diy solar

diy solar

An Enphase Ensemble Installation

... You must have been on the waiting list with deposit/pre-order months ago. Surprised the installer allowed the AC to be on the backup circuit!

I think it was more that I lucked out as one of the areas the Enphase wanted to deploy in first. I don't have a critical-circuits, I control what's on/off by their switches or the load center....you can read more on that here. If I pull to many amps it resets; no waiver needed. Someday I'll get the automation together....so much to do so little money.... <sigh>
 
Last edited:
'The warranty/internet connection seems odd to me. So is enphase telling me I can operate it in a way to harm it or will enphase adjust charge/discharge rates to conform as needed to keep warranty terms intact.

I bet it's carbon trading. I disconnected mine from the net. They just sell Carbon credits to polluters.
 
Warranty
What does the current U.S. warranty say? Have to look that up... it's that after 10 years or 4,000 cycles it'll have > 70% capacity provided the ambient temperature is neither to cold or to hot.

.. So is enphase telling me I can operate it in a way to harm it...
That's correct. The warranty is very specific about the expected ambient temperatures. No real surprise, operating at high temperatures degrades them, charging at low temperatures can destroy them.

...'The warranty/internet connection seems odd to me...
I think you're thinking Tesla's Powerwall. I've heard that even if you're offline for too long a period it'll void the warranty because it only stores so much ambient temperature data before it needs to overwrite. Enphase probably learned from that and has enough internal storage for its lifetime.

As far as I know, there's nothing about mandatory connectivity in the current Encharge warranty; if there was I'd expect the number of incommunicado days to be clearly spelled out. That's not to say the Enphase system doesn't send a ton of data back to their HQ, it does. I know they make that data available for researchers to study; but they could be selling it too. I'm okay with that, it keeps those researchers off the streets and from getting into trouble (idle hands and all that). Hopefully Enphase is learning from that too and making better firmware for it.

... will enphase adjust charge/discharge rates to conform as needed to keep warranty terms intact.
The only time the word adjustment is used in the warranty is the transfer fee, so it doesn't say anything about it.

There's nothing they can do to adjust the discharge rate, if it didn't meet the minimum spec in the datsheet then it could be considered defective under the terms of the warranty. They could affect the charge rate, and if it lengthened the life of the battery I'm all for it.

I'm pretty sure the batteries have unreported capacity as they'll last longer if you don't fully charge/discharge them. So, by altering those settings that headroom could shrink over time to ensure the batteries met the warranty conditions. As the battery would still be performing as advertised that seems fair.
 
Last edited:
Weird Service Entrance, Enphase tie-in, & Why there are no Critical Circuits here

Once upon a time there was a cozy little house here. But the owners thought it too small so they added on to the west, doubling the size. So, the house has one service entrance, one meter. From there it splits into two disconnects, one for the west and for the east located at the meter. From there it goes to a normal load center, err... except there are two of them (not a sub-panel). It was probably cheaper and more expedient to do it that way. When I moved in, it looked like this:
1598012053491.png
The base power draw of the house is under 1kW with the PCs, lights, fans, frig, etc. turned on. It's the air conditioners that consume the bulk of the power. So, my choice was to get a battery with enough kWh to last the night with just the fridge and a small mini-split on, and enough inverter to be able to start the AC (with Soft starter), so that on a sunny day I could power everything and recharge the battery. Or at least not have to reset my clocks for every power blip where the AC came on. I'll up capacity with EVs when V2H, V2G, or VPP happen (Stationary! We don't need no stinking stationary batteries!).

We don't need no stinking Critical Circuits
I never liked the idea of a critical circuits panel. They're nice in that you can ensure that the inverter can't be overloaded. But if I hear a commotion outside, I want to flip on the outside lights and see what's going on. Maybe I want to use the microwave, or see the news on the tv, or whatever. Why shouldn't I get to choose? This is America after all!

With overload protection, an inverter just stops working without doing any damage. It takes a minute for them to reset and you have to reset your clocks, but that's about it. Now that I have the soft starter, the main AC coming on doesn't cause the inverter to reset. So when the power goes out, no problems. Assuming I know about it (see foghorn), I can turn off stuff to conserve power.
The oft abused quote comes from the classic Treasure of the Sierra Madre
1598013123710.png
I suspect Enphase feels the same way since they normally recommend feeding directly into your load center. With two load centers and needing to disconnect from the grid, that wasn't in the cards. Even if I had wanted a critical circuits subpanel it would have been hard to do, the wires run in conduits embedded in concrete.

That was then, this is now
1598019044246.png1598017157586.png

One advantage to a supply side tap is it's not subject to the 120% rule as nothing can be overloaded. Behind the Encharge 10 you can just see the edge of the Encharge 3. Need to paint the conduit the house color so they don't stick out so much... it's an eye-sore, now you know why I started the camouflage thread.
 
Last edited:
Just got a good test in on the soft starter. As you may recall the device normally works fantastically, but they sent me a firmware update for the IQ8s shifting the frequency when curtailing power. It's a moving target because as the AC starts drawing power it causes the IQ8s to first shift back towards 60 Hz and then back up to curtail the solar again.

