Topic: Is the Access corrupting my rainfall data?

[BeaverMeadow]

Why does my model 06058 display show the true .03" rain I received yesterday whereas the data sent to MyAcurite and WU shows 2.58"? That erroneous rain data was received in two discrete instantaneous readings occurring at 6:04pm and then at 9:59pm (as shown below).

My Access is sending this info out to MyAcurite and WU. Is this a malfunction of the Access?

Here's the WU data for the erroneous rain readings:

6:04 PM   72F   51F   47%   South 2.0mph   3.0mph   29.80in   0.00in   0.02in   w/m²
6:09 PM   71F   52F   51%   South 2.0mph   2.0mph   29.80in   1.28in   1.30in   w/m²

9:59 PM   51F   50F   97%   NNE   0.0mph   0.0mph   29.86in   0.00in   1.30in   w/m²
10:04 PM   51F   50F   97%   NNE   0.0mph   0.0mph   29.86in   1.28in   2.58in   w/m²

Thanks for any ideas. I'm really curious about how this could happen.

[BeaverMeadow]

This is the third or fourth time this has happened-- My display shows correct rainfall of .03. The Access is sending out to WU and MYAcurite a reading of 1.19" of rainfall.

3:29 AM   60F   60F   99%   NNE   0.0mph   0.0mph   29.82in   1.19in   1.19in      w/m²


Am I the only person with the Access to have this erroneous reading occur? I am blaming the Access but if anyone has any other theories I'd like to hear them.

 

[worachj]

I really don’t know what’s going on. But I’m wondering if your 5-N-1 has periods of lost connections with the Access and the totals get messed up when it reconnects. While your display is always connected with the 5-N-1 and is always getting updated. I know the Access has battery backup for power outages, but don’t know what happens with the data gap during lost connections to the 5-N-1.

Do you notice any data gaps on WU when this occurs?

[BeaverMeadow]

Thanks for the theory. I don't see any gaps below in the uploads to WU. My signal strength is a consistent 3 out of 4 bars.

3:19 AM   60F   60F   99%   SW   0.0mph   0.0mph   29.82in   0.00in   0.00in   w/m²
3:24 AM   60F   60F   99%   NW   0.0mph   0.0mph   29.82in   0.00in   0.00in   w/m²
3:29 AM   60F   60F   99%   NNE   0.0mph   0.0mph   29.82in   1.19in   1.19in   w/m²
3:34 AM   60F   59F   99%   SW   0.0mph   0.0mph   29.82in   1.19in   1.19in   w/m²
3:39 AM   59F   59F   99%   NNW   0.0mph   0.0mph   29.81in   1.19in   1.19in   w/m²
3:44 AM   59F   59F   99%   WNW0.0mph   0.0mph   29.81in   1.19in   1.19in   w/m²
3:47 AM   59F   59F   99%   WSW0.0mph   0.0mph   29.81in   1.19in   1.19in   w/m²
4:04 AM   58F   58F   99%   NNE   0.0mph   0.0mph   29.81in   1.19in   1.19in   w/m²
4:09 AM   58F   58F   99%   SW   0.0mph   0.0mph   29.81in   1.19in   1.19in   w/m²

[worachj]

That is strange!!!

Added:
I still wonder what happens if the 5-N-1 loses connection with the Access and there's rain data during that period. There could have been a lost connection and rain data during the five minutes gaps that the data shows.

Wonder if there's a way to test for a rain event on the 5-N-1 during a lost connection with the Access.

Second Add:
I wonder if pulling the batteries on the 5-N-1 during a rain event for two minutes would simulate a lost connection with the Access and see what happens when the batteries are reinserted.

I may try it with my 5-N-1, but there's bad weather here today.

[BeaverMeadow]

I would think that if there is a lost connection between the 5 in 1 and the Access during a rain event the data would be gone since the 5 in 1 has no stored data to upload after the connection is re-established.

