the Current Cost plug-in is available here

[pogle]

The XML data is averaged out every minute and then written to the history database, and the serial communications that receives the XML data is what drives all of this. Thus, if your USB connection was bad, or if you unplugged the USB interface and when you plugged it back in your computer assigned a different COM port then it would cause problems, but otherwise there is no explanation. The graphs are generated from the history database data. The only other potential problem is that which plagues millions of PCs everywhere - Vista.

As was the case yesterday, after a HS restart yesterday morning, the CurrentCost graph plotted correctly right up till 1045 PM and abruptly stopped (same time as yesterday). I have attached copies of the graph from this morning as well as the section of the log around 1045PM. I dont see anything in the log that would reference the CurrentCost logging. Also, this is running on a quadCore Vista machine with HSPRO and the latest CurrentCost plugin. All other HS controls and events are working properly. The USB connection has not been changed and the machine is not allowed to go into sleep or standby. Any other ideas?

[Rick Tinker]

As was the case yesterday, after a HS restart yesterday morning, the CurrentCost graph plotted correctly right up till 1045 PM and abruptly stopped (same time as yesterday). I have attached copies of the graph from this morning as well as the section of the log around 1045PM. I dont see anything in the log that would reference the CurrentCost logging. Also, this is running on a quadCore Vista machine with HSPRO and the latest CurrentCost plugin. All other HS controls and events are working properly. The USB connection has not been changed and the machine is not allowed to go into sleep or standby. Any other ideas?

Nope - there is definitely NOTHING that is time related in the plug-in to stop data logging. Even though the computer is not set to go to sleep, could the USB port be getting put to sleep? That can happen as a "low power" option on some computers.

[JWW]

Current Cost Classic Monitor

Hi Rick
Here are the errors from the debug log:
03/08/2010 17:55:15 ~ CDConnection( aConn )
03/08/2010 17:55:35 ~ Verifying Devices.
03/08/2010 17:55:35 ~ Initializing Current Cost on COM port 5
03/08/2010 17:55:35 ~ Opening COM port at 2400 baud for older Envi units.
03/08/2010 17:55:35 ~ Current Cost Startup Complete.
03/08/2010 17:55:35 ~ Updating ALL projected values
03/08/2010 17:55:35 ~ GetADOConnection: C:\Program Files\HomeSeer HS2\Data\CurrentCost\History.db
03/08/2010 17:55:38 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:55:44 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:55:50 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:55:56 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:02 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:08 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:14 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:20 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:26 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:32 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:38 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
regards, JWW

[Rick Tinker]

Hi Rick
Here are the errors from the debug log:
03/08/2010 17:55:15 ~ CDConnection( aConn )
03/08/2010 17:55:35 ~ Verifying Devices.
03/08/2010 17:55:35 ~ Initializing Current Cost on COM port 5
03/08/2010 17:55:35 ~ Opening COM port at 2400 baud for older Envi units.
03/08/2010 17:55:35 ~ Current Cost Startup Complete.
03/08/2010 17:55:35 ~ Updating ALL projected values
03/08/2010 17:55:35 ~ GetADOConnection: C:\Program Files\HomeSeer HS2\Data\CurrentCost\History.db
03/08/2010 17:55:38 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:55:44 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:55:50 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:55:56 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:02 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:08 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:14 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:20 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:26 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:32 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
03/08/2010 17:56:38 ~ Error, sensor is invalid in UpdateValuesFromXML(probably different data frame): -1
regards, JWW

This means that the XML data coming from the Envi is not right - what version and type of the Envi controller are you using? Are you sure that the baud rate setting in the configuration of the plug-in is correct? Can you use the Envi terminal emulator program to see the XML so I can have you check to see if a specific element is present?

[JWW]

Current Cost Classic Monitor

Hi Rick

The monitor is the original Classic version which runs at 2400 baud and is the same unit that worked OK after you added 2400 baud support in one of the earliest versions of the plug-in. Here is its output: Thanks JWW

[Rick Tinker]

I found a problem with the thread that reads the XML data from the serial port - it was not protected from an exception killing it, which is bad. I made that change and the latest (.42) is in the updater.

Also, you can now select a HomeSeer device to use for external temperature logging - you can tell it to use the string for the device and it will convert it to a single integer numerical value for logging (so it cannot have any extra characters besides numbers in it) or you can tell it to use the device's value, and enter an optional divisor. So if the temperature of 76.4 is saved in the device value as 7640, then use 100 as the divisor, otherwise leave it as 1.

The external temperature is logged every minute, but the graphs/charts do not yet use the data being logged. I will be working on incorporating it.

Also, the scales on the charts change based upon the size of the chart, so you hopefully will not have unreadable scales if you are using a smaller chart size.

The monthly comparison chart also changes to being a bar chart if the chart size is large enough, because a bar chart can show single data points such as when you have the first day of the month and cannot draw a line between two data points, but it will revert to a line chart when the chart size is smaller because a line is easier to see on a small chart.