The weather here is mostly cloudy today with the odd patch of pure sunshine, so testing at full power to get a lot of frequency shifting has been hard. Usually by the time I run down and flip the power off, run upstairs to turn on the AC a cloud has come overhead.

I finally got smart and downloaded a light meter app for my cell phone. Then I left the phone is one orientation and watched the light meter's reading to the panel outputs. For the orientation/reflectivity ~300 lux was 6 kW. So, I waited until it looked like I might get a break, then flipped to battery power with the AC off and waited. When the lux hit 330 I turned on the AC, victory! It's clouded up again now... down to 2.6 kW.
 
Just got a good test in on the soft starter. As you may recall the device normally works fantastically, but they sent me a firmware update for the IQ8s shifting the frequency when curtailing power. It's a moving target because as the AC starts drawing power it causes the IQ8s to first shift back towards 60 Hz and then back up to curtail the solar

So the solar is curtailed when batteries are full, there is less load than solar is producing, and you are off-grid. Do I have that right? If you’re on-grid the power would be sent back to the grid with no need to curtail. So ideally it’s like an EV. You don’t want to leave home with a full battery when you have a 10 mile down hill commute. You always want some room in the battery for excess generation/regenerative braking.
 
So the solar is curtailed when batteries are full, there is less load than solar is producing, and you are off-grid. Do I have that right?
Solar is throttled back when the frequency is shifted. Everything else in that sentence is good logic for when to shift the frequency and by how much.

But there are other times to curtail power too. For example, California, Hawaii, Canada, Puerto Rico all currently have rules for how grid operators can use frequency shifting to curtail renewable energy from residential systems. Enphase is probably just making use of the same technique for off-grid control. For example, CA Rule 21 has this chart (ref) which shows
1598045596357.png
Ideally grid operators can also request additional power by decreasing the frequency. We'll probably be hearing more about that in the not to distant future (see speculation in Battery Day).
 
I thought I might have found an envoy api call to get the Encharge charge/discharge rate: GET http:/envoy.local/ivp/ensemble/inventory, here's a shortened response with the key data, the interesting value looks to be real_power_w, possibly the number of watts charging (negative) or discharging. Sadly the discharging the value of real_power_w is 0; although the charging data may or may not be accurate. (see post #128 for update).
JSON:
[
    {
        "type": "ENCHARGE",
        "devices": [
            {
                "device_status": ["envoy.global.ok"],
                "last_rpt_date": 1598198575,
                "admin_state_str": "ENCHG_STATE_READY",
                "percentFull": 95,
                "temperature": 33,
                "maxCellTemp": 34,
                "real_power_w": -32,
                "dc_switch_off": false
            },...
    },
    {
        "type": "ENPOWER",
        "devices": [
            {
                "device_status": ["envoy.global.ok"],
                "last_rpt_date": 1598198638,
                "admin_state": 24,
                "admin_state_str": "ENPWR_STATE_OPER_CLOSED",
                "temperature": 97,
                "mains_admin_state": "closed",
                "mains_oper_state": "closed",
                "Enpwr_grid_mode": "multimode-ongrid",
                "Enchg_grid_mode": "multimode-ongrid",
            }
        ]
    },
]
Where when off-grid Enpower changes to:
JSON:
    {

        "type": "ENPOWER",
        "devices": [
            {
                "admin_state": 21,
                "admin_state_str": "ENPWR_STATE_OPER_NO_GRID",
                "temperature": 98,
                "mains_admin_state": "closed",
                "mains_oper_state": "open",
                "Enpwr_grid_mode": "multimode-offgrid",
                "Enchg_grid_mode": "multimode-offgrid",
            }
        ]
    }

While switching back to the grid this state changed: "mains_oper_state": "open synchronizing" before "closed".

Also found GET http://envoy.local/ivp/ensemble/status which provides the Enpower state and the SoC in watts via agg_backup_energy.

JSON:
    "secctrl": {
        "freq_bias_hz": 1.3200000524520874023,
        "voltage_bias_v": 16,
        "freq_bias_hz_q8": 2123,
        "voltage_bias_v_q5": 512,
        "configured_backup_soc": 95,
        "adjusted_backup_soc": 95,
        "agg_soc": 95,
        "agg_backup_energy": 4750,
        "agg_avail_energy": 4750
    },
Interestingly, only 3.36 kW are "available"... sounds like they're running 90 to to 20% SoC.
Updated post #17 with the new finds.

Update: Curious... Post #17 already had the first api call... it was updated in post #100... so this is a rediscovery ;-). The real_power_w appears new. A firmware upgrade? Nope... Software version is now D6.0.80 (24c7d7) built on 7/8/20; but post #100 was from 8/4. Must have missed it on the first go round.
 
Last edited:
Seems odd that you have temps in F and C in the same Enphase system. I have C in my BMS and F on the Schneider now. My air tamp is showing in F on the Enlighten. I'll have to pull status from the iQ7's and see what unit they report.
 
....Prior to installation my in-rush meter read 86.4 amps on the compressor C wire...
I had a 22 amp start today during some tests I was running, lowest so far! Most are 26-32 as there's a spike during the wave form. MicroAir said they thought they could get it down to around that consistently with an update they're working on.
 