But this has nothing to do with my problem-- there was no rain event when the Access falsely reported 1.19" of rain to WU and MyAcurite websites.

Erroneous data is not a good thing when an instrument company such as Acurite claims accuracy. I mean, it's in their company name!

I'm still thinking something is intermittently amiss with the Access based on the fact that the bad data is going out to WU and MyAcurite but not to the Acurite display.

[worachj]

Thanks for clarifying that it happened on a no rain event. I won't waste time testing.

Its weird in your first post that 1.28 in was added each time. That's strange.

[galfert]

The 1.28 is in the rain rate column. The other column is the rain totals for the day as they accumulate. Seems that the Access is erroneously generating a 1.28 rain rate when it isn't raining. It did it 2 times, with the same value.

Reset the Access, pull batteries out and power cycle.

If it persists then contact Acurite and arrange for replacement. Plenty of people with the Access and I've not seen this issue before. So I doubt it's a firmware bug. Seems more like a hardware malfunction (because it doesn't happen to others).

Doubt the 5-in-1 is at fault because the display is reporting fine. The fact that the error shows up for for myAcurite and WU indicates that the problem is with the Access.

[nincehelser]

It wouldn't be a lost signal.

Rain data is sent as a rolling counter value from the sensor.

The receiving device (Access or display) computes the difference from the last received value with the current value.  That difference is the rain amount for that time period.

For example, suppose the current counter value is "123" and it hasn't been raining.  That counter value will stay constant until the tipping bucket tips with some rain, then it will change to "124".  Now suppose it is raining and the signal is lost for some reason.  Let's say the signal was reacquired an hour later, and the counter is now "200".  The receiving device will then compute the difference of "76".  That translates into "0.76 inches" of rain that accumulated during the signal outage.

This means you won't lose the rain accumulation during a signal outage, but your "rain rate" calculations will likely be goofy.

As for the consistent value of 1.28 (or 128 bucket tips), that's a power of 2.  That makes it sounds like a particular bit is getting flipped occasionally in the Access.

I'd definitely ask Acurite for a warranty replacement of the Access.

[BeaverMeadow]

It wouldn't be a lost signal.

Rain data is sent as a rolling counter value from the sensor.

The receiving device (Access or display) computes the difference from the last received value with the current value.  That difference is the rain amount for that time period. ........

.....As for the consistent value of 1.28 (or 128 bucket tips), that's a power of 2.  That makes it sounds like a particular bit is getting flipped occasionally in the Access.

I'd definitely ask Acurite for a warranty replacement of the Access.

I don't get a consistent value of 1.28". The first time I noticed it it was 1.28. This most recent time it was 1.19".

I'll contact Acurite.

Thanks.

[galfert]

You didn't say if you tried powering down and pulling the batteries as I suggested. Sometimes devices need to be restarted.

[BeaverMeadow]

You didn't say if you tried powering down and pulling the batteries as I suggested. Sometimes devices need to be restarted.

Oh sorry-- yes I powered down and pulled batteries. But the erroneous reading only shows up intermittently so I also contacted Acurite about the issue:

Our offices will be closed June 29th through July 7th for a company holiday. Please note that order processing will be delayed during this time, and customer support will be closed. All customer inquiries will be answered in the order they are received when we return on July 8th.

Thank you for your patience during our company holiday.

So I'll have a week to see if the error occurs again.

[mikemaps]

I've had my Atlas 5 months now & had that happened once when Access reported 1.17" and only .03" fell. That was a few months ago & it hasn't happened since.

[BeaverMeadow]

I received a reply from Acurite now that they are back from July 4th vacation:

-------------------------------------------------------------------------------------------------------------------------------------------------
Jul 8, 1:02 PM CDT

Thank you for contacting Acurite. We have had a small handful of users experience this same issue as you have described. We have provided your information to our development team for further investigation of the cause of this issue. We will update you when we have more information to provide at that time.