[Rick Tinker]

Hi Rick

The monitor is the original Classic version which runs at 2400 baud and is the same unit that worked OK after you added 2400 baud support in one of the earliest versions of the plug-in. Here is its output: Thanks JWW

Did it truly work OK with the older plug-in? The reason I ask is because there is no sensor node in your XML, which is what is breaking it now. I can see the old plug-in defaulting to a sensor of 0 and now it does not, so that explains that, but you also have date information coming in an entirely different format. Come to think of it, the old plug-in did not use the date/time info, so I guess that explains it. Let me see if I can make some adjustments to handle this other data format. If we can't or it is too complex, we may not support this old of a unit, but off the top of my head I think it is doable.

[pogle]

Nope - there is definitely NOTHING that is time related in the plug-in to stop data logging. Even though the computer is not set to go to sleep, could the USB port be getting put to sleep? That can happen as a "low power" option on some computers.

Well, I don't think so. This is a Dell XPS desktop server and I have all the sleep/hibernation/etc. etc. turned off. Plus, it seems highly unlikely that the data logging would terminate at exactly the same time (1045 PM) two days in a row. I restarted HS this morning (again) and it is happily plotting my power now.. we'll see what happens tonight...

I do run an event at 1045 PM every day, but there is nothing in it that has anything to do with this plugin.. just turning off a half dozen lights, etc.. Is there any timing window that would stop graphing if a new current value was not received when expected.. i.e. if HS was busy doing other things...?

[Rick Tinker]

JWW,

Try the enclosed and let me know if it works. Run the Setup.exe or just double-click on the MSI file to install it (when HomeSeer is shut down).

Rick

[Rick Tinker]

Well, I don't think so. This is a Dell XPS desktop server and I have all the sleep/hibernation/etc. etc. turned off. Plus, it seems highly unlikely that the data logging would terminate at exactly the same time (1045 PM) two days in a row. I restarted HS this morning (again) and it is happily plotting my power now.. we'll see what happens tonight...

I do run an event at 1045 PM every day, but there is nothing in it that has anything to do with this plugin.. just turning off a half dozen lights, etc.. Is there any timing window that would stop graphing if a new current value was not received when expected.. i.e. if HS was busy doing other things...?

No, but given my discovery on the serial read thread, load up the latest version just to make sure and see if it makes a difference. If it doesn't, then it is just your luck that you are running Vista - the OS where mysterious things fail at random...

[JWW]

Current Cost Classic Monitor

Rick
Thanks for the code, at first glance all looks OK, extra fields have been generated and data has been collected and looks good, as below (about 23:10). Graphs are also working.
What version do I use from now on? Thanks, JWW

[snedecor]

I have all the latest software and just updated to 1.14 on iphone and
STILL can't get ENVI graphs to show..same on HStouchPad 1.14

I have the same problem.

The graphs show up in the designer, and show up in the client that runs on the PC, but deployed to the iphone, I get the text of the http that is sent to chart.apis.google.com.

[psampson]

$Value per KW/h correct in realtime device?

Hi Rick,

Great work proceeding here thanks!

I have the latest version just from the updater and noticed even though my price per KW/hr is 0.19945 I see 25c in the realtime device when only drawing realtime 720W.

If I disable the plugin, and re-nable it, briefly the realtime cost is 14c and then changes back to 25c in about 30 seconds and does not fluctuate after that. Seems from then on, no matter what the realtime consumption is 960W or 720W the realtime cost is fixed at 25c and does not seem to change.

I have verified all my table entries are .19945 and that table is in use.

Also, I use a cent-a-meter device which uses the API to update the plugin every 30 seconds. This has been working well. I also have several rf and fixed temp sources, is it possible to use the api to update the temp value now as well?

thanks and regards,

Phil

UPDATE: it looks like realtime cost per hour $ is not updating at all I restarted HS, went out for two ours and it is two hours behind everything else on the last change time stamp.

[pogle]

No, but given my discovery on the serial read thread, load up the latest version just to make sure and see if it makes a difference. If it doesn't, then it is just your luck that you are running Vista - the OS where mysterious things fail at random...

Hi Rick,

I updated to the .42 around 1030 last night and restarted HS.. I watched it until around 1048 and noticed that it made it past the 1045PM stoppage seen the last couple nights..although there was an unexplained gap in the trend around 10:06PM (first picture) but on checking this morning, it appears to have stopped completely at 1049 - a couple minutes after I last looked at it. The last two pics are from this morning.. showing last update at around 10:49:46 PM

Really puzzled now...

[Rick Tinker]

Rick
Thanks for the code, at first glance all looks OK, extra fields have been generated and data has been collected and looks good, as below (about 23:10). Graphs are also working.
What version do I use from now on? Thanks, JWW

Keep using that one until I release the next one into the updater, and it will have those changes from here on out so you should be good to go.