JayMardern

OK so the manual indicates the DIPs are just for setting address (i.e. for paralleling), doesn't seem like this is configurable then. Could be a BMS/balance/cell issue - might be worth following Alex's suggestion in the other thread...

ShutDown is 20%. The restart value you see there (35%) is when the inverter must power loads back on after being recharged; after the shutdown occurs. There's a nice explanation of this in this thread!

The settings in your screenshot are for the inverter; based on your settings, your inverter is set to make an audible alarm at 35% but only shut down at 20% (and then power back up when the battery has been recharged to 35%). BMS comms/cabling looks OK since the inverter display matches the battery display. But you’re right - the shut-down looks to have been triggered by the battery reporting 0%. It’s weird that it reports 0% when it’s supposedly at 36% though - could this be a setting on the battery itself? The datasheet seems to indicate recommended discharge being down to 15% - so a cut-off at 36% seems high… are there any settings in the battery menu (or perhaps on its DIP switches) that might affect this? @MAGNETO RENEWABLE ENERGY Would you be able to assist by any chance?

Leroy Merlin is typically reputable, this is excellent pricing if legit. Looks like the datasheet is likely this one (555M)? The 'Questions' section has some of the specs listed.

Same here. Partly-cloudy/light-cloud-cover isn't usually a huge deal - but was raining a bit earlier and clouds containing rain seem to make a huge impact unfortunately. Yield yesterday in cloudy JHB was 7.8 kWh vs 26kWh the day before , for the same demand. 70% loss! Enough to run all the essentials for ~ 14 hours - but not enough to do any heavy lifting beyond that.

Agreed that we'd see an extra handful of units (2kWh maybe 3 per day due to self-consumption/charging-losses?) but the OP is reporting a tripling which doesn't sound right?

The AM-5's do need to be paralleled with the same model unfortunately; but 6000 cycles (assuming I got the right datasheet) on your Magneto is still nothing to sneeze at, They're also the same underlying chemistry (i.e. both are LFP). So far have been extremely happy (it's 1C and we run/cycle it relatively hard; we've peaked at 4kW [~ 80A/0.8C] during load shedding), but at only 6 months old it's too soon to tell how it'll hold up long-term. LFP is meant to maintain solid performance over time but you'll have to ask me in a few years 😀

Different BMS'es count cycles differently. Technically 1 full cycle for, say, a 5kWh battery, would be 5kWh charge followed by 5kWh discharge (100% in and 100% out); but different BMS'es may increment the count before hitting 100% (say, at 80% instead). It's up to the manufacturer. Compare your usage to the datasheet of your battery model (for example check if this is yours - if so, it's rated for 6000 cycles to 80%) to see if the cost savings over the lifespan of the battery are greater than the cost of a new battery. I daily solar-cycle my AM-5 because the 10-year warranty doesn't limit cycles and if it dies the day after it turns 10, cycling has still paid for itself - everything beyond that is a bonus!

This worked. Brilliantly. Thermostat was around 55-60, pushed it to the max (70) which added about 30-40 minutes of heating time during the day; cut it off at 4PM - the water was piping hot for everyone. Thought it might be fluke (because I absolutely couldn't do this last week before the thermostat change) - but 3 consecutive days have been perfect. Geyser consumption per day is roughly the same; but said consumption now happens entirely during daylight hours. Winter might require me to to do the same on the other geyser (and move one shower there). This is also one less thing for me to worry about since I no longer have to manually skip the nighttime heat-up when everyone's done for the day during daylight. Basically added a 2.5kWh battery to my system, at zero cost, which happens to hold water! Thanks PowerForum!

Interestingly inital hardware revisions did reportedly have an external fan. Presumably they were deemed unnecessary later; I guess fewer moving-parts to fail is a good thing if the hardware can handle it. Since these things have warranties up to 10 years (depending on battery choice), I'd expect them to be OK as-is... But external cooling definitely can't hurt - some nice external-fan ideas here for when you get around to it:

Nope, Manager doesn't give you the lower time-refresh rate, You have to fill in the form and you'll see the 60s option then appears (without any change to the 'role permission' for your user); took a day for me. Love the change, should be rolled out to everyone!

