Yeah, that’s a different issue. Just one more! 
FWIW, the TI MCU that’s used by Wyze in the contact sensors is interesting enough to me that I ordered a development kit from TI today. The CC1310 and related products look like they have a lot of potential for a wide range of sensing applications. Link below if you’re also interested.
3 Likes
FYI, received two replacement contact sensors today, two more are on the way. Spent considerable time before they arrived trying to get the last contact sensor that failed working again. As an experiment to see if the bridge was permitting new devices to be added, I deleted my motion sensor and added it back in. No problem, worked perfectly. Tried my contact sensor that failed, but still showed a flash on battery install, and 3 flashes on holding in the reset button. No matter what I tried, the app would timeout without ever detecting the sensor.
Got the new sensors, they have a different PCB datecode of 2019 than my other ones, which were 2020. Both sensors were quickly detected when adding them to the app, and both function as they should.
So, if you think that it’s you’re doing something wrong, that’s probably not true. These contact sensors definitely have some issues, either with going completely dead, or ceasing to be visible to the bridge. As there are hardly any components in these devices, it’s either a hardware issue with a specific revision of the TI CC1310 MCU, or a firmware issue. Hope someone at Wyze is listening and working on the root cause.
3 Likes
I was reading a bit in TI site, apparently you can do a reset written into firmware. May be Wyze flash them when they get the returns back.
Wyze should have a pin to reset the lockout function.
2 Likes
it is also good practice to ensure that a power cycling operation, such as a battery replacement, triggers a Power-on-reset by ensuring that the VDDS decoupling network is fully depleted before applying supply voltage again (for example, inserting new batteries).
Interesting, I wonder if this is why the new battery replacement instructions recommend waiting 20 seconds after removing the old battery, to let the caps in the decoupling network discharge?
1 Like
Could very well be. I think it would take a lot longer than 20 secs for natural discharge to occur in a capacitor. I left one of my dead sensors battery-less for a week thinking it might reset, but nope. Another will have been sitting un-powered for 21 days tomorrow. Going to give it another shot. If that doesn’t work… time to look at the schematics to force a hard (jumped pin) reset.
2 Likes
Curious to know what you find out. Also wondering how that development kit worked out. 
I did find that they have been shipping the Rev B die (referenced in the above caveat) since 2017, so I doubt that’s the issue. There are several other advisories for the chip here.
1 Like
I’ll post what happens tomorrow afternoon or evening. I didn’t order the kit. I think that was the OP (@pbanders ). I did read the Rev identification/production/distribution sheets and saw that Rev A isn’t likely in use here. I’ll dissect one tomorrow to verify the chip.
1 Like
Status update from me. I’ve replaced 3 of the 4 failed sensors, 4th sensor is here and will replace it later when I get back to my other house. I now have three of the Ti development kits. You need at least two to do TX/RX testing. That CC1310 chip is pretty amazing. It has a sub-1 GHz radio, which has range, power, and interference advantages over the commonly-used 2.4 GHz frequency. In long-range mode (low bitrate) in free air, the range is between 10 km to 20 km, pretty impressive. You can also use Energia as the IDE and load standard Arduino libraries and sketches to it. Lots of add-on boards, too. I’m hoping to spend some time doing development with it later this year.
One easy hacking project I’m going to do is to turn a Wyze contact sensor into a leak detector. Just remove the reed switch, add leads, and connect a leak detection sensor. Same is true of most open/closed sensors.
1 Like
I do not have any contact sensors. so I do not kniw which package Style Wyze sensor is using.
In RSM package the decouple cap pin is
- wonder if short to gnd will clear. make sure it is the correct package. the pin layout is different in every single packaging style.
there are tital of 3 style.
The spec say reset BOP onky when vdds are depleted.
another try is grounding these vdds pins.
itis just me speculate. it might damage device.
to be safe. may use a 1k ohm resistor to ground.
The thing is, with mine at least the contact sensor comes back, it has just lost its configuration - the MAC is gone, and it’s replaced with NULLs. Hard to say whether a reset would fix this or if somehow it got completely erased.
I think it is a big win if the come bk … config is just matter of setting.
Mac address missing is a concern. probably a firmware push may help.
FWIW, I’ve tried shunting the open battery terminals for several minutes, it didn’t get the units to come back. You’d be surprised how long after removing the battery that you still read voltage in the mV range on the terminals - try days and days. It’s not clear how to access the reset pin to see if the brownout protection is the problem - and, at $5 apiece, not really worth the required effort. Lastly, it’s not clear how Wyze flashed the firmware to the CC1310, or if it’s pin accessible after board assembly. The CC1310 does support over-the-air firmware updates, but it’s unclear how to accomplish this with the Wyze Sensor if it’s possible at all. Losing the MAC but retaining functionality is puzzling. I assume Wyze flashes every CC1310 with the base firmware, and then though some initialization process flashes in the MAC, so it might be possible for Wyze to upload a new MAC.
1 Like
It is not the battery terminals plus to neg. Since it is using coin battery, design must use dc/dc feature and a 22uf cap.
Try grounding the decouple pin and vdds pins. then battery. do a reset( I think). I am guessing here.
though it is a 5$ piece. returning is just a hassle
pin 19 is vdds dc/dc.
just saw your pic . may try shorting all caps with a resistor to see if it will clear. especially that big grey one at tge lower corner.
True, regarding the battery. I just shorted what was easily accessible. It’ll be a while before I can try doing pin level work, the parts are at one house and my electronics setup is at the other. As for “returns”, Wyze doesn’t want the bad units back. Tell them what’s happening and they’ll send you a new sensor with no return.
Fair … Let us know when have play time.
Sometimes these charges are are parasitic. and must be drained.
First time replacing batteries 2 dead will not reconnect
haven’t read the whole thread but looks like I might have to get new ones out I have and replace
I guess I should have replaced batteries when I got the low battery indication 
New sensor works just fine but now I have to edit rules again for wyze and alexa, 
I had a similar situation.
I started out deleting an Alexa routine and writing a new one.
Found that was not necessary. Just change the trigger, clicking on the 3 dots in the upper right corner.
When you pair a new senor the rule is still looking for the old sensor (even if it’s the same sensor as before)
Chas
I too have a sensor that is now non-functional since the battery died. Tried every troubleshooting step and it will not connect to the bridge. I cannot get it replaced under warranty since I am not in the US. I am hoping that some of the handy members on this forum can figure out if we can somehow reset the device. If not, at least I have one left and a package of new batteries.