Have a wonderful day,

Tasha
AcuRite Customer Support
--------------------------------------------------------------------------------------------------------------------------------------------------

BTW, I had the false rain reading again on July 6.

8:24 PM   72F   72F   99%   NNE   0.0mph   0.0mph   29.90in   1.11in   1.17in


I am wondering if the WU data comes from Acurite's server or directly from the user's Access? (I recall that there was a change in how the data was handled after the transition from the SmartHub to the Access). If the data is sent to WU from the Acurite server then the erroneous rainfall readings that I and others are having might not originate at the Access at all but instead at the Acurite server.

[mikemaps]

I had it happen again today. My display shows .54” of rain today which is the correct figure yet my acurite and WU are showing 2.46”. Most times it is correct but this is the second time it was way different. Hope they can figure it out.

[nincehelser]

I am wondering if the WU data comes from Acurite's server or directly from the user's Access? (I recall that there was a change in how the data was handled after the transition from the SmartHub to the Access). If the data is sent to WU from the Acurite server then the erroneous rainfall readings that I and others are having might not originate at the Access at all but instead at the Acurite server.

It is sent to wunderground directly from the Access.

I think the protocol change you're thinking of was the firmware upgrade from "Bridge" to "SmartHUB".  The bridge firmware sent rain data for the last 36 seconds.  The problem with this if you lost your network connection your rain data during the outage was lost.

SmartHUB corrected that by keeping track of the daily rainfall since midnight and the last 60 minutes like wunderground wants.

[BeaverMeadow]

......It is sent to wunderground directly from the Access.

[/quote]

I am wondering if the WU data comes from Acurite's server or directly from the user's Access? (I recall that there was a change in how the data was handled after the transition from the SmartHub to the Access). If the data is sent to WU from the Acurite server then the erroneous rainfall readings that I and others are having might not originate at the Access at all but instead at the Acurite server.

It is sent to wunderground directly from the Access.

I think the protocol change you're thinking of was the firmware upgrade from "Bridge" to "SmartHUB".  The bridge firmware sent rain data for the last 36 seconds.  The problem with this if you lost your network connection your rain data during the outage was lost.

SmartHUB corrected that by keeping track of the daily rainfall since midnight and the last 60 minutes like wunderground wants.

So that leaves the Access as being the originator of the erroneous rainfall readings.

[nincehelser]

So that leaves the Access as being the originator of the erroneous rainfall readings.

Yes. 

[BeaverMeadow]

I have heard back from Acurite today regarding this issue--

They said: "I understand your frustration however, getting a new Access will not correct the rain reading issues for you. At this time we do not have an ETA for when a solution will be found. We do understand it is very frustrating. We are working to find a solution as soon as possible. We do appreciate your patience and understanding while we continue to search for the best solution."...

I responded: "If only a handful of people are having this issue wouldn't that indicate that it is a faulty Access? If not a faulty Access then what else could be the cause?"
 

[worachj]

I wish they would state what the problem is. It could be a sensor/firmware/Access problem or a combination of each. My guess would be that they’re trying to come up with the cheapest solution like a firmware solution. Replacing hardware is expensive for them.

The Acurite Access can accept up to 7 sensors.  The following sensors are compatible:
 

• 06059M AcuRite Atlas Weather Sensor
• 06014M Pro+ 5-in-1 Weather Sensor: Temperature, Humidity, Rainfall, Wind Speed and Wind Direction
• 06004M 5-in-1 Weather Sensor with Temperature, Humidity, Rainfall, Wind Speed and Direction
• 06008RM 3-in-1 Weather Sensor with Temperature, Humidity and Wind Speed
• 06002M Wireless Temperature and Humidity Sensor
• 06044M Room Temperature and Humidity Sensor
• 00276RM Room Monitor
• 00275RM Outdoor Temperature and Humidity Monitor
• 06010RM Add-on Water Detector (requires a room monitor)
• 06011RM Add-on Liquid and Soil Temperature Sensor (requires a room monitor)
• 06012RM Add-on Spot Check Temperature and Humidity Sensor (requires a room monitor)
• 06034RM Wireless Rain Gauge Collector [/l][/l]
   

