Electronic Loads

[Hedges]

Those are the exact inverters I use for my real-world capacity test (2 of them).

For more meaningful real-world capacity test, I charge a top-balanced 8S battery to my 10A 8S charger’s limit of 28.8V, wait an hour, and then discharge at ~80A using a pair of maxed-out 1kW inverters each driving a 1kW space heater.

Oh? They are described as grid-tie. Designed to shove power into the utility grid, but don't think they would power a space heater.
(I thought returning the power borrowed for charging might be a better idea than just dissipating the kWh)

 

[fafrd]

Qualification of each individual cell capacity. There has been a lot of commentary about these 280AH cells being unmatched. I am curious to quantify just how unmatched my cells are are and if this changes over time.

There are better / easier ways to quantify individual cell capacity.

And the best way to quantify mismatches of a pack is to top-balance, wire them in series, charge with a standard charger, then power an inverter driving a suitable load (I like space heaters) while monitoring individual cell voltages (a monitor like BatGO is good for that if not already integrated into the BMS).

Capacity mismatch can actually be well-characterized at lower discharge rates, as well. Connect an 8S battery to a 2kW (~7 Ohm) heating element and over the course of 3+ days you can watch how closely the cells track).

The more important reason to discharge at higher rates (closer to intended usage rates) is to see the effects of internal resistance differences and especially how any weak cells drop under heavy load.

 

[fafrd]

Oh? They are described as grid-tie. Designed to shove power into the utility grid, but don't think they would power a space heater.
(I thought returning the power borrowed for charging might be a better idea than just dissipating the kWh)

The SUN GTIL inverters are truly a wonder to behold.

You modify an extension cord so that you can place the clamp sensor around the hot wire.

You plug that extension cord into any standard socket and plug your space heater into that extension cord.

You connect the GTIL to the battery and connect it’s clamp sensor (sensing the extension cord).

You turn on the GTIL by plugging it into the same extension cord.

It wakes up, displays the power flowing through the extension cord, then after a short delay, it shows the power it is generating and the net power flowing out of the socket and through the extension cord drops to 0W.

A GTIL relies on the grid to supply the PSW signal and then turns any extension cord (or an entire house) into a grid-neutral island.

You can use your own multimeter to prove that everything it is reporting is correct: the space heater continues to run and consume ~1kW while there is no current flowing out of the socket (and while DC current continues to flow out of the battery).

Who knows what the long-term reliability of these little SUN GTIL’s will be, and at 75% efficiency (at least at 24V), they are definitely substandard in that regard, but they are a true game-changer for DIY solar...

 

[Hedges]

I've figured out how you can make "zero export" instead do "zero export except for your approved 3kW NEM system".
Also, "zero export in excess of 3kW"

Can you use multiple SUN GTIL as a team to deliver multi-kW and still limit export?

 

[fafrd]

I've figured out how you can make "zero export" instead do "zero export except for your approved 3kW NEM system".
Also, "zero export in excess of 3kW"

Yeah, the key is how to set the AC current limit so it’s not kicking-in until exceeding 3kW.

Another alternative is to have all of my daytime consumption, primarily my 5 fridges/freezers, wired through the same subpanel. Then I could just feed the new Microinverter into than panel and only turn on the charger whenever that subpanel starts to export to the main panel.

That, combined with a timer on the charger so it’s only available during the peak hours of the day, get me to 95% of where I’m trying to go...

Just a clamp meter wired into a comparator driving a relay could solve the issue with no Arduino/smarts needed...

Can you use multiple SUN GTIL as a team to deliver multi-kW and still limit export?

See, that is the thing. The GTILs are fantastic for replacing consumption with battery power, but they don’t represent a load at all (except the battery),

So what I want is a GTIL-like load, which is a battery charger combined with a clamp sensor. An ideal charger would calibrate it’s consumption to offset export (like the GTIL calibrates generation to offset consumption), but I haven’t found any variable-power chargers and designing one goes way outside my competence level (or interest, frankly).

Since I’ll know the maximum power generation as well as the consumption of all the fridges, configuring a properly-sized charger to turn on as a dump-load to only turn on when all fridges are off sounds like the most straightforward solution...

 

[Hedges]

Yeah, the key is how to set the AC current limit so it’s not kicking-in until exceeding 3kW.

If that limit isn't programmable, only aims for zero kW, I can fool it by magnetically summing currents.
Trivial one is to subtract current produced by NEM-approved 3kW inverter.
Bit more complex is to subtract a fixed 3kW. But armed with ohms law and a spool of wire I could do it.

 

[fafrd]

If that limit isn't programmable, only aims for zero kW, I can fool it by magnetically summing currents.
Trivial one is to subtract current produced by NEM-approved 3kW inverter.
Bit more complex is to subtract a fixed 3kW. But armed with ohms law and a spool of wire I could do it.

That’s exactly how I’m fooling my GTIL inverters (total current into main - total grid-tied solar production).

so no, I don’t really need truly programmable, just ‘higher than 3kW so turn on battery charger to dump load’ and ‘below 2.9kW so turn off the battery charger’

How would you convert the current from the clamp sensor to a control signal closing or opening a relay?

 

[Hedges]

Clamp sensor probably has a resistor included. The ones I have which went with a power monitor produce 300 mVrms AC at 100Arms

You could build a circuit to buffer, rectify, and filter then feed a comparator along with reference from potentiometer.
But, you'll get magnitude of current not direction of power flow if you don't do more.
Ideally multiply by voltage. Could synchronously rectify with a pass FET enabled by one polarity of voltage so you multiply by 1.0 during half cycle instead of a sine wave.

uP digitizing volts and amps with at least 240 Hz sample rate and you might do it digitally.

 

[fafrd]

Clamp sensor probably has a resistor included. The ones I have which went with a power monitor produce 300 mVrms AC at 100Arms

You could build a circuit to buffer, rectify, and filter then feed a comparator along with reference from potentiometer.
But, you'll get magnitude of current not direction of power flow if you don't do more.
Ideally multiply by voltage. Could synchronously rectify with a pass FET enabled by one polarity of voltage so you multiply by 1.0 during half cycle instead of a sine wave.

uP digitizing volts and amps with at least 240 Hz sample rate and you might do it digitally.

Thanks for the suggestions but we should probably take this discussion to another thread (thread-hijacking alarm going off ).

 

[HaldorEE]

Thanks for the suggestions but we should probably take this discussion to another thread (thread-hijacking alarm going off ).

Probably a good idea, but I don't mind. I took a different path on my load.