How did my CO detector know it had reached EOL?

Question

I was recently awakened by my CO alarm beeping. I take CO exposure very seriously and attended to the beeping immediately. It was not a CO concentration alarm, but rather an end of life alarm indicating that it needed to be replaced. Note that there is a different alarm for a malfunction, which also requires replacement.

Let me state up front that I have no desire whatsoever to defeat this warning. In fact, I promptly ordered a new one and then proceeded to take apart the old one. Others have done a breakdown of one similar to mine.

The sensor is a TGS5042. I went through an app note.

My unit included a date [m/d/y] and it was nearly 8 years ago – generally in the window for how long these sensors last. It looks like the EOL alarm worked and I am happy that.

How did the circuit know that the sensor was too old?

Certainly it is unlikely that there is any kind of clock/calendar involved - that seems unlikely to me. It also seems unlikely to me that operating time could be collected and used reliably because the sensor life may not be dependent on operating time – e.g., a 25 year old sensor that has operated for 30 days only.

The sensor has an analog output proportional to CO concentration, but CO concentration should be zero or near zero in a normal indoor environment. Is some baseline property of the sensor checked for a signal that is too weak or something similar? How could that characteristic be distinguished from a malfunction?

Probably just a timer. Remember timers can be made to operate all the time with a small internal battery, or only when operational (plugged in). — DKNguyen
– DKNguyen, Commented Mar 23, 2023 at 14:27
Why do you think it's unlikely that there's a clock involved? I think it's actually quite likely there's a clock of sorts. — Hearth
– Hearth, Commented Mar 23, 2023 at 14:31
It could also be a self-diagnostic thing with the sensor since it has a gel which may or may not degrade. In that case there would need to be a set of neutral operating parameters considered to be normal. Unless this is built-in to the sensor or cheap and easy to do, it's probably just a timer. — DKNguyen
– DKNguyen, Commented Mar 23, 2023 at 14:32
it may have a chemical timer of some kind ... measuring the migration of ions through a medium — jsotola
– jsotola, Commented Mar 23, 2023 at 14:35
It's a timer. I had one go end-of-life within 2-3 days of the expected EOL based on the in-service date I wrote on it. Sensor degradation isn't going to be that accurate. — user71659
– user71659, Commented Mar 23, 2023 at 22:32

JRE · Accepted Answer · 2023-03-23 15:24:01Z

Pages 7-9 of the linked app note describe a self test procedure that the device using the sensor should repeat periodically.

Based on the results of the test, the device can determine if the sensor is still usable or not. There's a certain voltage range that the sensor output should reach at the end of the test. If it is outside that range, then the sensor would be deemed end of life.

While there will have to be a timer of some kind involved because the self test has to be repeated periodically, I wouldn't expect the timer to the only criteria because the gel in the sensor will dry out over time regardless of whether or not it is in use.

DrG · Accepted Answer · 2023-03-23 15:30:38Z

I think that JRE's answer is probably as good an answer as I am going to get. The patent that is referred to in the app note looks to me like it is very specific to the sensor only and within the capabilities of the circuit, which include a PIC MCU

and the second IC that

I could not identify but in the linked teardown, it was loosely speculated that it could be an op amp IC.

Not that I completely understand the app note or the patent, but it seems like there is enough there to accomplish the sensor diagnostic.

While under battery operation, doing a periodic diagnostic could easily be done by a counter/timer in the PIC. I still don't think using a timer to test the age of the unit is likely, but the the sensor diagnostic is what you really want to do anyways.

Not sure about what differentiates the malfunction alarm from the EOL alarm, but it is conceivable that it could test a number of other characteristics.

I suppose I could have spent more time on this and figured out more on my own, but my EE skills/experience is limited...so thanks.

Stack Exchange Network

How did my CO detector know it had reached EOL?

2 Answers 2

Your Answer

Hot Network Questions

How did my CO detector know it had reached EOL?

2 Answers 2

Your Answer

Sign up or log in

Post as a guest

Hot Network Questions