RobM

I was told by a Sunsynk technical advisor to permanently bond the inverter output to earth. He maintained that this was safer than a bonding relay as the relay could fail. A number of electricians I know who install inverters follow this practice.  

I was told by a Sunsynk technical advisor to permanently bond the inverter output to earth. He maintained that this was safer than a bonding relay as the relay could fail. A number of electricians I know who install inverters follow this practice.  

I was told by a Sunsynk technical advisor to permanently bond the inverter output to earth. He maintained that this was safer than a bonding relay as the relay could fail. A number of electricians I know who install inverters follow this practice.  

I'll get straight to the point. If you are thinking of using a Growatt SPF series inverter with stand alone Lithium batteries, DONT! 
Growatt, for reasons known only to themselves have a charging algorithm that abandons the battery after charge, leaving it to supply a load (solar) or just invisibly feed its internal circuits (about 1A), slowly discharging to some mysteriously defined threshold where the charger kicks in again and tops up. This is purported to be 95% SOC or 2V drop below -?? float apparently. The 95% SOC (State Of Charge) threshold seems to be linked to a Lithium battery with comms to the inverter. Otherwise it's the 2V drop. 
Now.. enter a lithium battery with no comms. These are all the rage now and selling like hotcakes. The Hubble S120 being a good example - internal BMS with external balancing unit when connected in series. Lithium batteries have a very flat discharge curve so a small drop in voltage represents a big drop in SOC. This is quite different to lead acid batteries which have a relatively bigger drop in voltage with drop in SOC. 
Enter the Growatt dreaded algorithm: Without comms to the battery the inverter seems to estimate SOC by means of voltage drop. The magical 2V drop translates to a big drop in SOC for a lithium battery. The upshot of this is the "clever" Growatt algorithm essentially lets the battery discharge to around 30% or less before re-engaging the charger. That said I have yet to see this in action. Perhaps I have been too impatient. Its more like a 2V-ish drop. What this means in this case is that the poor Lithium batteries, in grid bypass mode are left invisibly feed the inverter at about 1V for three days or more before they are topped up. The result: constant relatively deep cycling of the the batteries for no good reason. A similar scenario would play out with solar as the charging source.
Growatt claim that the logic behind this algorithm is to save the "charging relay" from wearing out. Seriously??? 
IFFF the battery were left completely to its own devices to rest with no load at all after charging this might make sense but to subject a battery to perpetual cycles (deep in the case of LFP) for no good reason makes no sense to me. The friendly Growatt "international engineer" has spent three weeks of confused exchanges trying to convince me this is normal and nothing to worry about. Nothing to worry about is also the Invest Solar guy's response. 
So there is my observation. I wonder how many others have unwittingly hit on what they thought was a cost effective long(ish) life setup.
Wow - can't anybody make a decent reasonably priced inverter?  

default customer code is inf01  worked for me

It seems one needs to be ultra careful about connecting any inverter with grid feed potential with the HXE310-P meter. If it trips there seems to be no way of resetting it short of getting City Power to reset. How did you get the problem solved? 

Thanks @zsde, I'll take another look although I am pretty sure its all voltages. The inverter in question is the SPF 3000TL HVM which is a somewhat lesser beast than the 5000ES which I assume you have. This is used as a backup system so no PV, just grid power. It spends most of its time in grid bypass mode (slowly wasting away battery capacity). I am not aware of a firmware upgrade and the Growatt people confirm it has the latest firmware. 

I found the Hubble support line very helpful. I logged a query through their web site and someone called me and answered my questions directly. The local Growatt support guy seems to be little more than a postbox with scant product knowledge. The Growatt "international engineer", Amos is responsive but frequently gets the wrong end of the stick and also has to refer queries to "Research". 

I am now thinking Kodak is your best bet for Axpert type inverters. They are distributed by Segen Solar who carry spares locally, have knowledgeable tech support locally and seem to be a well run outfit.  

With a battery with comms at least it only goes down to 95%. As much as this wastes solar power and cycles the battery unnecessarily at least it's still fairly viable. With dumb lithium batteries its a much bigger mess because the battery discharges deeply even when there is power available. 

I'll get straight to the point. If you are thinking of using a Growatt SPF series inverter with stand alone Lithium batteries, DONT! 
Growatt, for reasons known only to themselves have a charging algorithm that abandons the battery after charge, leaving it to supply a load (solar) or just invisibly feed its internal circuits (about 1A), slowly discharging to some mysteriously defined threshold where the charger kicks in again and tops up. This is purported to be 95% SOC or 2V drop below -?? float apparently. The 95% SOC (State Of Charge) threshold seems to be linked to a Lithium battery with comms to the inverter. Otherwise it's the 2V drop. 
Now.. enter a lithium battery with no comms. These are all the rage now and selling like hotcakes. The Hubble S120 being a good example - internal BMS with external balancing unit when connected in series. Lithium batteries have a very flat discharge curve so a small drop in voltage represents a big drop in SOC. This is quite different to lead acid batteries which have a relatively bigger drop in voltage with drop in SOC. 
Enter the Growatt dreaded algorithm: Without comms to the battery the inverter seems to estimate SOC by means of voltage drop. The magical 2V drop translates to a big drop in SOC for a lithium battery. The upshot of this is the "clever" Growatt algorithm essentially lets the battery discharge to around 30% or less before re-engaging the charger. That said I have yet to see this in action. Perhaps I have been too impatient. Its more like a 2V-ish drop. What this means in this case is that the poor Lithium batteries, in grid bypass mode are left invisibly feed the inverter at about 1V for three days or more before they are topped up. The result: constant relatively deep cycling of the the batteries for no good reason. A similar scenario would play out with solar as the charging source.
Growatt claim that the logic behind this algorithm is to save the "charging relay" from wearing out. Seriously??? 
IFFF the battery were left completely to its own devices to rest with no load at all after charging this might make sense but to subject a battery to perpetual cycles (deep in the case of LFP) for no good reason makes no sense to me. The friendly Growatt "international engineer" has spent three weeks of confused exchanges trying to convince me this is normal and nothing to worry about. Nothing to worry about is also the Invest Solar guy's response. 
So there is my observation. I wonder how many others have unwittingly hit on what they thought was a cost effective long(ish) life setup.
Wow - can't anybody make a decent reasonably priced inverter?  

The batteries make up for a big portion of the cost when installing solar, that’s if you also need backup after dusk. The cost of putting power into those batteries with PV are relatively cheap if you take into consideration the lifetime of the PV panels that will do most of the work over many years and possibly outlive any inverter and batteries. 
Batteries can almost be compared to a generator using fuel, to a inverter using batteries. The difficult part is to re-coup as much as possible of that battery cost over the shortest possible time. In my opinion it is a real work of art to put a system together that is efficient with the bare minimum.

I have managed to get successful comms to macbook using the usb connection and watchpower. It takes a little while to connect. It may also have something to do with the kind of USB cable you are using, they are not all created equal. 
What looks like an ethernet connector (computer port) is actually a serial port. Why inverter manufacturers still use serial comms is totally beyond me.