Clarification on charger in 6kW inverter

Frankly if Sid came out with a inverter that could handle a ev power pack, I would gladly snag several power packs and the inverter, instead of building batts.  Would make my system a bit stream line.

My setup is kind of a mix up Pup that don't know where his bone is, and well, thats what I get for starting in a hobby that I had no idea about, and buying things that really was just dumb in the first place.  Either way, now I know what I'm doing and understand a lot more on eletrical and all the crazy things about it, I can some what stream line it and not waist money on stuff that are just dumb.  If the GS inverter didn't came around I would have already bought two Victron 3k inverts by now.  But, when Sean announce the GS line and demo one, I jumpped in cause the fact is, the GS inverter from all I've seen blows even victron out of the water and its all thanks to Sid.

If you request 120vAC input, yes.

No, in charge mode, "inverter" is off.  Load is switched directly to the AC input; the transformer creates split 240vAC via the autoformer principle if configured for 120vAC.  Yes, this does limit the maximum input power...it is safe to say to 1/2 (i.e. 3kw), though for all practical purposes 4.5kw may be possible without overheating the transformer.

Absolutely.  Fully configurable.  (Gosh I need to write that manual...!)  Charge voltage (float and "bulk") are completely configurable, as is the charge amperage (based on AC input amperage FYI.)

Yes, charge current can be limited.

I'm not sure how many watts it can provide either 😉.  The limit will be the temperature of the FETs, as they are considerably more lossy in charge mode (due to the use of the backsplash diodes in the FETs, which drops up to 1.2v per quadrant).  I've run 33A back into a battery during tests (at least until my oscilloscope started to emit a terrible smell due to the noisy AC input!)...33 amps @ 55v is 1.8kw...really nothing to write home about--but nothing was getting warm/hot, so that's good.  It should handle 3.6kw without any issues.

I should note that the intrisic boost charge method (very common in LF inverters--it doesn't require any external parts) isn't particularly efficient...I was measuring 50% efficiency...which is pretty pathetic if I say so myself.  (Worth noting: I've never seen "charge efficiency" ratings on LF inverters, probably because of this!)  However, there's a "PFC" software tweak I can implement that reportedly will increase charge efficiency by 30%...which is quite significant.  Something to add to my list...

Why stop there?  https://theinverterstore.com/product/30kw-pure-sine-power-inverter-charger-300-vdc-240-vac-split-phase/

Haha.  Dangerous stuff, but...!

There's a good reason that the designs are very vehicle-specific: the battery systems are completely nonstandardized.  From a design standpoint, it becomes extremely difficult to try to support every little difference in manufacturer designs--not to mention battery voltages, connectors, interfacing, BMS systems, etc., etc.

You'll be perfectly dandy with most older switching-power supply appliances (yup, bridge rectifier -> filter cap -> switching circuit), however, some designs with PFC correction might not be too thrilled with DC input.

Something you MIGHT consider, would be to "hack" an existing high-frequency inverter setup.  As they're all different in their own ways, there may be a few "gotchas" to keep in mind...BUT most of them have a boost converter to generate 180-200vDC (for 120vAC output) or 360-400vDC (for 240vAC single-phase output).  Technically you might be able to use a 240v inverter, sneak power to the high voltage filter caps, figure out how to power the control module (most often on the cheaper Chinese units, it's an EG8010 / EGS002 board...horrible with surge loads by design, but not too shabby otherwise)...and have a solution of sorts at your own risk.

Of course, the 280-300v range is just simply difficult to work with--but you might figure something out.

Hey Sid, so when my Gen kicks in, the invert will be relying on my gen to provide the house the power?  Or can I have it just go into charge mode and have the slave take the load for the few mintues?  Dang, I wish my gen was a 240v now, I do got a honda with a turn switch to crank on it, and its got 240v, but I don't think its pure sign as its one of those construction type that frankly I can not even move it, I have to use a dolly to even move it and it takes two people just to even pick it all up. lol.  Its a 8500 watt model. I wish I could use it, as its got a massive tank and can run for nearly 24 hours on a full load, problem is its franken loud *grin*.  I'll stick with the 120v

1 hour ago, Sid Genetry Solar said:

No, in charge mode, "inverter" is off.  Load is switched directly to the AC input; the transformer creates split 240vAC via the autoformer principle if configured for 120vAC.  Yes, this does limit the maximum input power...it is safe to say to 1/2 (i.e. 3kw), though for all practical purposes 4.5kw may be possible without overheating the transformer.

Is there a mode that allows charge and inverter at the same time with an ac input or is that not possible due to the FETs not being able to do both at once?

Sid,

Thanks.  That answers all my questions.

