Hi list, I am happy to report that this problem is solved! Please see below (inline) for the details...
On 03/02/2011 09:49 PM, Eloy Paris wrote: > On 03/02/2011 05:49 PM, Paul Alfille wrote: > >> Maybe some static charge in the Vcc lead confused the sensor in the >> lightning storm? > > You know what, I want to get to the bottom of this so I have decided to > do nothing now to try to fix the problem I experienced a couple of > nights ago when we had lots of rain and thunderstorms. Allow me to explain: > > Whatever happened a couple of nights ago surely will happen again. When > that happens I'll go outside and start with the simple stuff, i.e. > grounding Vcc. I am hoping that the problem was caused by either high > humidity, or by some weird electric problem caused by not having Vcc > grounded. If it is the later I will be able to confirm it by grounding > Vcc when the problem is happening. As I said, I have not had a single 85 > degree reading since that night, so I have no need to try to fix things > right. > > I'll report back as soon as I know more. It happened again last night. It started around 11:45 PM. The forecast indicated that we were going to have some rain but the 85 degree readings started even before rain started. When I woke up this morning I had lots of emails, 5 minutes apart, indicating that my outside DS18B20 was reporting 85 degrees (I read temperatures every 5 minutes and have the polling script email me if a sensor reports a temperature outside a range). It seems like we had some rain overnight because the ground was wet. However, it did not seem like it was a heavy rain, and we did not have any thunderstorms. So, with these perfect conditions, and knowing that the problem was occurring, I decided to take a first stab at fixing the problem. As I mentioned before, the easiest thing to try first was going to be to ground Vdd, which I've had floating since I installed this sensor the last week of November. And that was it! I've been running for 9 hours now and have not received another 85 degree reading. Furthermore, about 30 minutes ago we had a big thunderstorm with strong winds and heavy rain, and that did not cause the sensor to report 85 degrees. So there you have it, grounding Vdd is what fixed the 85 degree readings. I think I won't be needing fancy conformal coating after all -- the sensor is not directly exposed to the elements and it seems to be performing great now that I grounded Vdd. Regarding the root cause, I think some of you mentioned something about increased capacitance when humidity is high? It must have been something related to increased humidity since we did not have lightning when the problem started last night. I am not rushing to ground Vdd in all my other sensors right now, though, since they have not reported a single 85 degree reading in several months of operation. I guess it does not get that humid inside, although now that summer is close (and it gets really humid where I live) that may happen, so I shall see. In any case, the minute I get an 85 degree reading from a sensor I am grounding that sensor's Vdd. Guess DS18B20 sensors make for cheap high humidity detectors when Vdd is left floating ;-) Anyway, thanks to all for all the great comments and help. It's definitely been a very interesting thread. Cheers, Eloy Paris.- ------------------------------------------------------------------------------ What You Don't Know About Data Connectivity CAN Hurt You This paper provides an overview of data connectivity, details its effect on application quality, and explores various alternative solutions. http://p.sf.net/sfu/progress-d2d _______________________________________________ Owfs-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/owfs-developers