These temperatures do seem normal; mine are similar, Thread on the Deye (same hardware) here where others report similar temperature range: https://powerforum.co.za/topic/16954-deye-5kw-hybrid-inverter-running-temperature/ As an aside, my 5kW Sunsynk unit also doesn't have an external fan - but on really hot days (>32 degrees), under high PV production, there's an internal fan that I can hear kicks in. It has to be really hot, though.

Exactly; peak PV will now be beyond the 5000W AC rating of the inverter. Hence if PV goes above 5000W (whilst AC load is equal to - or even greater than - 5000W), I'd need to make sure the battery isn't full so that we can dump the balance of the PV (up to an additional 1500W - for a total of 6500W PV supported by the inverter) into the battery to minimize wasted PV. This is since DC-to-AC is capped at 5000W, whilst DC-to-DC can do the full 6500W if I'm not mistaken, Yes!! And push up the thermostat to the max (anyone know if there're limits on this?) - to maybe score that last 30% extra heat needed to handle the evenings using the existing geysers. (Or so I hope!)

We can't invert more than 5000W to AC - so the balance would go to the battery via DC-to-DC, correct? So if I pushed up the array to 6500W (another 5-ish panels - not all that expensive), then on peak-production days, as long as the battery is below 100%, the 1500W balance above 5000W could go to the battery - and then it'll only get clipped to 5000W if the battery is full? And of course there'd be benefit in the hours further away from mid-day where were're further away from that maximum; right?

Heat pumps are incredibly efficient but their short-ish lifespans (especially under heavy use) and high-ish upfront cost is a bit of a downer: I've got three reverse heat-pumps (aka air conditioners!); and the two that are most frequently used are out of action; including a nice, overpriced Carrier unit that was only 7 years old (and had a non-mechanical electrical problem that no technician could economically repair without an expensive mainboard replacement; such is the nature of these things!). I'm thus inclined to avoid equipment with moving parts for this reason. I'm currently heating 350l of water almost entirely from the sun using cheap, easy-to-replace (insurance will do it for free!) regular geysers; via inverted PV. It's just those last extra 50-100l or so I need to cover; to treat another geyser like it's a giant, cheap battery for nighttime use! Or I need to enforce daytime-showering in my house! PV is typically rated at 25 years - during which time I'd probably go through 3 heat pumps - so the cost savings are likely to be at least partially negated... or am I mistaken? What would the requirements be for heating 400-500 litres of water per day via heatpump? In my case, the distance between the two geysers is about 15m. Could I get away with a single 2kW unit? Or perhaps a pair of 2kW units running sequentially? Do they last longer than I've been lead to believe if properly maintained? The other potential deal-breaker is that the smaller geysers would be run via the inverter Aux, which would hard-power-down depending on PV availability - to do heating primarily from the sun. Even on an overcast day, this would work nicely. That's fine for a geyser but cutting power to air cons is considered destructive, are heat pumps the same? I'm pretty close to the max already (4.5kWp on a 5kW unit - unless I add another battery to do DC-to-DC); I could push it up and just clip in the middle of the day, though. And of course paralleling a second inverter is looking tempting (to double the max!) Yes, this takes into account increases based on the historical 5-year average. And right now, hardware costs are even lower so this figure would be even better if I purchased now. To anyone on the fence: go for it! One other idea: perhaps I should crank up the thermostat of the current geysers to let them get hotter in the day? Is this possible/legal? Then this, plus a few extra panels (even with clipping) could potentially sort my household out?