So when charging from the mains the load needs to be 3600W or less, due the ability of the transformer to form 120VAC into 240VAC, with a surge to 4500 or so, but that would depend on having that much available at the plug.

I could use a more efficient separate 30 amp 120 VDC (ie 3600W or nominally 150 amps into a 24VDC bank) to maintain the battery charge from the mains as my solar will be initially undersized for my generator panel loads.

I could also recharge the batteries from the mains when the loads are smaller, like possibly at night.

And it may make more sense to use a DC-DC converter/charger directly into the 24VDC battery bank when away from home.

... or maybe just come up with the $60k Cdn for an F-150 Hybrid :-!

I think my next step is to start monitoring the loads on at least the generator panel to see what my real world requirements are before I go any further.

Thanks again, that helped a lot.

13 minutes ago, Baz said:

So when charging from the mains the load needs to be 3600W or less, due the ability of the transformer to form 120VAC into 240VAC, with a surge to 4500 or so, but that would depend on having that much available at the plug.

Yes and no.  Continuous 3kw for sure.  But it might be able to sustain 4.5kw total input draw, depending on the heat load of the tranny (this hasn't been tested).  I specced the output windings for 25A...not saying they can't do more, but they'll generate more heat at higher amperage...which limits the total power.

33 minutes ago, Cali Carlos said:

Is there a mode that allows charge and inverter at the same time with an ac input or is that not possible due to the FETs not being able to do both at once?

Sorta yes and quite a bit no 😉.  In the "charge" mode, it is not possible to produce any AC output (all AC output is a direct connection from the AC input).  In the "inverter" mode, it can't regulate battery charge.  Basically, it's kinda impossible to try to regulate 2 things with one throttle (i.e. AC output voltage AND backcharge rate.)

One half-exception: technically the grid-tie master functionality on a GS inverter allows for battery charge during inverter mode--but the crucial point to keep in mind here is that the inverter is the master device.  It throttles down the slave grid-tie system via frequency shift (either linear throttling if supported, otherwise it's an effective on/off) based on the charge settings and loads.  If the "slave" device does not accept/handle a frequency shift throttle, then it is not possible to regulate battery charge voltage/current.

But obviously, this cannot function with either AC Mains (grid) or a generator--neither accept a frequency shift, nor can either be throttled down.

Yes.  Generator has to take 100% of the load, otherwise it's not possible to charge the battery.  If the inverter is drawing power from the batteries, obviously, that's classified as "charge" in at least all of the books I've read 🤪.

The autoformer principle of the transformers' secondary will provide 240v loads from the 120v input--but obviously, all the power has to get funneled into the inverter through that single 120v phase winding--there's a lot to keep in mind here.

Technically yes, the slave will continue running (which might be a problem of its own)...BUT if the load on the battery is not offset by the charge current, you'll still be draining your batteries, even with the generator running.  Maybe less of a drain, but definitely not a charge.

Not to mention that battery charge mode is less efficient than inverter mode (due to system losses).  Kinda a catch-22.

36 minutes ago, Sid Genetry Solar said:

If the "slave" device does not accept/handle a frequency shift throttle, then it is not possible to regulate battery charge voltage/current.

At which point I assume the inverter would error out and shut down? 

Even if that wasn't an issue, and say a 10kwh bank was nearly depleted and you wanted to hook up a 1800w generator to the output and set the inverter to grid tie master mode for a few hours just to top up the bank a little but nowhere near long enough to fully charge the bank to where it wouldn't need the frequency shift. Could that work? And if so would it need to be a 240v generator or could a 120v work?

Well, I've decided to run the Master on the LI batts, and run the slave on Lead  So one can charge while the other is running off batts on its own, then I can have the lead batts charge in the morning on its own charger.

The slave won't get used that much to effect the batts as much as the master, as I don't hit over 6k the offten, matter of fact during the summer, only the dryer and the hot water heat is the only two devices that would push me to useing the slave, and that would only last a hour on the dryer water heater I think only kicks on for like 10 mins at the most unless after I take a show then I think it takes 30 mins to heat back up, never really timed it. going to change it anyhow to a heat pump heater.

No.  Only SLAVE devices can be safely connected to the AC output.  (Generator, AC Mains, etc. are MASTER devices--they can't be pushed around, and must be connected to AC Input.)

The inverter can only sync to an AC Input signal.  In order to function with a MASTER device (generator, AC Mains, etc.), the inverter must be a SLAVE device, meaning that it follows the MASTER signal.  Only the AC input has full wave detection (polarity, etc.), not to mention a disconnect switch in case things go awry.

