reading. I just now cut out the small tab in the plastic case holder to access the sensor contacts on cases from the old XMTR and on a stored, new, never removed from case. The New one reads 0.000 VDC still in the case while the current one reads 0.193 VDC both outside and inside the old case (left in case for 8 minutes). As a rough check, isolated probes read 0.000 and a new AAA Duracell reads 1.566.
What bugs me now is that while disconnected from the sensor, my meter read 41 for 1 measurement period before the hourglass showed up.
Do I need a super strong magnet to put it to sleep? The magnet in the case is as strong as the stronger magnets on my fridge.
I looked at the circut and the reed switch indeed cuts the ground side of the supply so there should be absolutely no draw.
So with my operational XMTR, perhaps I was reading residual charge 5+ minutes after putting it into the magnetic field case. Next sensor change I'll take measurements again using the case including momentary shorting of the contacts to eliminate a potential residual.
After review of reed switch characteristics I recognize that they are NO or NC, changing state only when exposed to the magnetic field, so I would ASSUME that there are no special procedures to put it to sleep.
The reed switch is open when no magnetic field and closed in magnetic field. Therefore it can be only info for controller and software controls the power. One possibility is that when the transmitter is packed to the box with magnet, then a short magnetic pulse, that compensates the magnet, is applied to the transmitter. When the cpu detects this pulse it goes to low power state. This is only my idea but they must have some way to get it in low power mode when packed.
You reminded me that I forgot to retest after eliminating any potential residual charge.
I think you guys are correct and I was wrong on that one. I put a very small ammeter on mine with it hooked up to a lab power supply. With the magnet in place the current did not drop to 0 as I had expected.
It looks like Dexcom has a new slim low profile design on the G4 transmitter, I found this picture on the web...
Rumor has it, that it now has one circut board instead of two...
It may be similar design to the old 7+ transmitter. :-)