Safe Grid Use of the 5000ES and transformer

[kjswiley]

That is what Will points out, the sticker still show 5000 ES, the user manual is shown as 5000 ES.

Yeah caught it later as I caught up...what a cluster F. Updated my post to ignore the question.

 

[Bud Martin]

We bought "ES" version per Signature Solar, it is even on the receipts. That is not jumping to conclusions that is being given incorrect information straight from the manufacturer we bought them from. They should figure out what they are selling.

SS needs to at the very least put the sticker on the unit showing what you bought, if you have two of them side, one from SS and one from other vendor, how can you tell which one is the US model? They do not even have the manual for the 5000 US.
So basically SS has about 5000 of SPF-5000 ES out there with the 5000ES sticker and 5000ES user manual to go along with it, just imagine that some one will be working on the installed system done by previous installer and try to troubleshoot the problem based on model as shown on the sticker and on the user manual!

 

[hipringles]

He should also show us his user manual, see if it says SPF-5000 US or not.

All of my paperwork/stickets from SS show 5000 ES. No where on my invoice, stickers, manual, does it say US

FWIW I will not be using mine in passthrough mode ever based off this. And when I get the LV's these are going to just power a 240vac mini-split offgrid in my garage.

Waste of money, but I mostly blame myself for not digging deeper than the info on SS website

 

[Samsonite801]

You can only by U.S model from Signature Solar

To me it doesn't matter what these Growatts are or not, I've seen so much garbled information and documentation that is clear as mud, to the point that, if I did buy any of these 5k units, no matter what they said on the plate or the label, I would still take it apart and see for myself how the neutral is grounded to the frame, I wouldn't necessarily trust that it has adequate separation to where a temperature change could make the board expand or flex and ground back out on the pedestal or something.

I'm not saying this is a bad unit to use for me in some application (I am an advanced user and can figure out how to safely install it), but I would rather not try and recommend this kind of setup to other new-to-the-industry kind of folks, other than to say use the EU (single phase output) inverters in USA at your own risk, which is what I've always kind of thought about this hack anyways trying to use wrong equipment when there are better choices out on the market already for same or similar money.

Even the supposed US model for this, the 5000 US or whatever they call it, well does it support split-phase out of the box? So how does that mean it is a model made or intended for the US? No... I prefer to just recommend that a DIY'er in the US should just use stuff that operates in split-phase without needing any external add-ons, it stays easier and less convoluted that way. I like to keep things simple in my recommendations to others.

Again, if somebody gave me one of these to have for free, I might go ahead and try to learn how to wire it right (i.e. safely) in a split-phase setup using the autotransformer (I am off-grid), and even one example, I have a friend who keeps trying to say he might want this Growatt/autotransformer solution for his off-grid lot, but I keep saying why bother, why not just use a pair of LV6548s or a single LVX6048 (or stack up to 9), since they just work split-phase right out of box? I think I finally have him onboard to the MPP inverters though now.

It's simpler that way, than to try and teach people how to put a square peg into a round hole, especially a beginner to this DIY solar space.

(It's the wild West out here)

 

[robby]

NEW VIDEO BY IAN

Thanks Ian.
I hope some of these naysayers finally understand what they are dealing with.

 

[Mr-Sandman]

That's a good educational video, that should scare the hell out of most people using this setup.

Lessons for those that don't take the 25 mins to watch all of Ian's video or understand what his trying to tell you (I think he did a good job of explaining this).

1. Autotransformers should NOT be used in systems that connect to the grid "OFF-GRID Use ONLY"
2. If your autotransformer quits functioning for any reason, half the crap in your house is probably going to get fried with 240 volts. (IE breaker trips to it, thermal shutdown, bad wiring, or it simply fails)
3. These 240v non-split phase inverters used with autotransformers probably are NOT a good idea. Yes they work, but the potential for disaster is one breaker trip or autotransformer failure away.

If you already have this setup and can't afford to get a proper inverter, I'd recommend at a minimum:

1. DO NOT put a breaker on neutral legs. This is given, but I had to say it.
2. Make sure the breakers hooked to the L1 L2 of autotransformer are large enough to handle full load on either side. Update: It might be best to just hardwire the autotransformer to the panel, you never want it disconnecting. You risk overcurrent to it, if you have more inverters than autotransformer rating. Choose which is less risky I suppose.
3. If you do have breakers on the autotransformer, put the inverter OUTPUT on a breaker and Install a breaker trip bar across the 240v inverter output breaker and both legs of autotransformer breakers. That way if the autotransformer was to trip a breaker, the OUTPUT from your inverter will trip at same time. The goal here is you want the inverter 240v output to be killed if any one of the legs off the autotransformer was to fail or be shutoff.

