32 minutes ago, edmundp said:

True, but Plonkster already has all the equipment.

Lies he said he really wants a bigger inverter

3 minutes ago, viper_za said:

Lies he said he really wants a bigger inverter

Want, yes. Don't need it though. For the cost of the anti-islanding, I can take the next step in the original journey I envisioned three years ago, which is basically the story you can read in the self-consumption whitepaper. I sort of lost track of that vission, this thread brought me back to it. I was planning a bigger battery bank. Now I'm wondering what the heck I was thinking.

The plan: Small efficient inverter, remove baseline and other "always plugged into the same outlet" things from the grid. Get back into a space where "expansion" means "buying another PV panel".

But right now I'm sorting out other areas re finances. Financial Discipline and all that. Seeing that the Hakko Soldering Station you want is back in stock and the price looks good... it took a LOT of will power to tell myself NO. Thou shalt do only planning. The big stuff comes next year. Maybe.

(In other words, talk to me again in three months and I'll tell you something different).

1 hour ago, SilverNodashi said:

It has dual MPPT trackers, do you could effectively have poly and mono crystal panals mixed, or west/south orientation mix if you need to. OR just have two strings of the same panels. Quite a nice feature. But it's a high voltage system - 250V - 900V, so you need minumum 11x 250W panels. 

That is either the 5KW Plus or 10KW model - The 3KW can handle panels of up to 3200W and the 3KW Plus can handle panels up to 4500W.

V_max MPPT on the 3KW Plus is 500V.

MPPT optimal working range is between 250V and 450V on the 3KW (Plus)

1 hour ago, edmundp said:

Could you not just use one string of 9 x 250W panels all in series? Thus 37Voc (approx) * 9 = 333V? Or would this result in terrible MPPT usage?

That should be perfect - the nominal V_MPPT is 360V and the MPPT range is 250V - 450V.

2 hours ago, edmundp said:

How responsive is the MPPT portion in terms of demand/power usage changes? Let me explain - one of my pet peaves with the Axpert is the slow response time of the MPPT. Quite often when the batteries are nearly full and due to either a substantial change in load or sudden light condition change the MPPT will first maintain current level (either boosting charge volts by up to 1.5V above setpoint for a few seconds or dipping the same) then followed by yet another boost or fall, followed by a fall and then only settle down. Each phase lasts a couple of seconds. Repeat this a couple of times a day and it leads to all sorts of data anomalies. Plus the DC bus voltage on the inverter side ranges between 396VDC and 451DC when this occurs - I do not think this is so healthy.

And then the ever important question in my mind - what is standby usage in idle?

Any comments on the Infini's MPPT?

I have not monitored V_BUS, but it responds very quickly to changes in PV. I have a power meter between the infini and the grid right next to the inverter and if clouds move in font of the sun or out of its way, you can immediately see the change in PV power on the inverter's display and also on the power meter - there are no delays as you've experienced on the Axpert.

During the night the infini 3K Plus draws about 60W - roughly the same as the Axpert 5KW.

2 minutes ago, superdiy said:

I have not monitored V_BUS, but it responds very quickly to changes in PV. I have a power meter between the infini and the grid right next to the inverter and if clouds move in font of the sun or out of its way, you can immediately see the change in PV power on the inverter's display and also on the power meter - there are no delays as you've experienced on the Axpert.

I see the same on my Axpert when looking at the display(The software does take a second or two longer to display the same).

@edmundp
I am pretty sure it updates way faster than the software does and this let's you think it takes longer then it actually does for the MPPT to ramp up and down

13 minutes ago, superdiy said:

During the night the infini 3K Plus draws about 60W - roughly the same as the Axpert 5KW.

My Axpert 5kVA only shows a max of 26W(23-26) draw on the BMV when in bypass mode

8 minutes ago, viper_za said:

My Axpert 5kVA only shows a max of 26W(23-26) draw on the BMV when in bypass mode

On the BMV - that is then from battery.

It is about 60W from the grid.

Just now, superdiy said:

It is about 60W from the grid.

Don't think it does both.
As far as I understood the Axpert powers up from your bank and not the grid?

27 minutes ago, viper_za said:

I see the same on my Axpert when looking at the display(The software does take a second or two longer to display the same).

@edmundp
I am pretty sure it updates way faster than the software does and this let's you think it takes longer then it actually does for the MPPT to ramp up and down

The software will be slower to respond, because you have to poll the inverter constantly to get the data out of it. The time to update PV and power data will depend on where in the list of poll commands you are at that moment and what type of connection you have to the inverter, i.e. serial or USB. In the case of the 3KW infini you have to poll about 8 commands in a continuous loop and even via USB without any pause in-between commands it still only give you a refresh time of between 1 and 2 seconds.

 

7 minutes ago, viper_za said:

Don't think it does both.
As far as I understood the Axpert powers up from your bank and not the grid?

