Thanks for this information. It’s disappointing to see that the Envoy itself is only polling every 15 minutes. I was hoping that they would keep the 5 minute interval at the Envoy and then condense it to 15 minutes for storage on their pathetic overloaded servers.
I’d be interested in how your conference call with Enphase goes.
The reporting interval in Australia is 15 minutes. In the US in recent times it was 5 then it went to 15 and now I believe it is back to 5 minute intervals. In Envoy it samples every 60 seconds.
Things seem to be in flux. Several months ago, my reporting switched to 15 minute intervals giving a “stairstep” graph with identical power values for each of three 5 minute reports then jumping to a new value for the next three values.
At a later time the reporting switched to a “sawtooth” graph with identical values for 2 of the 5 minute reports then a lower value for the third before jumping back up for the next 15 minutes.
I note that just today, it is reporting different values for each of the 5 minute intervals. However, I still have a “sawtooth” graph as every third report for power drops down below the previous and subsequent values.
Here is a sample of my data for today. Power is the last column and you can see, for example, the sawtooth in the first five values where reported power goes from 3577 down to 3514 then up to 3790, up to 3969 then down to 3911.
27/09/17 8:45AM 4.447kWh 0.713kWh/kW 3,911W
27/09/17 8:40AM 4.121kWh 0.660kWh/kW 3,969W
27/09/17 8:35AM 3.790kWh 0.607kWh/kW 3,790W
27/09/17 8:30AM 3.474kWh 0.557kWh/kW 3,514W
27/09/17 8:25AM 3.181kWh 0.510kWh/kW 3,577W
27/09/17 8:20AM 2.883kWh 0.462kWh/kW 3,367W
27/09/17 8:15AM 2.602kWh 0.417kWh/kW 3,093W
27/09/17 8:10AM 2.344kWh 0.376kWh/kW 3,138W
27/09/17 8:05AM 2.083kWh 0.334kWh/kW 2,994W
27/09/17 8:00AM 1.833kWh 0.294kWh/kW 2,780W
27/09/17 7:55AM 1.601kWh 0.257kWh/kW 2,821W
Well… it’s even more screwed up than I thought.
I decided to log the power reported on the local Envoy web page and compare that to the MyEnlighten web site and to the values that PVoutput pulls from Enphase.
The result… they are all different numbers, not even close!
For example:
between 13:00 and 13:15 Envoy reported 4.31 kW, MyEnlighten reported 0.1 kW(!yes) and PVOutput reported 4.30, 3.32, 3.49 kW
The Envoy web page reports the exact same value for each 15 minute interval.
Seems to be a random number generator once it gets to Enphase. I don’t know who to trust.
I guess I’ll have to build my own monitoring system.
Within the 15 minute time interval the system is generating watts at predetermined intervals I believe those intervals to be 60 second. The question is how is Enphase aggregating those 60 second values and what formula is it using to convert it to Wh? The other issue is the question of what precision is being applied to the calculations.
Whilst this does not apply to Wh or kWh conversion of kW just to share a methodology based on my experience in the world of finance in the calculation of interest on an investment. The first step it was performed to the maximum precision allowed by the operating system which was then truncated from the 4th decimal point and rounded 5/4 to 2nd decimal point before credited to the clients account. So in the calculation of Interest calculated in the first step $5.4585614326 becomes $5.4585 becomes $5.46.
What I have not defined is the calculation of the interval rate for the annual percentage rate applied to values where the interest rest may occur at intervals other than annually.
Yesterday For Energy Used PVOuput reported 25.982 kWh Monthly Net Energy Report [MyEnighten] 25.52 kWh and for Energy Generated 32.129 kWh compared with 31.72 kWh. In my view the difference is not material and could be related to any number of data processing strategies which are beyond the control of the end user.
And if you round the numbers up to an integer for Energy Used for both reporting systems you get 26 kWh and for Energy Generated 32 kWh.