BTW, this autotransformer issue is not specific to Growatt inverters. The discussion of the bounded grounding screw in the Growatt 5000ES is a separate but related safety concern.

 

[Shale MacGregor]

That's a good educational video, that should scare the hell out of most people using this setup.

Lessons for those that don't take the 25 mins to watch all of Ian's video or understand what his trying to tell you (I think he did a good job of explaining this).

1. Autotransformers should NOT be used in systems that connect to the grid "OFF-GRID Use ONLY"
2. If your autotransformer quits functioning for any reason, half the crap in your house is probably going to get fried with 240 volts. (IE breaker trips to it, thermal shutdown, bad wiring, or it simply fails)
3. These 240v non-split phase inverters used with autotransformers probably are NOT a good idea. Yes they work, but the potential for disaster is one breaker trip or autotransformer failure away.

If you already have this setup and can't afford to get a proper inverter, I'd recommend at a minimum:

1. DO NOT put a breaker on neutral legs. This is given, but I had to say it.
2. Make sure the breakers hooked to the L1 L2 of autotransformer are large enough to handle full load on either side.
3. Put the inverter OUTPUT on a breaker and Install a breaker trip bar across the 240v inverter output breaker and both legs of autotransformer breakers. That way if the autotransformer was to trip a breaker, the OUTPUT from your inverter will trip at same time. The goal here is you want the inverter 240v output to be killed if any one of the legs off the autotransformer was to fail or be shutoff.

BTW, this autotransformer issue is not specific to Growatt inverters. The discussion of the bounded grounding screw in the Growatt 5000ES is a separate but related safety concern.

I am one of the noobs that did research on one set of products(Victron Quattro European), then thought he knew enough and applied it to another set of products(5000se). I am using 3 5000se and 1 Victron Autotransformer. This is an off-grid setup. so I am clear on #1... For #2, The Victron AT is wired directly into the Panel as L1 L2 and Neutral. Are you saying that the AT could fail in a way that still passed L1 and L2 through but not the neutral? #3 The Autotransformer has a breaker built into it, so if it trips it wont pass on any of the legs or neutral.

Am I thinking correctly in the above? If not please clarify if you can. I am going to consider what to do next, but not sure I should resell these to people unless I understand if it is even possible to safely use them in an off grid split phase situation.

 

[Hedges]

Are you saying that the AT could fail in a way that still passed L1 and L2 through but not the neutral? #3 The Autotransformer has a breaker built into it, so if it trips it wont pass on any of the legs or neutral.

If an open circuit occurs, e.g. breaker to auto-transformer if NOT set up as you describe, or wire to auto-transformer, or winding of auto-transformer, then there is a problem.

Connected as shown in image of video couple postings ago, if bottom winding opened, "Load" is connected to 240V through an inductor. At zero load it will see 240V. At modest load, voltage gets pulled down to near zero. (Inductance of transformer winding won't pass much current at all, perhaps 10 mA, unless second winding is able to carry equal and opposite current.)

Connected instead to inverter with two floating leads L1 & L2, centertap connected to Neutral bar of breaker panel (which is probably grounded, it converts 240V to two 120V phases, 120/240V split-phase. If any one winding of auto-transformer becomes disconnected, what you get are loads on phase 1 connected in series with loads on phase 2, and 240V across them.

If that's one, 100W incandescent bulb on phase 1 and one, 100W incandescent bulb on phase 2, they will agree to divide voltage equally, and both will light normally on 120V. Zero current flows in neutral wire because equal current from one load to the other. Same as if auto-transformer still connected; zero current in neutral.

If one 100W incandescent bulb and one air compressor, almost all the voltage, 240V, is dropped across the lightbulb. It burns brightly, for a very short time. Meanwhile, compressor experiences brownout.

This is exactly the same thing that happens with house wiring if convenience outlets are wired with 12/3 + ground Romex to two breakers. Black Line 1 and White Neutral feeds some circuits, Red Line 2 and White Neutral feeds other circuits. If a wire nut comes loose and disconnects White Neutral, same uneven division of voltage as with disconnected autotransfomer.

I think a 25A breaker feeding a sub-panel with hardwired auto-transformer is a reasonable thing to do. If over-current trips breaker, power to loads is shut off.
If there is a thermal protection device on the auto-transformer, that needs to go to a remote-trip of the breaker, not just interrupt power to auto-transformer the way it would for a heating load.

For 3-phase, there are "lost phase" detectors to turn off power. Something like that could serve as added protection. But it is probably wired as additional circuits which, if they became disconnected, would once again not provide protection.

 

[Mr-Sandman]

Honestly, I agree with Hedges, it sounds like it would be best to hardwire the autotransformer to the panel. You never want them disconnecting. If there is a short or overcurrent situation, you want the supply current to trip "Inverter output to panel". That is probably safer than trying to use breaker trip bars across those breakers.

