powerjack 24-volt 15000W lfsp inverter new 5-14-2021 manufacture

No.  The problem is that the regulation code has no ability for "no change."  This means that every half-wave when the SPWM throttle is adjusted, it HAS to be incremented OR decremented.

The end result is that the throttle has a never-ending stream of +-+-+-+-+-+-+-+-+-+-.....ad infinitum.  This results in one side of the MOS boards running hotter than the other--as one side is always doing just a tad more work.  (Let's not get started on the mismatched MOS drive and PCB layout that also compounds this.)  Worth noting that when I revised the firmware to add a "zero spot" (i.e. no change), the low-side FET temperatures evened out...

Oftentimes, the zero-hysteresis regulator will get into a bigger oscillation with itself, i.e. ++--++--++--++-- or even bigger +++---+++---+++---.  These will result in very noticeable oscillations (significant flickering) at some ratio to the AC mains frequency (i.e. 30Hz, 20Hz, 15Hz, etc.)

I've seen wild oscillations like this on GS inverters before (solved with a firmware update, haha!), when Sean was running his smaller A/C unit on a GS6.  When the inverter regulation was oscillating, the registered output load was double what it should have been--so your reporting PJ alarms is completely understandable.  My solution was to slow down the regulator--which has the side effect of more noticeable brief "flickers" when large loads are turned on/off.  (Sometimes there isn't a perfect solution 😉)

When the inverter regulation was oscillating, the registered output load was double what it should have been--so your reporting PJ alarms is completely understandable

I use lithium ion battery  at 62  v DC  and no flicker .   Your lithum IRON battery  will max out at  58 v DC .   When the DC voltage go to 52 vDC the light flicker and it is time to recharge the  battery .    I  use the flickering light to tell me when to shut down the inverter and recharge the battery .    That is one use for a poor  PJ cpu  design and no wifi board .     

This is coincidental behavior, likely resulting from mathematics and the CPU regulation.

My v10.3c/ASL3 upower 8k 24-230V does intermittently develop some 12.5Hz  (fOut/4) voltage oscillation under light load (<200w). Above that it seems pretty stable. This only causes a noticeable issue with cheap chinese LED bulbs that don't have a proper AC-DC converter inside. All the decent quality LED bulbs that I have bought from uk retailers are fine. The direct-from-china bulbs don't have a proper converter chipset - just a rectifying diode, capacitor (usually undersized so some noticeable 50Hz flicker even on grid, which most people don't even notice but really disturbs me) and a dropper resistor.

The really old PJ V1.4 of same specs with 2x AS2 transformers in series doesn't exhibit any oscillation. But Sid said in response to one of my older posts that has a completely different MCU and hence firmware

Yep earthing and fusing is simply there to protect the installation (wiring etc) from overheating and potentially catching fire in the event of a fault. And as you rightly state, the maximum disconnection time is 5 seconds for most circuits, in UK regs anyway, not sure about other countries. 

When rewiring the house shortly after buying it in 2013, I opted to fit RCBO's on every circuit rather than the clumsy dual-RCD and a bunch of MCB's which seems to be the uk electricians default choice. Reason being is that although my choice costs a bit more, it allows the use of a smaller consumer unit by saving 4 unnecessary slots, makes any kind of earth leakage fault traceable to a single circuit, and crucially, prevents an earth leakage fault on one circuit from killing other circuits. So for example the washing machine develops an earth leakage fault, it doesn't put out the lights!

Regarding the off/on grid switching I utilise a 2PCO switch with centre-off to ensure no risk of grid-inverter connection. I have the inverter neutral tied to the grid earth, which is actually combined with the grid neutral as the supply is what is called PME in the Uk. So my RCBO's will still trip in the event of a >30mA earth leakage on any circuit. The inverter case is also tied to the same earth. I have no earthing on the DC side other than the metal framework on which the PV panels are mounted being about 2ft into the soil.

Using several RCDs instead of a single master also addresses low level leakage, which is perfectly normal, accumulating and pushing the master RCD closer to its trip current resulting in the last straw, which may also be completely safe levels, causing a false trip.

Indeed, that is another benefit. When I bought the house it had a single RCD. On damp evenings this would sometimes 'buzz' and very occasionally trip. Yet without an expensive tester I couldn't find which circuit (or combination of circuits) was causing this

During the rewiring I found two rodent-chewed wires on different circuits in the void between ground and 1st floor. The damp air circulating here must have occasionally condensed on the wire and the sum of the two was enough to cause a trip. Obviously both lengths of cable were replaced!

the 7.5KVA transformer arrived yesterday via Fedex. unboxed it in the middle of the night. heavy at 133 pounds to man-handle. it is a 240x480 primary to 230x340 secondary general purpose transformer. i am planning to hook it into the 240 out put of the inverter then either use 240 or 120 to the main breaker AC panel. all off-grid.

thanks for the grounding discussion, I am still reading and studying those points and will incorporate the grounding on the AC side of things with hopefully some type of GFCI type protection.

the 7.5KVA transformer with be about 95 percent efficient so I will loose/waste some solar generated battery capacity with this(5 percent). I still need to make more room in the off-grid solar power shed while I wait for weather to allow the bigger outside equipment builds.

right now it is still a floating neutral without an earth ground like in the RV systems; I will change that in a bit of time.

