Max/Min Operating DC Voltage of 48v

Ideal system range for a 48v inverter is roughly 40-60v ideal (maybe up to 62v if necessary).  For Li-Ion, 14S is ideal; for LiFePo4 (LFP), 16S is ideal.

Programmable high/low cutout ranges are from the nominal voltage up/down roughly to the 40v / 60v limits.  After further testing, we may be able to handle higher voltages with a 48v setup.

What are your thoughts on a 17s LFP if it's not charged to 100% capacity? I was thinking of getting 17 280AH cells to keep one as a spare but then I started thinking that being as my system was going to be primarily as a backup system and not cycling daily I should probably only have them charged to 3.4 or so and thus a 17s system would be under 60v? I've never built a lithium battery bank so any input or advice would be appreciated.

I know a 3.2v battery its chargeing curve is at the 3.4v to 3.55v that is where the longest charging curve I've seen.  Anything pass that seems to be kind of a waisted power.

So I personally have a 12kwh 16S LFP bank (51.2v, 248Ah) that is my "daily driver" for living off grid.  Had it for almost 4 years now, with absolutely no issues (though it does help that I have some home-made 3A balancers on it; I plan to offer these via the Genetry Solar website once I get enough other things taken care of.  They can safely handle a much bigger battery than the usual "cheap Chinese BMS" that usually gives you 0.03A [30mA] balance current.) 

My charge settings are currently:

• 3.53vpc (=56.5v total) "bulk/absorb"
• 3.50vpc (=56.0v total) "float".  (The 32650-sized cells I have were rated 3.65v full charge; that's 58.4v for the bank; I am purposely quite a bit below this, again, for longevity.)
• do not use "equalize" settings on the MPPT
• no temperature compensation whatsoever
• keep in mind that Lithium-based batteries cannot safely be charged in freezing temperatures

The problem with using a super-low charge voltage on the LFP batteries, is that if the charge voltage is below the "knee" (there is a "knee" on both ends, both for fully charged, and fully discharged)...you will never be able to charge your batteries.  Nor will you know until the voltage starts to fall off the LFP "empty" knee.  This is because the charge voltage has to exceed the battery's present voltage for any power to flow into it.

For example, my LFP bank seems to have ~53v nominal voltage (after the sun goes down, this is where it sits all night)--that's ~3.30vpc idle.  I haven't seen it below 50v except for the 2 occasions that I've pretty well run it completely out due to a week of extremely gloomy cloudy weather.  My thought (without testing) is that charging at 3.4vpc would be an extremely weak charge at best.

A 12S LFP bank (on a crude DIY "gator") I have tends to full-float at 40v, or 3.33vpc--in other words, charging at 3.40vpc would only give you 0.07v of difference to get power into the cells with.  (That's one mighty tiny trickle charge.)

I would definitely suggest staying with the 16S + 1 backup cell idea.

The ones I'm looking at have a nominal voltage of 3.2, are full at 3.65 and empty at 2.5. There's only around 10AH from 3.0-2.5 so really no need to go that low. Freezing isn't a concern in my area. Does the gs inverter have a float charge setting? If so would it work with backing feeding the power through the output or would I have to wire it through the input for that?

49 minutes ago, Cali Carlos said:

Does the gs inverter have a float charge setting?

Yes, it can charge batteries from AC input.

50 minutes ago, Cali Carlos said:

If so would it work with backing feeding the power through the output or would I have to wire it through the input for that?

So technically...

If the AC source is a generator or AC grid, it will have to connect to the AC INPUT terminal (you will need to specify 120 or 240v for this on the order).

If the AC source is a grid-tie system, it will have to connect to the AC OUTPUT terminal (the GS inverter will throttle it down / shut it off as necessary for the battery voltage.)

11 minutes ago, Sid Genetry Solar said:

If the AC source is a grid-tie system, it will have to connect to the AC OUTPUT terminal (the GS inverter will throttle it down / shut it off as necessary for the battery voltage.)

Yes when using the system for backup power it'll be a grid-tie system but if I wanted it to float the batteries to make sure they're charged when in "storage" then I assume I'd want it wired into the input for that? 

Should I always have them floating or is the self discharge on LFP so low that it only periodically needs it?

Can I have the input wired for 120 but still keep the grid-tie capabilities?

Can I have it in grid-tie mode and also charge via a generator if my panels aren't quite producing enough?

5 minutes ago, Cali Carlos said:

Should I always have them floating or is the self discharge on LFP so low that it only periodically needs it?

Assuming they're good LFP cells, the self-discharge should be practically negligible.  Running a charge cycle at least every year would be a good idea, just to be safe.

I have experienced Li-Ion ("recycled" 18650s, mind you) batteries randomly failing and draining themselves to 0v...but those were anything but brand-new cells.  LFP batteries can do similar, but only if they are very old and/or quite abused.  (If a cell is at 0v, "recovering" it is not recommended...yes, it may be possible, but the cell will never be the same again.)

7 minutes ago, Cali Carlos said:

Can I have the input wired for 120 but still keep the grid-tie capabilities?

Absolutely.  If the inverter is a 240v split-phase output, I'd recommend running 240v from the grid-tie system into it.  The AC input is a separate circuit, and can be configured for 120v when ordered.

8 minutes ago, Cali Carlos said:

Can I have it in grid-tie mode and also charge via a generator if my panels aren't quite producing enough?