@flatfourfan @GreenFields Insights much appreciated! 5kWh Battery is set to cycle down to 65%; and you're right: on this particular day cycling was not beneficial at all (other days are better though!) In this case the heavy-afternoon-use-culprit was the most usual of all usual suspects - the geyser. Daily hot-water-consumption in our house is around 12 kWh per geyser: a 3kW 150l and a 4kW 200l pair; so total of around 24kWh of heating per day. It's high. The challenge is that we also use a lot of our hot water at night; so the geysers get done heating during the day (consuming about 9 kWh [larger geyser] and 15kWh [smaller geyser] respectively by sunset, since the smaller one is used by the kitchen during the day). Unfortunately they still need to be topped up later in the evening for an hour or so otherwise they run cold; which burns at least 8 kWh a night directly from Eskom. On a typical day (graph attached), Smaller geyser (3kW) powers on at 8:30AM Larger geyser (4kW) powers on at 10:30AM, powering the smaller one off while it heats Larger geyser is typically at full temperature by around 1:00PM; when it's power drops to 0 the CBI Astute Timer switches the Smaller geyser back on immediately (details on this rule, here) Smaller one finishes by around 4PM. They both get heated again (over shorter periods) at night since we'd unfortunately run out of hot water otherwise. These are the evening spikes visible in my attachment to this post: a combination of a some Geysers plus some night-time electric dinner cooking. When the larger geyser is done during the day there's often enough solar to recharge the battery; so cycling makes sense on those days (like in this attachment); as well as when there's load shedding (since both geysers are on non-essentials and hence powered down during grid-down). On the day in my previous post, the kids all showered during daylight so the geysers never finished heating in the day (which is why there wasn't spare capacity to charge the battery). That's not usually the case though; but it did allow me to manually cut the evening-heating that day. I've considered moving the smaller geyser to Aux (to keep it powered when grid goes down); this would potentially take us a bit close to the inverter/battery max (which I could deal with by undervolting - though that has other side-effects). Downgrading the element is another option but that would cause heating it to overshoot daylight hours. The two potential longer-term solutions would be: The Small Upgrade Option: Add a pair of Geyserwise's to target lower temperatures in the evening. No solar benefit, but this reduces grid-waste in the evenings and means it matters less if I forget to bypass the night-time heating timer if the family happens to finish with the hot water in the day. Add a few more panels to cover daytime load-vs-PV overshooting; maximum array size is unfortunately just 5kWp (barely more than my current array) if I want to avoid clipping at mid-day. (I'd need a second battery to take advantage of the 6.5kW DC-to-DC PV-to-battery-charging capability on a larger PV array offered by the Sunsynk). Alternatively: Double the Whole System (aka The Nuclear Option!): Double battery to 10kWh ...and use the two batteries to power all essentials at night (via time-of-day power limiting) Double panels to 9kWp This would provide enough extra capacity to charge the new battery, new geyser, and cover daylight-hour load-vs-PV over-shooting Double inverter (2 x 5kW in parallel) Add an additional small 150l geyser (with a small 2kW element) in series to leave that heating the whole day; and power-down all geysers at night regardless, knowing we've absolutely got enough hot water for the night-time under all circumstances (and hence reducing geyser grid-usage to zero on a sunny day) ...and put the two smaller geysers on Aux (without worrying about inverter/battery capacity limits) to have them stay powered on during load shedding The biggest benefit would be had from the second option; but since the current system dropped us below the magic 1000-unit-per-month breakpoint, it's all diminishing returns from here. Doubling up would take us down to just a handful of units a month but because they're now cheaper units than before, return on investment (even with reductions in hardware pricing), will still climb somewhat for the system as a whole; from around 5 years (with current system) to 7 years (if doubling up). Still worth it in the long run , I guess!

This has been previously found to be rather common in equipment that contains MPPT’s for solar. I have lousy 4G (using Vodacom) where I live and haven’t noticed any difference (day or night) since installing my unit. (My DC runs across both MPPT’s and isn’t in shielded-metal conduit). But there’s a useful thread here describing this occurrence on Axpert/Synerji/Voltronic inverters, that also details solutions (including the above-mentioned shielding) for those who do find themselves affected and wish to curtail it:

