Open givecoinsupport opened 11 months ago
Can you please explain better where do you call COMCHECK
against and NEWSCAN
?
@Matthew1471 maybe you know how to achieve that the inverters are reporting every minute instead of every 5 minutes?
Apologies for the delay, I've been upgrading my Enphase to another 7 more panels (and experiencing PLC incompatibility with Tesla PowerWall 2).
Have a look at these lines, I previously noticed calling the communication check would trigger that.. I haven't tried running this recently however. I know I say it a lot but the examples are a really good place to check for API usage.
Given it's in /installer/
usual caveats apply on the token role type. Would prefer to not say this too loudly (hopefully burying this in an issue should be discrete enough) but Enphase don't actually strictly enforce who is an installer or not, so it isn't too difficult to use a different e-mail address, register yourself as an installer and then as the owner "invite" your new "installer" to also have access... that's how most of us have access to the installer
functionality.
A new scan
is a PLC scan, basically a broadcast request for provisioning.. and probably inadvisable to run routinely (stress on the box, noise on the wire and the potential to accidentally provision your neighbour's microinverters!). I'll do some more digging on what other things trigger a PCU / inverter comm check, there might be a path from an owner
token but I'd need to spend some more time reading the source and doing a Find In Files.
Enphase themselves moved over to CT clamps and meters for production stats (Enlighten lets installers choose where the measurements come from but CT Clamps are much more preferred), the 5 minute granularity is usually appropriate for the panel by panel breakdown (Enlighten itself is every 15 minutes don't forget!) but for second by second power measurement the IQ Gateway Metered with properly configured CT Clamps are the preferred way to go. Again, in the examples folder is some code to get those meter readings.
I'm already using the CT clamps for measuring. I would only need it for some specific situations as I'm discussing since over 6 months with Enphase for a bug in the microinverter firmware and they continue to not understand me and to don't believe me. The only thing they do is testing different grid profiles for 48 hours, since months...
I have no shading and all panels are exposed to exactly the same amount of sun, still a lot of panels do not produce correctly in a micro grid.
Will test that, since I already have the needed rights. Thanks!
I assume you have also performed a PLC noise test and checked to see if the panels are producing but failing to report (the panels produce so long as they have a grid profile and favourable grid conditions, they don't really need the IQ Gateway - the IQ Gateway and it's PLC chatter is just for data)? That's the issue I have at the moment, et al this person. 7 panels on the same cabling as the Powerwall and its heavy duty switching inverter get drowned out by all the Powerwall (EMI? EMF?) noise on the wire. Enphase recommend filtering it with a line filter or trying to chuck some ferrite cores on the Powerwall AC to absorb its noise.
I haven't delved that deeply into the panel reporting (although I was excited to find they report their individual temperatures too).. the other thing to check might be the event logs in the installer portal, if they're shutting down due to something they don't like they should issue a dying gasp and the gateway should generate an event in the log (it will mark it as resolved but you can still go into either System Dashboard or unfiltered Homeowner Events to view them). If they fail to report then the IQ Gateway itself will issue a "failed to report" message for every microinverter that doesn't check in.
I haven't attempted to look for microinverter uptime stats... I'll try to improve the documentation when I have more time and hopefully we can figure something out. Hope the comm_check trick works for you in the meantime!
Unfortunately the error/status reporting of Enphase sucks. It always only says DC power too low
. Even if there is full sun. There is no noise with the Victron Energy MultiPlus, but I still added some ferrite cores.
The problem raises after the inverters go over 50 W in the morning. If then I turn off power production for 60 seconds and turn it back on, all is working as it should.
/installer/pcu_comm_check
calls commcheck.lua
on the IQ Gateway with --installer
as one of the command line arguments which is documented in the script as Flip the installer flag for back-to-back polling for 60 secs
. I couldn't find any other meaningful calls to this script apart from maybe /ivp/ensemble/comm_check
.
The commcheck.lua
itself is reliant on /ivp/peb/pmi/app/ssi
which I believe is a native process and also locked down to roles higher than owner
so you'd still need an installer
token in practice anyway but potentially could shave some time off the calls... although you are running pretty low level close to the PLC modem then (potential for bricking goes up).
There is also /ivp/peb/pollperiod
which may allow you to change how frequently they poll - but I'm not sure changing the internals of the IQ Gateway would be wise if you're trying to get them to support the device - they could argue the issue is caused by non-standard config.
