ISY-994 IZ PRO ENERGY RELATED ISSUES:
During the last two years of trying to integrate the ISY with the GEM. There have been some consistent problems being seen. To date there hasn't been any forward movement in identifying or resolving these issues.
I am hoping very much that in Beta 5.XX which is due next year will address some of the feature disparities and resolve some of the on going issues I have seen, and endured.
DEVELOPMENT:
Currently development of the Energy Module for the ISY has come to a screeching halt. Its safe to say nothing worth mentioning has appeared for the last two years for the module. In the image below there is a field called *Current Current* to date this box has never worked in reporting the amount of current / amps the house is using.
The next issue relates to energy values that are available from the drop down list that can be applied to programs. To date, channels 1 & 2 can use Amp, Volts, Watts, KWH, Polarized Watts from the list.
The remaining 30 channels only have the two options which are Watts & KWH. Having the ability to select all of the listed energy options would prove extremely helpful in crafting programs and allow flexible monitoring.
There are instances where the current / amperes need to be measured opposed to watts. There is also circumstances where the end user has a dedicated circuit that is generating power (NET Metering) and having the ability to monitor and track polarized values would be beneficial.
As it stands the above options are not available.
A long time feature request has been having the ability to reset the counters in the GEM. As you can see there is a option to reset the counters but this is a manual method. There is no method to add this to the programs and have the system reset all the counters for the end of the month.
The current work around is to set a reminder on my smartphone and do this task at 12:00 AM each month.
Failure to reset the counters each month in the GEM will affect the Total Energy column in the ISY. This impacts any programs that rely on the KWH total which you see in red. So if your intent is to track the total energy for that month for the furnace and have some kind of action performed by the Insteon network. This would cause a false flag to occur.
Not good if devices are going to be turned on / off.
One of the features I had requested was to have another Energy Total counter which recorded the daily KWH consumed. Opposed to what you see here a rolling counter that in essence is a global counter. To be fair I am uncertain if a individual channel counter can be reset or that a daily counter could be added.
But, then again this conversation hasn't progressed very far when asked.
PROBLEMS: SPIKES - UNKNOWN VALUES - FALSE FLAGS
So now we are getting to the meat of the problems I have been seeing and trying to determine the root cause. For the past several years I have been receiving alert messages that my solar array is outputting insane amounts of power.
These values sometimes are reported in the range of millions, and billions of watts. As of this writing I have not seen further spikes in this channel. Some of the things I did to be proactive was to include all listed companies to the alert message.
So, in essence they see what I see and receive the same notifications. The importance of doing this is having factual data to reference to. Nothing is better than receiving e-mails with time stamps along with the output from the system itself.
One of the things I have also done is limit the voltage range the system is monitoring. I am unsure if this has played a role in not receiving the alerts. Opposed to simply hiding the problem because two conditions must be met before the flag is set to true.
In this screen capture if it isn't obvious its 8:XX PM in the evening pitch black. There is no sun, the inverter's are asleep, the circuit has been disconnected from the grid.
I will be updating this test to see what bares fruit.
This same problem also appeared in my sump pump circuit. As you can see there is no way such a power output could have been achieved on any scale.
This is another example of the same issue on the sump pump.
To try to isolate and confirm what I was seeing the following things have been deployed and put into place. I have systematically placed monitoring equipment on key circuits that are showing these erratic and unknown values to appear.
The following steps have been performed and completed to help isolate, reduce, and stop these spikes.
GEM: PT Voltage has been measured and calibrated against a high quality true RMS DMM. CT parameters have been calibrated against known measured loads. CT's have been replaced for the circuits impacted by this. UPS and filtering systems are in place.
Additional line measurements via point of use CT's are in place to confirm and collaborate the primary circuit wattage.
The next problem I have seen is where the ISY has reported my refrigerator is in a negative wattage condition? as seen below
One of the things I did to confirm if this condition existed was to program the ISY to take the same data from the DB and record the value when the message was sent. As can be clearly seen in the message below the DB shows a power output of 111 watts, where as the ISY at the time of the alert indicates a power value of -136.63 watts?
The next issue is receiving alert messages that my dryer has been turned on / off. I have many programs that monitor and alert me to the current status of the appliances in the home. This helps the family members know when the next load is done and budget their time accordingly.
This clearly can't be done if the person is receiving a message at the wrong time due to a false flag condition. What is strange is that the program(s) absolutely work in terms of telling me when something is turned on, or is completed and goes into standby, or turns off.
The issue here is random as you see below, supposedly the dryer was done at 1:01:25 AM.
To further illustrate how impossible this could be is the shear fact the power to the breaker is turned off!
I have done this to several circuits and waited to see what happens. Of course it happens even with no power flowing to the device which clearly indicates this is not a power spike in that circuit.
The next problem appears to be the system at random times reports a low / high voltage condition when none exists. Over all this particular problem does not happen very often and other monitoring devices I have at my disposal never reflect this issue.