Of course if you have more than one of those 5000 watt inverters hooked up, you have the potential to overcurrent the autotransformer wiring before the inverters breakers trip.

 

[JohnGalt1717]

Just add another or multiple SCC's. You don't have to run all PV thru the inverter MPPT.

Any suggestion on a good high voltage MPPT? Preferably one that allows to be define battery voltage much higher than 48V? I can't find any stand alone MPPTs on the market ATM. Everything that is high power is integrated into the inverter and the inverters that are high power all have serious limitations too unless 240 only.

 

[JohnGalt1717]

Not if you are off grid, UL is not a requirement

That isn't true. UL 1310 isn't required. UL 294 still applies as does NEC standards on high voltage lines both because of the 240VAC and because the inverter boosts 48VDC to 340VDC to supply the inverter to get 240VAC out without a transformer. And it does so with PWM which makes it pulsed DC which is regulated.

 

[JohnGalt1717]

Thanks Ian.
I hope some of these naysayers finally understand what they are dealing with.

I have to say, after watching these videos, and even looking at the units that do split phase, it really seems to me that AC passthrough is a bad idea.

The better solution in my mind is to rectify the AC to 340VDC using a ideal rectifier (4 fets pulsed which can also apply PFC) which would generate a 0.5% loss, and then feed that into the SPWM inverter.

I.e. you have the following:

1. AC/DC mains input using a push/pull or Resonant LLC with transformer ideally that is isolated and UL listed. This outputs 340VDC to a internal DC bus. (about 95% or so efficient)
2. Battery to DC Bus that boosts using a full bridge design at 98% efficiency whatever your battery voltage to 340VDC, which is required for the inverter anyhow. (this design is symmetrical, which is different than what is in there now which is forward only)
3. MPPT that boost/buck as the second stage to the 340VDC bus internally. I.e. you use a half bridge to MPPT with shift and read, and then whatever the output of those maximized watts are, you buck/boost to 340VDC. This would actually be significantly more efficient with high voltage strings than MPPT to 48VDC AND once it's at 340VDC if you're running off of solar there is no further conversion and any excess after the inverter is taken care of will go out to the battery using #2.
4. Dual 120VAC outputs capable of taking 100% of the wattage from the internal DC bus as needed to either leg. (and international markets would just be 240VAC doing the same).
5. Output 340VDC (you can hack heat pumps and anything else that advertises "DC Inverter" to bypass the rectifier (or use the rectifier harmlessly) and feed this in without having to have huge inverters and save yourself a TON of in efficiency. Same with hot water heaters),
6. Output 120VDC (ELV) for things like high dryers etc. (The great problem children of design, but ELV is unregulated up to 120VDC with less than 5% ripple)
7. Output 48VDC for USB C 240V plugs. (I'd expect virtually all devices in the next 5 years like TVs etc. to only come with a USB C input because then they don't need UL certification at all, just FCC and any wall wart will work. If you switch the wall to be coming from 48VDC anyhow with USBC ports then you don't need the inefficient wall wart, the inverter nor the rectification. USB C will output 3.3, 5, 9, 12, 18, 20, 24, 28, 36, and 48VDC so you have everything that anything else DC would need so everything with a wall wart would be covered and work.
8. Build an on wall DC/DC cart charger for CCS Type 2 that boosts/bucks the 340VDC to whatever the car DC voltage request is at 98+% efficiency thus eliminating the entire need for inverter and then rectification just to charge a car making it 15-25% more efficient which is a big deal on solar off grid.

If you did this, you'd be VASTLY more efficient and need VASTLY less inverter power which would save a TON on the cost. (i.e. half/full bridges are always VASTLY cheaper and more efficient to build so even if you do need to convert voltages with DC, it's always cheaper than the copper in a transformer to do it purely AC and more efficient and easier than SPWM inverters.

This is a FET design in a pulsed environment so using GanFETS (600-900VDC max) for most of the design, and SICFets for the MPPT (1800VDC max) works really well and isn't very expensive because the pulsed continuous load rating on these FETs is in the 100s of amps per FET on a 6MM copper board trace.

Honestly thinking about designing one of these as a modular design in a rack... The only part that gets expensive is the AC/DC mains input because of the UL and FCC certification to do it right. Given ENNOID-BMS, this becomes super compelling in design because we can easily scale our batteries to higher voltage with this design and reach better efficiencies on the bus over time as we can afford more batteries, while minimizing inverter power required.

 

[Mr-Sandman]

Any suggestion on a good high voltage MPPT? Preferably one that allows to be define battery voltage much higher than 48V? I can't find any stand alone MPPTs on the market ATM. Everything that is high power is integrated into the inverter and the inverters that are high power all have serious limitations too unless 240 only.