Technically no, but functionally...sorta, at your own risk.

Grid-tie master mode is not available in anything other than "inverter mode".  In charge mode, the inverter will function as a pass-through (even if only 120v input, it'll still output 240v thanks to the autoformer principle of the main transformer secondary.)  So technically the grid-ties will see an AC signal and try to start--that is, if they don't mind the hash/noise on the AC line resulting from the inverter running battery charge. 

But in AC charge mode, the inverter has no way to adjust the output frequency to throttle/shut down the grid-ties if necessary.  It also cannot identify or prevent power from getting backfed into the generator.

16 minutes ago, Cali Carlos said:

Yes when using the system for backup power it'll be a grid-tie system but if I wanted it to float the batteries to make sure they're charged when in "storage" then I assume I'd want it wired into the input for that? 

I'm still working on the "grid-tie charge" code, working to perfect that...haha.  OTA updates are so handy 😉

Yes, I'd recommend charging the batteries via AC input.  You would only charge batteries via grid-tie master mode IF there was no connection to the AC Mains; it won't coexist with AC Mains.

53 minutes ago, Sid Genetry Solar said:

If the AC source is a generator or AC grid, it will have to connect to the AC INPUT terminal (you will need to specify 120 or 240v for this on the order).

Eh, I don't remember this being a part of the order options. You're planning for charging to roll out after or before these ship?

I'm going to need to figure out BMS control before I can enable inverter charging, right now all I'm hoping for is decently fast ATS function.

7 minutes ago, kazetsukai said:

You're planning for charging to roll out after or before these ship?

I've been testing it lately...working alright.  Will probably need some polishing before I'm fully happy with it...but current limit and voltage settings DO function as of right now.

7 minutes ago, kazetsukai said:

Eh, I don't remember this being a part of the order options.

Unfortunately I think you're right.  Hey, we can make mistakes 🤪.  It affects the internal wiring, a solder jumper, and CPU configuration...I'll buzz Sean on that one.

9 minutes ago, kazetsukai said:

right now all I'm hoping for is decently fast ATS function.

This fast enough for you?  Roughly 1/8th of a second from loss of AC input power to full inverter output voltage (at no load).

Blue is AC input...purple is AC output.

<a class="ipsAttachLink ipsAttachLink_image" href="/monthly_2021_03/image.png.a7d298f6fe1bc572c02715cee00c8e6d.png" data-fileid="156" data-fileext="png" rel=""><img class="ipsImage ipsImage_thumbnailed" data-fileid="156" data-ratio="75.00" width="1000" alt="image.thumb.png.6dc36f72cc20497970705cadf361f3c2.png" data-src="/monthly_2021_03/image.thumb.png.6dc36f72cc20497970705cadf361f3c2.png" src="/applications/core/interface/js/spacer.png" />

Ha, looking at the 'scope readout, I see that the relay hadn't released by the time the inverter AC wave started.

Think I need to tweak that a bit.  It needs to detect a lost wave a bit sooner...without being cantankerous and tripping out with noise on the input AC signal (yes, it will disconnect from AC input if either frequency or voltage are out of range).

4 minutes ago, Sid Genetry Solar said:

Unfortunately I think you're right.  Hey, we can make mistakes 🤪.  It affects the internal wiring, a solder jumper, and CPU configuration...I'll buzz Sean on that one.

Bummer it can't do both, but I get why based on your previous statements. If I had to choose I'd pick 240V.

6 minutes ago, Sid Genetry Solar said:

This fast enough for you?  Roughly 1/8th of a second from loss of AC input power to full inverter output voltage (at no load).

I'm curious to see what happens when AC power comes back as well, if you got that. And your thoughts on what might happen with a modest load in both scenarios.

I currently use a manual changeover, it'll be nice to just plug in. Going to be putting a hardwired surge protection device upstream of the inverter.

Don't think I have a screenshot of that, though I could take one at some point here.

Modest load...not really sure, though it'll probably have a bit of a power "splash" when the relay pops on.  I have tried to programmatically get the relay to actuate at the zero crossing--but you know how unpredictable a mechanical device is.

I've toyed with the idea of having the inverter keep running for a cycle or 2 after the relay pops on (as it will only switch to an input AC wave if it can sync to it) for a seamless "on" transition...not implemented yet, but this would be possible to implement via firmware update.  Unfortunately, a seamless "off" transition is kinda impossible with a mechanical relay being its unpredictable self.

1 hour ago, Sid Genetry Solar said:

But in AC charge mode, the inverter has no way to adjust the output frequency to throttle/shut down the grid-ties if necessary.  It also cannot identify or prevent power from getting backfed into the generator.

Definitely not worth the risk and not really a big deal. If I need a generator it's either because it's late in the evening or because there's to many clouds. In either scenario I'd shut the solar breaker off.

1 hour ago, Sid Genetry Solar said:

Yes, I'd recommend charging the batteries via AC input.  You would only charge batteries via grid-tie master mode IF there was no connection to the AC Mains; it won't coexist with AC Mains.

When you say no connection to the AC mains I assume you mean through the input correct? For grid-tie master mode I assume I'd have the output wired into the AC main via a transfer switch or through a breaker directly on the main in which case the input would not be connected? And if I wanted to charge the bank via the AC input from the grid then I need to make sure that the output isn't connected to the mains?