[Mabcmb]

I to had a false reading today for the first time on my Atlas system . Showed .12 rain when it's sunny and no wind . Funny part is my display shows no rain at all . Plus as you can see it disappeared later in the morning .
 [ You are not allowed to view attachments ]

[nincehelser]

I have heard back from Acurite today regarding this issue--

They said: "I understand your frustration however, getting a new Access will not correct the rain reading issues for you. At this time we do not have an ETA for when a solution will be found. We do understand it is very frustrating. We are working to find a solution as soon as possible. We do appreciate your patience and understanding while we continue to search for the best solution."...

I responded: "If only a handful of people are having this issue wouldn't that indicate that it is a faulty Access? If not a faulty Access then what else could be the cause?"

It might mean they think it is an issue that will be fixed in the next version of firmware.  I could be something like filtering out a spurious reading.

[worachj]

An anomaly happen to my setup after getting my Atlas repaired for false lightning strikes and a new replacement Access.

Both of these false readings happen at the same time. A spike of 256 lightning strikes and .01 inch of rain was picked up by the Access and sent to MyAcuRite.com and WeatherUnderground, but it was not picked up by the HD display. Luckily it was only a one time occurrence.

So it does seem to be related to the Access since it doesn't seem to be happening to the HD display.

[BeaverMeadow]

Another erroneous rainfall reading today.Was told by Acurite to be patient. Very irritating. I keep cvs data and have to manually edit out Acurite's errors. Now, reviewing back to when I initially set up the Access, I wonder whenever I see rainfall data of around 1" whether it is a false reading or not.

Do these types of sporadic erroneous readings occur with Ambient or Davis pws' too? Maybe this level of accuracy is the norm for home based weather stations and I should just accept it.  

 [ You are not allowed to view attachments ]
 [ You are not allowed to view attachments ]

[galfert]

Do these types of sporadic erroneous readings occur with Ambient or Davis pws' too? Maybe this level of accuracy is the norm for home based weather stations and I should just accept it.  

I have a few Ambient (Fine Offset) consoles (4 versions); WS-2902A, WS-2000, ObserverIP, GW1000 and send data to Meteobridge from the GW1000. None of the consoles has ever had a false reading from any sensor if the sensors are in working order, meaning the consoles don't corrupt data. I've never heard a console being the source of corrupting data to any other Ambient user nor to any Davis user. That is when there is no problem with a sensor. So sure with a bad sensor like a bad humidity sensor then false data is being generated by the sensor itself. Also with Ambient when the batteries die and the when solar panel charged super capacitor isn't enough power to run the unit then you end up with saw tooth data from the sensor coming on and off sporadically. What also does happen is if you are at the fringe distance is that you lose packets so there is missed data. But with a good signal nothing like this ever if the sensors and batteries are in working order.

EDIT: Oh I almost forgot. One other time I've seen false data is when there is morning water drop that hasn't evaporated after a night rain or a lot of morning dew. This only happens sometimes. What happens is that this water drop  (which is small and does finally evaporate quickly as the sun comes up). But for a brief moment can cause the UV and Solar Radiation to spike as it acts like a lens that focuses the morning sun in such a way that the sensor gets a bad read. Again in every instance I've spelled out occurrences that originate from the sensor unit.

No console has ever been the source of erroneous data....at least not that I've heard of. Well except for some early firmware builds where things like once the temperature was indicating zero at midnight. Little things like that but they were very short lived, which makes them insignificant. So short lived that I can't even remember these issues. They are bugs that are sometimes introduced by new firmware. The neat thing about Ambient and Ecowitt is that firmware fixes do occur very often (mostly to introduce new features) and they fix issues within days. I've reported numerous firmware bugs and they get patched very fast.