I have checked the Monthly Net Energy Report with that reported in PVOutlook and it is my opinion that because the data for individual 15 minute time intervals does not accord then what Enphase is reporting to PVOutlook is not exactly the same value for the time interval that it is using to report in MyEnlighten. The real test is that there is no material difference in the values reported by both systems which after all could be attributed to rounding at the end of day.
You might find this data smoothing technique helpful. So is Enphase pre-processing the data to be graphed? http://www.itl.nist.gov/div898/handbook/pmc/section4/pmc42.htm
Hi mspohr, well the telcon that was to be initiated by my installer never happened. I called them around 2:20PM to see approximately what time they were planning on making the phone call as it was getting quite late in the day. I was informed that they were very busy with other priority issues mostly those brought about by hurricane Irma. I was told that they would notify me within 1 hour of when they intended on making the phone call which would not be on that Monday and they had no idea as to just when the date of the conference call would be. Additionally, the wording of just how the “conference call” was going to take place changed from the original method which was that I would be listening in the background to the entire call…the new method is that I will be conferenced in IF Enphase needs additional support info from me. By the way I neglected to mention that my supplier is FAFCOSOLAR Energy located in Cape Coral, Florida. I pass this information along just so any other customers or poternial customers of FAFCOSOLAR who happen to be reading this Forum will be aware of the customer support you will most likely receive AFTER you have paid for your system!
I will continue to update the status of my customer support on regular intervals on this site.
Just a follow up on yesterdays comparison of Envoy, MyEnlighten and PVOutput.
As I reported yesterday, the Envoy internal web page reports the exact same value for each 15 minute interval. It may be collecting data from the inverters more frequently but the web page doesn’t change for 15 minutes. The MyEnlighten “Recent Production” reports data at 5 minute intervals and this follows the “sawtooth” 5min/15min pattern that is reported to PVOutput.
The MyEnlighten report also clearly shows the sawtooth with every third value dropping down before jumping back up for the next interval.
There is clearly some data manipulation taking place on the Enphase servers beyond what is reported on the Envoy web page. There also seems to be a 5 minute time shift.
(PVOutput mirrors the MyEnlighten values)
So, you can’t trust the Envoy and/or you can’t trust MyEnlighten to give accurate values.
Of course, Enphase could clear all of this up with a little communication but I guess that’s too much to ask.
Here’s an example:
MyEnlighten report Envoy
2017-09-27 13:00:00 -0700 4305 4.31
2017-09-27 13:05:00 -0700 3322 4.31
2017-09-27 13:10:00 -0700 3349 4.31
2017-09-27 13:15:00 -0700 3349 3.35
2017-09-27 13:20:00 -0700 2209 3.35
2017-09-27 13:25:00 -0700 2153 3.35
2017-09-27 13:30:00 -0700 2153 2.15
2017-09-27 13:35:00 -0700 2250 2.15
2017-09-27 13:40:00 -0700 2349 2.15
2017-09-27 13:45:00 -0700 2349 2.35
2017-09-27 13:50:00 -0700 2748 2.35
2017-09-27 13:55:00 -0700 2923 2.35
2017-09-27 14:00:00 -0700 2923 2.92
2017-09-27 14:05:00 -0700 2111 2.92
2017-09-27 14:10:00 -0700 2085 2.92
2017-09-27 14:15:00 -0700 2085 2.08
2017-09-27 14:20:00 -0700 1880 2.08
2017-09-27 14:25:00 -0700 1916 2.08
2017-09-27 14:30:00 -0700 1916 1.92
As soon as my PowerWall 2 can report to pvoutput I think I’ll turn off the envoy and see if Enphase notices 
Then I’ll turn it on periodically to confirm panels and microinverters are working correctly.
Well, an additional week has gone by and nothing heard from FAFCO Solar regarding my conference call with Enphase. I guess hurricane Irma must have wreaked havoc with all the FAFCO Solar installations…and by the time Irma got into this area she was downgraded to a Cat 3 hurricane (120mph max)…doesn’t bode well for the installations (FAFCO told me their installations were rated to 160mph) if that is what is causing the delaly in my telcon (so I’ve been told). I will continue to update this forum with the progress (or lack thereof) on a regular basis.