 I do really appreciate the discussion and feedback. thanks all

hard to say what it really weighs, the labels say 156 pounds , the shipping label says 139 pounds from Fedex, eBay seller said 133 pounds but it is a heavy brick that is for sure. the primary wires are smaller diameter and the 4 secondary wires are larger diameter. the wires are all labeled. there are 10 primary wire to give you various input options and the 4 secondary options to give you 120 or 240 volt out put. they have a green bolt and green wire connected to the case also. it is encapsulated so you can not actually see the wound transformer, but I suspect/believe it is made of copper.

I will have too straighten the mounting brackets a bit but had to peak inside this morning. this is the bottom  as it is upside down on the 2 wheel hand truck. i think i will mount it on some unistrut maybe so I can service the wires on the bottom and i can either have them exit out the back or side through he knockouts. i will bring in to the primary through one knock out ant go out through a different knock out to separate and add some type of organization to the connections. 

the wife will not like it if I damage the high priced ceramic tile in the kitchen so I will soon be banned to the cold outer world. more later.

... and here's a detailed response from Will Prowse where you can also find a link to Signature Solar's response.  TLDR summary, Signature Solar's Growatt inverter is provided to them modified by Growatt to not have the N/E bond internally.

Executive summary:  Be aware of what you are doing, what you buy, and where you buy it from.

yes, I saw that will prowse video of the cheap grow watt brand inverters, the one I found on eBay was being sold by signature solar the 12k grow watt for $2049 plus $175 shipping. and it has a built in MPPT solar charge controller. all in one type of inverter, but I am still assembling things separately. 

separate LiFePO4 batteries, separate inverter, separate solar charge controllers, breakers etc, 

presently buying GFCI AFCI breakers for a Square D QO plug-on-neutral panel for the AC side of things.

I am still waiting on the 26000 AMG version PowerJack inverter to arrive some time in March 2022. (January 12th 2022 score for $527 including shipping and tax direct from china via Ebay auction). I am planning to run 220 out of it. for the bigger appliances!

so the question comes around: is the neutral and ground bonded or not on the AMG version? still trying to get my head straight on that issue!!!! the LF inverters made by PowerJack AMG version???

supposedly a lighter guage SS case than genetry solar inverters,  so that would decrease the shipping weight some. see what arrives and let you know then.....

thanks for the reply😎

the 5000 model grow watt imported inverters are HF inverters that will prowse is saying are dangerous, so may not be comparable to the LF 26,000 watt AMG PowerJack LF inverters.

signature solar also came out with a video about theirs being specifically made for the USA market different than other vendors???? 

like i question???  i am not sure they would be comparable top a Powerjack LF AMG inverter???

grow watt 6000 watt HF all in one inverter vs AMG 26,000 watt PowerJack LF inverter.????? the 26,000 watt Poweerjack AMG inverter is strictly an inverter without charging or mppt options. I am doing that separately. Personally I am not interested in the charging from the grid. I am only interested in the inverter capabilities for this off-grid build.

my thanks for your thoughts, replies, and information!😎

the 12k grow watt is a LF inverter as i understand it though.....

so the question comes around: is the neutral and ground bonded or not on the AMG version? still trying to get my head straight on that issue!!!! the LF inverters made by PowerJack AMG version???

NO .   Neutral and ground are NOT bonded in any PJ inverter that I know .   PJ all have floating neutral  so the PJ inverter neutral  has one wire and only one back to the  main service entrance and is grounded  there to a  ground rod  at only one place .  My one  neutral wire  from  inverter to service entrance is over 100  feet .    The ground go to the inverter case and is NOT connected to neutral at the inverter like  the  Europe  and China  Growatt 5kw  with the green bonding screw .   

Nope, they aren't directly bonded.  They are loosely coupled with 1-5 0.01uF (10nF) caps depending on how the factory workers felt that day...

I think people tend to give far more credit to the mythical concept of "inverter grounding" than should be.  It's just a binding post to connect to the inverter chassis--that's it.  Inverters don't "Generate" ground--only the ground generates ground...

Can you link where you found this info?

I personally doubt it; they likely have the same chassis thickness as the GS inverters.  The reduced weight is due to a severely undersized transformer.

I'll be very surprised if the "AMG 26,000W Power Jack" inverter will do 6kw continuous.  Can also say pretty certainly that you aren't going to get anywhere close to 12kw continuous output either.

Listed shipping weight for an "AMG 26,000W" is 42kg.  It's also got a bigger chassis than the GS 12kw, as well as 2 mainboards--both of which increase weight without increasing output power.

The current GS 12kw inverter prototype (without ANY electronics) shipped at 43.5kg; actual shipping weight when fully completed should be closer to 48-49kg.  Difference being that it will actually DO 12kw continuously 24/7.