Finally, we have one of the most frustrating things I have encountered just in this last year. Below is a screen capture of the energy readings from the GEM to the ISY. Pretty much anytime now when I am in the Admin Console when I am sitting there the energy readings will simply just stop coming in and the system will stop updating the UI.
Being a long time manager, technician, and just a over all handy man in various industries. Its fair to say I have lots of common sense and technical know how to first trouble shoot, diagnose, repro, and deploy a fix.
I have from the ground up re-imaged ISY, my computers, replaced countless pieces of network related hardware, and combed the interwebs. To date I have not found any reason why the energy readings stop streaming into the ISY.
People have often asked me *Teken how do you know its simply not the GEM causing this delay in power values from streaming in??*
Well that is indeed a valid statement and question that needs to answered. I know its not the GEM because the same data is streaming into other devices in my network and to the cloud. For the sake of eliminating possible things that could be a contributing factor.
I have systematically removed every device from the GEM and only allow the ISY to communicate. The problem persists with this set up. I have replaced the Zigbee card, brought the entire device one foot from the ISY to GEM module, same problem.
I have disconnected every piece of electronic gear in my entire house that could possible cause any RFI/EMI that could cause interference, same problem.
I have even brought in a RFI / EMI meter to confirm my home is completely dead and void of any interference, same problem. Multiple computers spanning four different work sites using Windows XP, Vista, Windows 7, Windows 8, all show the same problem.
Going into the Dash Box, SEG, and a local Data Server all show no dropped COM's or missing data, ever.
Fast forward to November 2014 even with every firmware update this problem persists??
So, why did I just state all of the above with such clarity and exactness? Because its not just the energy readings that stop coming in. Its the entire ISY UI stops responding to user interaction or commands.
Reviewing the Level 3 Error Logs show the system simply stops. Why? I don't know why and have removed every program to see if the UI freeze stops, it doesn't. So, when people ask what have you done? Well, pretty much everything that can be done and more.
More details to come . . .
APRIL 02, 2015 UPDATE:
Well it looks like after what seemed to be eternity many of the problems I have been seeing have on the surface been resolved!
Paul at Brultech and Michel at UDI have finally determined how Zigbee packets coming from the GEM and how they are handled and translated in the ISY Series Controller. Have on the surface been able to resolve two issues.
OUT OF BAND WATTAGE:
Alpha 4.2.29 firmware was installed on March 30, 2015. Since that date no out of band values have been reported by the ISY Series Controller via e-mail alerts. These out of band alerts impacted voltage, solar, fridge, dryer, sump, and dish washer.
It has been two solid days and nothing has come in or has been recorded.
ADMIN CONSOLE CONNECTION:
Another apparent side of effect of patching up the Energy Module has been the restore of being able to stay connected to the Admin Console. This issue was two fold in that in the past I was able to stay logged in and review live events of status and energy readings for an entire day.
Prior to Alpha 4.2.29 I was hard pressed to stay connected from seconds to minutes never mind hours!
I have been able to stay connected and view streaming energy readings in an excess of three plus hours with out a single drop out. This aspect can not be understated because when a person is coding and looking for a change of state in the programs and nothing is changing.
This leads to confusion and a lack of confidence of what is actually happening in the system as a whole. So the connection appears to be more solid and streaming energy readings in the Admin Console are timely and accurate.
VARIABLE MISCOUNTS:
Some where down the line of various Beta updates the system has started to miscount. I have hundreds of state / integer variables that track and count when things are on, off, duration, cycles, and run time.
For what ever reason (loosely speaking) the system ignored extremely small wattage changes. But now even the most minute drop in wattage triggers a change in state. This unknown sensitivity was the root cause of my variable miscounts!
Below is a perfect example how powerful the DB / GEM combination is. Using the *Real Time Data* option this is a instance where the fridge dropped off for just a few seconds.
Zooming in and expanding the time interval for this session clearly shows the (average watts) being well above my set threshold of 105 watts. In this case the image capture shows the fridge overall was consuming between what I have come to expect from 109 - 111 watts.
Which is obviously impacted by voltage rise / fall.
As we look closer its apparent during the fridge cycle the appliance dropped off to 104 watts. This in effect caused the preset threshold of the fridge monitor program to restart and add in another cycle to the daily count!
I have since recalibrated my programs with a lower starting wattage to ensure any dips in line voltage or power draw does not cause this miscount to happen again. I have to thank once again Ben at Brultech for helping me and also pointing out this issue.
Having the ability to monitor, track, capture, and recall live events as they occur is simply amazing to me. Its safe to say Brultech has pushed the envelope of their product and offered the end user one of the most versatile and powerful data loggers in the open market.
With out Brultechs help and insight I may have never found what was causing all of these miscounts which in the big picture made me lose confidence in the ISY Series Controller.
With this knowledge and insight at hand I can continue to redeploy my Home Automation project with more confidence.