516 rfxcom problems

[mhwlng]

my motion detectors send their state 5 times,

I always see errors in some of the commands :

------------------
B1:ON
B1:ON
B1:ON
B1:ON
>>>>> Error decoding block No matching handler for a block with 33bits
Attempting resync from 7
Attempt 2 partition data 70 8F 00 FF 00 p�.ÿ.
--------------------
B1:OFF
B1:OFF
Not a 1-wire
>>>>> Error decoding block nil
Attempting resync from 12
Attempt 2 partition data 70 8F 20 DD p� Ý
>>>>> Error decoding block No matching handler for a block with 112bits
Attempting resync from 2
Attempt 2 partition data 70 8F 20 DF p� ß

-----------------------

Also, my power meter data is not parsed properly :

this device also sends the same data 5 times, and this is the result :

Internal serial number = 00 F0 31 64 00 86 .ð1d.†
Internal serial number = 00 F0 30 BC 00 86 .ð0¼.†
Internal serial number = 00 C8 18 A6 00 43 .È.¦.C
Internal serial number = 00 F0 30 3C 00 21 .ð0<.!
Internal serial number = 00 F0 31 78 01 0C .ð1x..

-------------------------

one difference between your code and mine is, that you initialise the rfxcom with : F041

and I use mode F02C....

I don't know if this matters...

------------

also, what does the X10 RF Transmitters component do ?

I don't see any different with or without this component enabled ?


Marcel

[Rob H]

The RFXCOM component is in a transitional state at the moment and doesn't currently interact with the X10RFTransmitters class or the RFXCOMTransmitters class, but that should be sorted soon. Once this has been fixed then the RFXPower should work (I don't have one of these to test against but Mike does).

Re the errors - I see these quite often myself, and I assume that they are due to RF interference corrupting the signal. The RFXCOM component attempts to resync, but that doesn't work all that often. The X10RFTransmitters component should take care of the repetition of events.

I initialise with F041 because I want the variable length data for Visonic sensors while still receiving X10 - I'm still not clear whether I need to initialise the X10 side first to get the full length data from X10 security devices - I've noticed using the example code from RFXCOM that there seems to be a difference in X10 output depending on the initialisation order.

[Promixis]

Marcel,

too clarify - we have separated the RF devices into Tx and Rx.

Rx we will support are the W800RF32 and RFXCOM

TX we will support are X10 Standard and Security, RFXCOM devices, and Some Visonic +/- others

[mhwlng]

I still have problems with 518

I have x10 motion detectors that work fine
rfxcom power meter data is not parsed properly :

B10:OFF
B10:OFF
B10:OFF
B10:OFF
B10:OFF
Internal serial number = 00 F0 8E 7A 02 07 .ðŽz..

after the rfxcom power meter data is received (sent every 30 min.), also the x10 motion detectors stop working..

Marcel

[Rob H]

Looking at this now - my code attempts to resync if it finds unexpected data in the buffer, but this doesn't appear to work as well as I'd like.

[Rob H]

Can you try this replacement for RFXCOMReceivers.lua (goes in the lua\Components folder).

[mhwlng]

when the power meter data comes in, I get results in attached :

the x10 motion detectors still work afterward (I'm not sure if some motion detector events were lost during the resync process..)....

Marcel

[Rob H]

You shouldn't lose any motion detector events during a resync - I have however seen it produce an occasional spurious result e.g. I got an L13 ON while testing this version and I don't have any such device.

I may change the way that this is handled, e.g. just throwing away any block that doesn't have valid data.

Not sure about the power meter data - I don't have one of these devices to test with which makes it a bit tricky.

[mhwlng]

Not sure about the power meter data


ok, I'll revert back to my old rfxcom component.....

If you need me to test something in the future, let me know...

thanks,

Marcel

[Rob H]

Mike has one though ;)

[mhwlng]

still problems with 519 :

I now see X10 devices, that I don't have :



Time Date Source Details Payloads
11:19:18:595 5/14/2007 X10 RF Transmitters L10:OFF
11:19:18:595 5/14/2007 X10 L10:OFF
11:19:18:515 5/14/2007 X10 RF Transmitters J15:ON
11:19:18:445 5/14/2007 X10 J15:ON
11:16:26:938 5/14/2007 X10 RF Transmitters J13:OFF
11:16:26:858 5/14/2007 X10 J13:OFF
11:16:11:917 5/14/2007 X10 RF Transmitters L9:OFF
11:16:11:817 5/14/2007 X10 L9:OFF
11:15:27:102 5/14/2007 X10 RF Transmitters L10:OFF
11:15:27:092 5/14/2007 X10 L10:OFF
11:15:12:541 5/14/2007 X10 RF Transmitters J14:OFF
11:15:12:441 5/14/2007 X10 J14:OFF
11:14:20:957 5/14/2007 X10 RF Transmitters L10:ON
11:14:20:867 5/14/2007 X10 L10:ON
10:55:21:619 5/14/2007 X10 RF Transmitters L10:OFF
10:55:21:609 5/14/2007 X10 L10:OFF



no sign of power meter data (which could be above, I'm not 100% sure)

also, no messages in the console

Marcel

[Rob H]

I was afraid of this, as I think I said earlier. I'll see what I can do to fix it.