Sticking to the standard installer friendly /installer/pcu_comm_check
is the way to go I suspect. There won't be any log lines suggesting you've run any more esoteric queries then that could come back to haunt you.
Unfortunately the error/status reporting of Enphase sucks. It always only says
DC power too low
. Even if there is full sun. There is no noise with the Victron Energy MultiPlus, but I still added some ferrite cores.
Ferrite cores apparently dampen the AC noise but they won't eliminate it and if the noise is that high might be too much. Do you have an export of your PLC noise scan from either Toolkit or System Dashboard to work out if PLC AC noise is low/medium/high. For me simple debugging was turning off my batteries and doing a PLC noise scan showing the before/after (not that I needed to as the microinverter comms red lights instantly went green on the IQ Gateway then anyway). Along with the "microinverters failing to report" message self clearing as I turn off the Powerwall batteries.
DC Low doesn't sound AC grid related tbh anyway, that will be solar panel to microinverter connection.. the panels are supported for the microinverters (using Enphase compatibility calculator online?)? There's some DC params that the inverters need that are beyond my electrical skill.
DC Power low will occur quite naturally early in morning and late at night as you'd expect.. I'm not an electrician but I'd be looking to see if I could measure the DC voltage/current on those wires to rule out if the panels are having a wobble or internal short or something that is causing DC power loss to the PCU (Power Conditioning Unit) / Microinverter. I almost wonder if you're able to take a good working panel and give that to the microinverters that are crashing out?
The problem raises after the inverters go over 50 W in the morning. If then I turn off power production for 60 seconds and turn it back on, all is working as it should.
Is this some of the panels or all of them? Are there some good panels and some bad panels? Have the panel makers been much help?
There is no issue of the panels. Just the message DC power too low
is wrong. It also appears, if I have full sun, grid connected and the inverters are throttled. There is no message like power reduced because of zero feed in selected
, power limited because of frequency shift
or anything else. Deep troubleshooting is not possible. Also the Enphase support is quite primitive with the tools they have to analyze a problem.
An Enphase engineer confirmed me that a service in the microinverter gets stuck. Therefore the power freezes to the current value and cant go up or the power goes to zero. Only a reboot of the inverter fixes the problem. There is no hardware issue, just a firmware bug they don't believe in.
Sometimes 12 of 16 inverters have the issue, sometimes only 8. Mostly at least 8 of 16 have the issue. Every time other ones. This also confirms me that this is firmware issue. I also tested all panels, they are all brand new and fine.
I'm using the IQ7A, just in case someone likes to know.
Can you please explain better where do you call
COMCHECK
against andNEWSCAN
?
in D5 Envoy firmware, when access(reading/get) http://envoy.local/installer/pcu_comm_check, I guess, envoy broadcast a command to report the status by inverters. so, just access it every minute, can get 1 minute data.
in D7 Envoy firmware, I guess, no more broadcast a command with /pcu_comm_check anymore. when NewScanPD http://enovy.local/ivp/peb/newscan, I guess, enovy broadcast a command to report the status by inverter. if Enable NEWSCAN every minute, can get the status of inverters in 1 minute. ( I have verified in Firefox - Just sending the packet again to force to update data from inverters)
You need 'INSTALLER TOKEN' to do this job. When I get a token using "https://enlighten.enphaseenergy.com/login/login.json?", I just can get an owner's token. When I try OAuth method on "https://entrez.enphaseenergy.com/login" to get an installer token , I can login and can get a session id, but, can not get a token using this session_id. it makes an invalid format/data error.
PUT method for NEWSCAN has explained in Mattew's API. and, just need 3 parameters to enable it
duration-in-minutes devices-to-be-discovered nsrb-to-be-discovered
[Edited]
http://www.bomijoa.com/solar/images/chart_panel_30.csv http://www.bomijoa.com/solar/?size=1&debug=1&detail=0&offset=0
Hi, How are you ? Thank you for the great effort. Just Let me share a small top to force to refresh report from Inverters for 1 minute interval.
In the old Local API, if call, COMCHECK. it can refresh report time to 1min.
BUT, in new API (v7) COMCHECK method is not working, have to use PUT NEWSCAN to force to refresh Inverter Data Report. and, can get 1 minute resolution.
if using Enphase's document to get a token with 'user[email]', only can get an owner token. I am trying to find a way to get a installer token to get installer access