Dramatic reduction of received packets

[Digit]

I just don't know what's happening here...[:(]

Today i saw some really strange things going on. Suddenly the amount of packets received by my RFXCOM receivers dropped dramatically.
I noticed this by motion sensors not going back to 'no motion' state the way they should (that is, by receiving a packet that says so).
I looked at my logfiles and somewhere around 06:25 today, suddenly the packet-rate on both of my RFXCOM receivers dropped enormously.
When i came home after work this evening, i thought it was RFX1 (see below) that wasn't functioning anymore (based on the sensors that were not being received anymore), so i dis- and reconnected power and left home again. Now that i'm back i had some time to analyze what the **** is going on. Here some statistics:

Code: Select all

                  RFX1   RFX2  Total
25-11-2008 02:00   506    534   1040  (these are normal )
25-11-2008 09:00    34     23     57 
25-11-2008 23:00   130    343    473


The numbers represent the number of packets received from a RFXPulse module during 1 hour. I focused on that module since that produces a relatively constant packet-flow. And it's not on batteries, so i could rule that out right away too.
But it's not just the RFXPulse module(s), everything has dropped dramatically regarding the number of packets received...

What is going on here? What can i do? I just don't get it.... [xx(]

[Bwired]

The only explanation I can think of in your case is heavy interference!
Did you check this.

[Digit]

Hmm, with the means i have available..., yes.
I checked if there was a switch or something else constantly sending out packets (i have had that once before); no, there isn't.
I checked the amount of parity errors on incoming packets. They were less, ofcourse. Related to the amount of packets you could say it is the same as before.
I checked the amount of unrecognized packets. They increased dramatically too, from 37 per hour, 295 per hour to even 354 per hour!

So now i think i need some kind of 'geiger-teller' for 433 MHz so i can walk through my house (and surroundings) to capture that interferer... but how?

[Bwired]

Is the rfxmeter causing problems as well, if so you can cut the antenna on one side so the range gets better (always do this).
Device you can easily move, move them closer. I had similar problems once a while back.
The devices you receive are they close or far from the receiver?

[Digit]

Yes, the RFXMeter is received worse (see the statistics for that).
I only checked 1 RFXPulse module, the RFXMeter has 3 of them (Power, Gas, water). Distance between RFXMeter and receiver RFX1 = approx. 2 meters. No concrete walls, just wood between them. So reception should be no problem at all i'd say.

[flexer]

<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by Digit</i>
<br />I just don't know what's happening here...[:(]

Code: Select all

                  RFX1   RFX2  Total
25-11-2008 02:00   506    534   1040  (these are normal )
25-11-2008 09:00    34     23     57 
25-11-2008 23:00   130    343    473


<hr height="1" noshade id="quote"></font id="quote"></blockquote id="quote">Digit, how do you measure and generate these statistics?

Flexer

[Bwired]

We have our own driver and every time a certain packet comes in we can count it.
It's also very handy for checking like: if in X minutes a packet is not received then report it.

[flexer]

Hi Pieter,

I understand that you have your own driver...

But which driver is it? The RFXCOM HS plugin, RFCReceiver? Or something custom made? I'm also using the RFXCOM receiver, but I'm not aware of a standard RFXCOM driver/utility that can do this.

Regards,

Flexer

[Bwired]

We are not using Homeseer.
We have our own made driver, Rfxcom comes with a VB.NET driver as well. Its possible to adjust this code with counting the different received devices (packets).

[Digit]

Flexer,
Pieter is right, just as Pieter i have written my own code to handle the RFXCOM Receiver & Transmitter. And then these things like counting packets are very easy to do.

[Digit]

Small update: our garage door opener, which has 433.92 MHz remotes, has also stopped responding to those remotes.. (

[b_weijenberg]

Digit,

It is possible that the only sensor that is received OK is the one that creates the problem. If that sensor transmits rubbish between the correct packets it disturbes all other RF transmissions.
If you can't find the bad sensor you can rent me for a few hours to help you next week using the spectrum analyzer.

Bert

[Jfn]

The time it started is very strange, that early in the morning.

A few possible options:

Do you live close to a construction site where for example a crane is operated using wireless control? Starting that early in the morning is not unusual for these people.

Do you live very close to a mobile-phone base station (within a few meters that is, for example in an appartment with a base station on top) These base stations are very well filtered, but 800 - 900 MHz radiation is detectable on 433 MHz when getting close to the antenna's).

Did your neighbour start using his/her wireless headphone on 433 MHz? Although I doubt it considering the time of day, who knows?

Did it stop at a certain time, or is it an intermittant problem, or is it still causing troubles?

[Digit]

Problem solved !! It was a MS13 motion sensor sending garbage.
This evening i decided to 'pull the plug' on everything that could send. Slowly removing the batteries from every device, closely watching the number of bad packets coming in. This was the situation when i started:


As more and more devices were taken off the air, the number went down, but not completely, untill i removed, ofcourse, the battery from the last motion sensor:
[img]uploaded/Digit/20081127215932_Capture2.jpg[/img]
RFXMeter was received perfectly, no bad packets anymore, so after a few minutes i tried the garage door opener. It worked again. It produced 3 bad packets, nevermind... now i'm clean for over 40 minutes [8D]

Now i can start programming all the MS13 sensors again, replacing addresses of the Oregon sensors, etc. etc. but still, you could say i'm very happy now
And thanks for thinking with me!

[Bwired]

Bert was right
Did you see these packets earlier in your receiver software, guess not otherwise you solved it already.
Why the bad packages on multiple ms13 so suddenly?