It may be projectile damage. The panels will probably stay attached to the roof in winds up to 160 mph, but flying debris can wreak havoc on the panels.
Hi johnwelch…well I sure would like to believe that it was something like projectile damage that has put me at the “bottom of the list” as far as support is concerned…however another week has gone by and I haven’t had any additional response from FAFCO Solar regarding my telecon with Enphase to discuss the modification of my Envoy firmware by Enphase that has changed my Envoy sampling rate from its previous 5 minute sampling rate to its new 15 minute internal sampling rate.
1 Like
Enphase is in California. No projectile damage here.
They just don’t offer support… anywhere, anytime, to anybody.
1 Like
I finally heard back from my installer this week. They did make contact with Enphase and as might be expected, I was not invited to participate in the telcon. Enphase informed FAFCO Solar that they did push a firmware modification into my Envoy which did change the internal sampling time from 5 minutes to 15 minutes and they would under NO circumstance be willing to reset it back to the original 5 minute sampling. So much for any support coming from Enphase.
So based upon all of the fine work done by so many of the contributors here on this forum, I find that these unwelcome modifications by Enphase to my solar system leave me with one option. In order to protect myself from the “NEXT” modification I have isolated my Envoy from the internet. Additionally this action will protect me from hackers that can take advantage of the VPN tunneling method that Enphase uses to transfer data to their servers. As it has already been pointed out…one good hack into the Enphase server and ALL Envoy users could be reading in the news about how their personal information had been made public.
In the past I have looked at a few of the hacks to tap into the Enphase data from the Envoy.
http://forums.hackaday.com/viewtopic.php?f=10&t=4324
http://blog.oddbit.com/2012/02/13/enphase-envoy-xml-data-format/
http://blog.oddbit.com/2012/02/22/capturing-envoy-data/
This latest crippling and corruption of the Envoy data might be a good impetus to look at this project again. It would be nice to capture the raw data stream from the panels.
The Enphase data anomalies continue.
Just noticed that for several days in the past week, Enphase is reporting solar production during the night! ?? It will report a steady 500 or 700 watts starting about 2 or 3 am through to the morning when the sun actually comes up and generates power.
I also noted that if I look at the previous day, reported power goes to zero in the middle of the afternoon when there is still power being generated.
It seems to me that Enphase is missing power in the afternoon but tries to compensate by reporting power much later that night. (The MyEnlighten site has the same anomalies)
I am part way through writing a windows service that pulls data from the Envoy-S and pushes it to pvoutput. If you have a pc you could leave on and the Envoy-S I could send you a copy to try and maybe you will be able to debug your issues. I haven’t run the service on anyone else setup so you may have to be patient while I fix any bugs in it.
My service can output the the inverter readings and meter readings to separate pvoutput systems
inverters https://pvoutput.org/intraday.jsp?id=36097&sid=56015 (inverters individually in extended data)
meters https://pvoutput.org/intraday.jsp?id=36097&sid=56468 (consumption and generation meters for different phases in extended data)
This sounds great. It would be good to have individual inverter data.
The problems I see are that I have an Envoy-C, not Envoy-S and I don’t have a Windows PC running. I could set one up dedicated to this project but the Envoy-C might be a problem.
Thanks for you help.
Hmm. Not sure if my code will work with the older Envoy.
If you are willing to find out try this.
In web browser from a computer connected to the same LAN as your envoy go to http://envoy.local from that page write down your Envoy serial number.
Then go to http://envoy.local/api/v1/production/inverters you should be prompted for a username and password. “envoy” should work as username then the last 6 numbers of the serial as the password.
Hopefully this will return your inverter spot data in JSON format. If not then sorry my software wont work with the older envoy
Thanks, I’ll give it a try.
I’m on vacation for a week so don’t have local access. I’ll give it a try when I get back