I might be wrong - I have not tested this since I do not have one, but that was mentioned by other users in this and other forums.

The infini 3KW draws about 60W from grid during the night though.

2 minutes ago, superdiy said:

I might be wrong - I have not tested this since I do not have one, but that was mentioned by other users in this and other forums.

The infini 3KW draws about 60W from grid during the night though.

Thinking about this, I'm almost 100% sure it does not use the grid for power.
I will have to test this with the prepaid meter just to make sure

Edit
What I think they saw as self consumption was when the inverter was allowed to charge the bank with Grid and to float the bank you would see usage around that wattage 

Actually I see the result from the BMV display 😯

Sent from my SM-N900 using Tapatalk

2 minutes ago, edmundp said:

Actually I see the result from the BMV display

OK, that is on the battery side only - I have not really monitored battery charging/discharging that closely on the infini - changes in PV reflects instantly on mains power though.

27 minutes ago, superdiy said:

The software will be slower to respond

@Coulomb corrected me on another post where we were discussing the number of uarts on the microcontroller, the Axpert has two, one for comms to the PC, another to talk to the SCC, and for the comms with the display it bit-bangs. So it would make sense that the display is somewhat slow to update.

9 minutes ago, edmundp said:

Actually I see the result from the BMV display

Sent from my SM-N900 using Tapatalk

If you are talking about the voltage spikes when high loads suddenly turn off, I would say it's normal in all systems that use PV to carry load and charge the bank at the same time.
The excess power generated needs to go somewhere and the best place I would say is into your bank as it is for such a short period it should not cause any harm

1 minute ago, plonkster said:

@Coulomb corrected me on another post where we were discussing the number of uarts on the microcontroller, the Axpert has two, one for comms to the PC, another to talk to the SCC, and for the comms with the display it bit-bangs. So it would make sense that the display is somewhat slow to update.

Although that might be true, by software I'm actually referring to monitoring software e.g. on a PC. 

Mmmmh, so in all this, I'm now wondering about the Conlog prepaid meters (I have one). It seems some of them have reverse-power protection built in and can either measure it or switch off (and I'm sure all of them are configured to switch off), so now I'm wondering how long (in milliseconds or seconds) you can push back before it becomes unhappy, and this is all related to this "how quickly does it respond" stuff because, if you run your infini without batteries (as I've been told you can), for those few milliseconds until the MPPT pulls back you will push into the grid.

So as I understand it, the thing with AC is that the current doesn't actually flow in any particular direction, but the power does. The current changes direction every 0.02 seconds, but the power vector (voltage times current times power factor) will generally point one way or the other. So when the power is going the wrong way, it means cos(theta) is negative or current and voltage is more than pi/2 (aka 90 degrees) out of sync. So now I'm wondering if there are every any times that this happens for legitimate reasons (big inductive loads switching off perhaps?) and whether there will be some tolerance which would give a grid limiter the space to back off?

For educational purposes of course... because one sort of has to know before you even waste the money going in a certain direction (in this case, hub-4).

Edit: I think @Wetkit is the man to ask. He had an infini and a Conlog issue. I think he said 10 seconds on, 50 seconds off or something like that? 10 Seconds is ample space to back off... :-)

@plonkster What I also heard was that some prepaid meters do not measure direction and what then happens when you push back to the grid, it actually sees as you using electricity and deducts from your units loaded 

13 minutes ago, viper_za said:

@plonkster What I also heard was that some prepaid meters do not measure direction and what then happens when you push back to the grid, it actually sees as you using electricity and deducts from your units loaded 

Yes, I heard that too, and that would actually be fine, because paying for a kilowatt going the wrong way for half a second (because you turned the kettle off) isn't going to break the bank. But the Conlog meters (not all, but many of them) specifically have "Load/Line Reversal Protection" built in, which will trip when you go over 40 watts. It's in the manual.

The manual also says:

Overload Indication
The consumer is disconnected when the load power / current being consumed, averaged over a 15
second period, exceeds the load / current limit. When this happens all ten segments of the rate
consumption bar-graph flash as an indication to the consumer to switch off appliances. The load limit is
programmable allowing the supply authority to limit the maximum power an individual consumer is able
to draw.

Now unless someone got fancy with the hardware, my guess would be that the reverse protection is in this same general area of the software, so you will probably have up to 15 seconds before it goes. But I summoned @Wetkit, maybe he knows :-)

Edit: (Man I do a lot of edits after the fact...), watts is units of energy per unit of time, Joules per second... so any power measurement implicitly implies that there is a time window in which the measurement is done, so there might be some wiggle room.

Perhaps I should visit the local demolishion guys (Doempie) and see if I can score an old meter.

I have noticed spikes picked up by the BMV lasting a couple of seconds do not translate to spikes in individual battery voltages. Admittedly the voltmeters are only accurate to 100mV but spikes of over 1V should translate to at least 100mV change per battery.