[Promixis]

I was afraid of this, as I think I said earlier. I'll see what I can do to fix it.

note: some random x10 events always will happen...

[Promixis]

Mike has one though ;)

and we will be working on it :D

[mhwlng]

note: some random x10 events always will happen...


no, I never see this on my own rfxcom component, this must be caused by some parsing error related to resyncing or power meter packets...

Marcel

[mhwlng]

in build 520, I now see this each time I start G5 (I added an rfxpower device)



...\Girder5\luascript\components\RFXCOMTransmitter s.lua:222: bad argument #1 to `ipairs' (table expected, got nil)
stack traceback:
[C]: in function `ipairs'
...\Girder5\luascript\components\RFXCOMTransmitter s.lua:222: in function `Enable'
...iles\Promixis\Girder5\luascript\ComponentManage r.lua:305: in function `EnableComponent'
...iles\Promixis\Girder5\luascript\ComponentManage r.lua:552: in function `StartEnabledComponents'
...iles\Promixis\Girder5\luascript\ComponentManage r.lua:219: in function <...iles\Promixis\Girder5\luascript\ComponentManage r.lua:218>
[C]: ?
...iles\Promixis\Girder5\luascript\ComponentManage r.lua:218: in function <...iles\Promixis\Girder5\luascript\ComponentManage r.lua:217>



and I don't see any rfxpower events in the logger

p.s. unrelated to this :

can you add this image to the next distribution :



Time Date Source Details Payloads
09:54:49:459 6/2/2007 DUI Subsystem Could not load requested image: C:\Program Files\Promixis\Girder\plugins\UI\netcallerid.jpg



Marcel

[Promixis]

bug fixed

for _,event in ipairs (self:GetItemCondition (item,'CounterEvents') or {}) do

[mhwlng]

I still have problems (build 520)...

1) I don't see an event in the logger, with the current power meter counter value.

how do I subscribe to this data ?

2) I see a number in component manager rfxcom transmitters current condition :

the value was 4085.2 kWh

which is wrong, because the value should be around 2177.23

so, I think there might still be some parsing error ?


Marcel

[Promixis]

RFX = ComponentManager:GetComponentUsingID (10047)

rfxeh = function (event,deviceaddress,...)

print (event,deviceaddress,unpack (arg)

end

RFX:Subscribe (rfxed)


the code for power is in the rfxcomrx.lua component, line 600

[mhwlng]

mike,

I found the error in your code :

you need to change from

local countervalue = bytes [3] + (bytes [4] * 256) + (bytes [5] *256 *256)

to

local countervalue = bytes [4] + (bytes [3] * 256) + (bytes [5] *256 *256)

We'll discuss my fee, later :)

and in your above example, you use rfxeh and rfxed (took me a while to figure out why it didn't work :rolleyes: )

Marcel

[Promixis]

thanks. changes made to svn.

[mhwlng]

the rfxcom device sends 5 RF packets every 30 minutes,

If I print the value from the event, sometimes see this :

17:24:16 RFX 2374.57
17:24:19 RFX 2374.57
17:24:19 RFX 2374.57
17:24:19 RFX 2374.57

17:37:40 RFX 2374.57

17:54:47 RFX 2375.19
17:54:47 RFX 2375.19
17:54:47 RFX 2375.19
17:54:48 RFX 2375.19

17:57:10 RFX 2375.19

you see that one event is delayed by 13 minutes and one by 3 minutes...

all 5 packets should have the same timestamp ????

Marcel

[Rob H]

That's really weird - I'll try to see if I can work out what's going on tomorrow.

[mhwlng]

Here's another strange one Rob

RFX 2386.19 10:27:21
Error validating Visonic PowerCode B0 parity error 10:27:23
RFX 2386.19 10:27:23
Error validating Visonic PowerCode B0 parity error 10:27:23
RFX 2386.19 10:27:23
Error validating Visonic PowerCode B0 parity error 10:27:23
RFX 2386.19 10:27:24

Error validating Visonic PowerCode B0 parity error 10:27:33
RFX 2386.19 10:27:33
Error validating Visonic PowerCode B0 parity error 10:27:33

here, the event is called normally with a valid kWh value, so I don't understand the parity error at the same time ?
(I don't have any visonic equipment, so if the packet is corrupted, why is my event still called)

I also sometimes see errors like this :


VCS 95 1A 65 80 20 F5 0A F7 08 •.e€ õ.÷.
100 111 false true
{ -- #0
[1] = "Light",
[2] = "Arm",
[3] = "Disarm",
} -- #0
VCS 00 20 D5 2A 34 CB 20 D5 2A . Õ*4Ë Õ*
6 205 false true
{ -- #0
[1] = "Light",
[2] = "Disarm",
} -- #0

Error validating Visonic PowerCode B1 parity error
Error validating Visonic PowerCode B0 parity error



could really be bad packets, but I believe that the latest version of RFXCOM firmware doesn't send bad packets with parity errors etc. to the pc anymore ?

[Rob H]

Looks as though it's receiving a bitstring of the appropriate length for a powercode signal that isn't a powercode signal but rather an RFXPower signal. The error/debug message is clearly inappropriate.