GTM mode is Grid-Tie Master--it's for grid-tie inverters only.  And in that case, the GS inverter is the master device, and the grid-tie units the slave devices.

Either 240v or 120v generators can function on the AC Input, just the inverter needs to be set up for the desired input voltage.

It's actually an automatic function...there is no mode setting for it.  If the inverter detects negative AC output power, it'll automatically switch into GTM mode.  (LCD Status line will switch to "GTM: Running (...%)" indicating the level of frequency shift throttle currently applied.)

If that is the case, I'd be better off running the motor/alternator setups I have. Current testing shows at about 50% at the worst unit to 75% on the best. Funny thing about the worst one is that it is only that bad when turning at lower RPM rates and as it climbed in RPM, the input to output ratio got better. 50.9% at 375 RPM  to 62.9% at 500 RPM on one setup. And 75% at 750 RPM which stayed the same at 1000 RPM on another setup. The diodes are happier at the 375RPM as that is only 150Hz while the 750RPM is 300Hz that they have to deal with. But that is what they make fans for.

Another question here that I'm not sure has been answered:

For:

• 240V charge / ATS function
• 120V charge / ATS function

Can AC input be hot plugged/hot unplugged fine? Will hot(un)plug trigger the inverter to react, and if so, what are the specific behaviors there?

I'm thinking of a workflow where I'm plugging into/unplugging from shore.

12 minutes ago, kazetsukai said:

Can AC input be hot plugged/hot unplugged fine?

Absolutely.  That's how I test it 😉. 

When you connect an AC input, the inverter will analyze it for 2-3 seconds before deciding whether or not (by settings/configuration) to switch over.  AC input voltage and frequency have to be within the required parameters for it to switch over--and if they're out of range, it will refuse to switch.

EDIT: The inverter also will sync it's output wave/frequency to the AC input BEFORE switching.

12 minutes ago, kazetsukai said:

Will hot(un)plug trigger the inverter to react, and if so, what are the specific behaviors there?

Currently, if the inverter's system Mode is set to "Normal" (default), it will instantly restart back into inverter mode when AC input is lost; total duration in my tests from loss of AC until full inverter regulation voltage is ~130mS.

The only part of the transfer that I haven't quite figured out is how the inverter should decide to switch back to DC battery inverter mode.  I'm thinking of an option:

• never switch back
• switch back when battery reaches FLOAT stage
• switch back if battery voltage exceeds BULK setting (i.e. MPPT solar input, assuming the inverter is set to a lower charge voltage)
• ???ideas?

19 minutes ago, Sid Genetry Solar said:

Absolutely.  That's how I test it 😉. 

When you connect an AC input, the inverter will analyze it for 2-3 seconds before deciding whether or not (by settings/configuration) to switch over.  AC input voltage and frequency have to be within the required parameters for it to switch over--and if they're out of range, it will refuse to switch.

EDIT: The inverter also will sync it's output wave/frequency to the AC input BEFORE switching.

Currently, if the inverter's system Mode is set to "Normal" (default), it will instantly restart back into inverter mode when AC input is lost; total duration in my tests from loss of AC until full inverter regulation voltage is ~130mS.

The only part of the transfer that I haven't quite figured out is how the inverter should decide to switch back to DC battery inverter mode.  I'm thinking of an option:

• never switch back
• switch back when battery reaches FLOAT stage
• switch back if battery voltage exceeds BULK setting (i.e. MPPT solar input, assuming the inverter is set to a lower charge voltage)
• ???ideas?

Having otions for FLOAT/BULK settings would be a good start, but I don't think its enough to be smart about using excess energy. Lets say its 1PM, the battery reaches FLOAT, switches to DC inverting. You don't know if you'll be getting solar generation in or not. Also, when does it switch back from DC battery mode? Lets say you deplete 1/4, 1/3, or 1/2 of the bank between 1PM and 8PM, now its sundown and oh, there goes shore power due to an outage, and you can't depend on solar until tomorrow.

I just don't think the data points are there for the inverter to make a smart choice on its own.

You don't have any mechanism for SoC calculation IIRC. I was planning to have the integration suite I've written send a signal to the inverter, switching to DC inversion based on the following variables:

• SoC calculations from a smart shunt (VE Direct data)
• Battery voltage
• Average AC draw
• Known solar data for the day up until that point
• Current known solar generation patterns and time of day
• Weather service API calls

And even then, post-decision (after the integration suite changes the inverter to DC battery mode), I may have it watch to see if I'm actually generating enough to offset the costs incurred from the inverter. Lets say the day goes from  sunny to rainy shortly after the inverter switches- oh, my solar output dropped to zero, I'm going to move change back to ATS function.

I really need to be able to talk to the inverter =).