Turns out I was just impatient in post #108. While running some tests off-grid this morning before the sun came up real_power_w did go positive and was representative of the current drain in watts: "real_power_w": 634,

But it's only getting updated about every 5 min. When you look at the "live status" screen, the field is getting updated about every 5s. So, there's a better source yet somewhere.

The other take-away is that agg_avail_energy appears to be the remaining power whereas agg_backup_energy appears to be the capacity.

JSON:
    "secctrl": {
         ...
        "agg_backup_energy": 4750,
        "agg_avail_energy": 4650
    },
 
Last edited:
This is interesting... from above we have:

JSON:
    "secctrl": {
         ...
       "configured_backup_soc": 95,
        "agg_backup_energy": 4750,
        "agg_avail_energy": 4650
    },

The max power is set by me to 95% currently. If I take 4750 / .95 =5000. Which is a suspiciously round number. The datasheet for the Encharge 3 says it has 3.5 kWh Total capacity and 3.36 Usable capacity. As 3.5/5.0 =70% what's probably going on is that they're taking 30% off the true total capacity to operate at an SoC between 90 and 20%. The 3.36 is probably some accounting for a 4% loss from inverter/battery inefficiency.
 
Thanks again for your insights. Apologies if you already covered this...
In off-grid mode:
How well does enpower transition between solar and battery?
What is the max load you have seen run by your solar-only in relation to its nameplate (AC) capacity?
I’m aware that environmental factors(clouds,temp) are a factor here.

Thanks, WL
 
Updated FogHorn to show the charge/discharge rate and time remaining to full charge or depletion.
Sun's out and currently making more power then I can use... so I could only test the battery discharging case in the debugger...
1598364566489.png
 
...How well does enpower transition between solar and battery?...
Like a UPS, no lights flicker, no PCs reboot, no clocks need resetting. The enpower is fairly slow to switch back to grid (5 min?), but again no issues.
That's assuming you're within the load. One of the reasons I started looking into soft starters was a momentary power loss causes the Enpower to switch over to battery and if the AC tried to start while on battery it reset. Then I was without power for 5 min while the Enpower transitioned back to the grid.

...What is the max load you have seen run by your solar-only...
I've run the house (mainly PCs, fans, dehumidifier, hot water, mini-split, fridge, freezer... varies upto 3 kW depending on what's running, but is usually around 1.2 kW), AC (2 kW) and charged the battery simultaneously while off-grid from solar.

Watts is watts, so load isn't as important as generation and what the Encharge can switch/handle while curtailing power.

The array is 7.8 kW of 340W panels and the inverters are rated for 295W ...but I typically see 300W off them, so 6900W is the most I can expect from it. A good day is over 50kWh, but I've seen peaks at 6.8 kW, that is to say I've rarely seen clipping (more here). Considering my suboptimal tilt that's not to bad.

It looks like the Encharge can hold it's rated surge while adjusting the solar panels. When it was resetting I was pulling 86 amps (the Encharge setup here can surge to 31 amps). Throw in the max solar and that's another ~28 amps.
 
Last edited:
Now that I have storage "working" I think I will add the extra 4 panels. I didn't like the idea of pushing more out to the grid, but with a place to put the power, it just gives me more power all day, especially if I have to go off grid during a failure.

The Ensemble is certainly nicely integrated, and the total system operation is looking pretty good. And the best part seems to be that Enphase is working out the issues. I have not gotten any response from an Schneider engineer yet.

The maximum power I have pulled out of the XW so far is only 2KW, and I have not topped 2KW of charge power either.

Once I have my loads move over, this will change a bit. But getting the charge current to follow my over production is going to be a real trick. That is one place the all Enphase setup should do much better.
 
Had the FLiR out... so why not...?
Ambient= 90°F, Pressure= 29.95 inHg, Wind Speed= 2.0 mph ESE, Humidity= 79%
Side of IQ8 in Encharge 3
1598373747649.png
Front of the Encharge 3:
1598373799320.png
Enpower
1598373900053.png
Service Entrance:
1598373960709.png
Junction Box, IQCombiner, Enpower
1598374024357.png
Under the panel, IQ7 (The hot spot is 139°F)
1598374163386.png
I wonder it it's lack of air-flow above the IQ7?
Western facing
1598374324406.png
East Facing
1598374473165.png
EasyStart
1598374565542.png
Whew! Is my neighbor's roof hot or what?
 
This is a blog post by Enphase detailing Power Start. A feature enabling the starting of high inrush current loads. Scroll down to view a chart detailing how many batteries are typically required to power certain loads and the accompanying soft starter that will help decrease initial start up load.
 
Thanks @WildLeg! The part numbers match up to Hyper Engineering soft starters. Oh duh... it even calls Hyper Engineering out in the article.
Hmmm, those devices only have a 1 year warranty. They're physically smaller which is nice for crowded condenser interiors. Their numbers might be conservative, for their 3 ton they have 75 amps before and 23 amps after. I was seeing 86.4 amps before and the after value still varies a bit, between 22 to 26 amps .
 
Back
Top