Hit 30 kWh today (highest ever) from our 4.55kWp array; system installed May/June; located in JHB. Winter averaged around 20-25 kWh per day; spring has been closer to 25-30kWh per day. Per-Month solar generation is 800 kWh; usage from Eskom dropped from as high as 1700kWh (yeah the 6 of us were pretty heavy!) down to 750kWh this month as submitted to CoJ; the drop is a combination of PV (saves 800kWh) and using the inverter graphs to curb general usage (killed some old light bulbs/servers/night-time geyser use - added a further 150kWh of non-solar-related savings). Thus current usage is around 50%-50% Eskom-to-PV; with most of our Eskom usage happening when the sun is down (mainly cooking and additional water heating). Might be time to move over to gas cooking and add an additional geyser (and extra panels to heat it); with perhaps a second battery to handle night-time base-load. Bar-chart for the last few days attached as well as today's usage graph. Water usage costs are now a tad higher than electricity usage costs (never thought I'd ever see that happen)... so is it time for a borehole? Or should we further upgrade the PV system to start covering our night-time usage? Tough decisions! 😆 ROI (including projected CoJ tariff increases) should be less than 5 years. Planned properly, specced properly, and installed properly - a good solar installation can certainly make a significant difference; and we installed on a relatively conservative budget to boot...

Sounds odd - self-consumption on these units, coupled with a battery, shouldn't exceed about 150W or so. Are you perhaps exporting to grid unintentionally? Can you describe your setup - what battery(s)? Could you share photos of your inverter home-screen and Work-Mode + Time-of-Day settings?

@Scorp007 Yup! Lead Acids and LS just don't mix unfortunately; I had similar experiences with my old pair of 200aH's - rated for 1500 cycles to 50% but I only got about 700 cycles (most well above 50% SoC) before I ran into the same behaviour you're experiencing. Happened to 3 different pairs; all different brands. And even just a handful of run-lows can severely cripple their lifespan. This coupled with how hard it is to tell exactly what their current SoC is (due to voltage sag under load; and the same voltage being produced under the same load for extended periods before sudden drops like you're seeing here!), means it's challenging to determine if they're being well-looked after or not. Lithium (with their longer cycle life and their BMS'es handling all this for us!) is the way...

Since the 'consumption' values are so different between the two: is it possible that the first graph might be including export-to-CT (ie, the whole house) whereas the second graph is only from the LOAD/UPS 'essentials' perspective? Thus is it possible that an 'export' of 2.51kW in the second graph covers half the of the household non-essentials load (total including non-essentials 5kw) with the balance (including non-essentials: ie the remaining 2.5kW since the total load is 5kW) coming from grid to service the non-essentials balance that can't be covered by the CT export? Could you post a picture of the inverter display homescreen?

It looks like your load is exceeding PV generation capacity, So yellow is PV, Blue is total load, Red is the amount of that total load being serviced by the grid (due to insufficient PV to cover the full load)

On the Sunsynks (and presumably Deye as well since they’re the same OEM) you can adjust the grid connect/disconnect voltages in the Grid menu. I had to slightly lower mine to compensate for low post-load-shedding voltages from the municipality. Too low can damage equipment so it’s a good safeguard. You could see if you can check the V-ac graph in the app to identify how low (or high) it’s going before the inverter disconnects.

@Zodiac69 @Yuri Thank you! I'm not running any HA instances (I've decommissioned my always-on-servers in an attempt to reduce my energy footprint to deal with load shedding!); presumably this isn't possible with just the native Tuya/SmartLife app - but perhaps it's time for me to spin something up again for this kind of thing 😉

From Deye's service desk (this was quick!): It's ok to use, no problem.And gen ports are paralleled,outputs maximun 10kw. I'm assuming the documentation is out of date (likely there were firmware issues at some point that have since been solved)... this would also explain why using it as an input seems to be working OK now. But looks like we're good to go! From their reply, we can see that the gen port (which is the Aux port) would be paralleled; and the maximuim allowable output would be the combined output of both inverters (5kW + 5kW in my example) - exactly as we'd expect. Fantastic news. As load shedding gets worse and further impacts solar water-heating time, I do believe that it'll start making sense for pretty-much everyone using an electric geyser to move them from non-essentials (which are still powered by excess solar on these units) to Aux instead, to allow solar-based heating in the absence of the grid with a level of protection for the battery level that Aux offers. This even makes it possible to do heating in phases when there's insufficient PV to do it directly; since with an appropriate battery level set, a 3-hour heat could be split over 6 hours that take place automatically as the battery level rises. (So for example, a battery range of 70%-90% would have the port activate at 90%, heat the geyser down to 70%, switch off, and then repeat again when the battery gets back up to 90%). It'd take longer but the fact that it's even possible - without depleting the battery completely - is great news.