Schneider Electric Conext MPPT Solar Charge Controllers

The Conext MPPT solar charge controllers come in 60A, 80A, and now 100A models. Get the most power out of your home solar system and improve battery life!

www.altestore.com

It's a good one that is 100% adjustable DC output to batteries up to 67 VDC. Very high quality, although pricey and you may need the Insight Home gateway to configure and monitor it.

 

[Zwy]

Any suggestion on a good high voltage MPPT? Preferably one that allows to be define battery voltage much higher than 48V? I can't find any stand alone MPPTs on the market ATM. Everything that is high power is integrated into the inverter and the inverters that are high power all have serious limitations too unless 240 only.

Define high voltage. 250V max works well, let's take #8 wire, 200 feet of distance with 4 to 6 conductors in conduit. 3% allowable voltage drop. You can run 19 amps and load voltage would be 194V.

While you think the higher 450VDC would be advantageous, you have to remember the inverter has a max wattage rating. A balance needs to be found between amps/volts where wattage remains within limits. It doesn't work to have 450VDC but can only carry 10 amps or less.

Many are using this MPPT. https://watts247.com/product/sc48120/

Note that it will take 250VDC now, the older models were 150VDC. Contact Ian for more information and tell him what your PV array will be.

 

[JohnGalt1717]

Define high voltage. 250V max works well, let's take #8 wire, 200 feet of distance with 4 to 6 conductors in conduit. 3% allowable voltage drop. You can run 19 amps and load voltage would be 194V.

While you think the higher 450VDC would be advantageous, you have to remember the inverter has a max wattage rating. A balance needs to be found between amps/volts where wattage remains within limits. It doesn't work to have 450VDC but can only carry 10 amps or less.

Many are using this MPPT. https://watts247.com/product/sc48120/

Note that it will take 250VDC now, the older models were 150VDC. Contact Ian for more information and tell him what your PV array will be.

I have 600W panels. I have 8 in series (and because of location I can' just do 4S2P). So I need 450V. My design had 4 strings of these which fit in the Growatt profile for wattage.

I would love nothing more than to find separate MPPTs that will handle this and then inverters that can be paralleled that have nothing to do with anything else and then one or more separate AC/DC chargers....

 

[Zwy]

I have 600W panels. I have 8 in series (and because of location I can' just do 4S2P). So I need 450V. My design had 4 strings of these which fit in the Growatt profile for wattage.

I would love nothing more than to find separate MPPTs that will handle this and then inverters that can be paralleled that have nothing to do with anything else and then one or more separate AC/DC chargers....

Let's get the specs on the panels.

 

[Struc]

@JohnGalt1717 @Zwy

Let's get the specs on the panels.

And let's put it in a different thread please?

 

[Zwy]

@JohnGalt1717 @Zwy


And let's put it in a different thread please?

Maybe, but one of the main reasons people buy the 5000ES is due to the high PV voltage input. This discussion in the thread is relevant as some like John don't see alternatives. If he wishes to start another thread and provide a link here, I'm all for moving the discussion. That way members can be aware of all aspects.

 

[JohnGalt1717]

Maybe, but one of the main reasons people buy the 5000ES is due to the high PV voltage input. This discussion in the thread is relevant as some like John don't see alternatives. If he wishes to start another thread and provide a link here, I'm all for moving the discussion. That way members can be aware of all aspects.

High voltage MPPT Charge Controllers preferably with flexible battery output?

So I've fallen victim to the Growatt scandal... and what I really always wanted was separate of everything from MPPT to inverters to AC/DC charging. I have 8x605Watt panels in series (and I can't parallel because of major losses going to 4S2P because of shading etc.) They're 48VDC max open...

diysolarforum.com

Parallelable SPWM (transformerless) split phase inverter without hybrid?

I'm a refugee from the ongoing Growatt 5000ES/US ... issue. Anyone know of a unit that doesn't have MPPT or AC/DC coupling in it? (or has a proper push/pull design like Ford Lightening have?) Preferably it would accept any voltage of battery from 42VDC to 400VDC since this design has to buck...

diysolarforum.com

 

[copec]

I have to say, after watching these videos, and even looking at the units that do split phase, it really seems to me that AC passthrough is a bad idea.

The better solution in my mind is to rectify the AC to 340VDC using a ideal rectifier (4 fets pulsed which can also apply PFC) which would generate a 0.5% loss, and then feed that into the SPWM inverter.

I.e. you have the following:

1. AC/DC mains input using a push/pull or Resonant LLC with transformer ideally that is isolated and UL listed. This outputs 340VDC to a internal DC bus. (about 95% or so efficient)
2. Battery to DC Bus that boosts using a full bridge design at 98% efficiency whatever your battery voltage to 340VDC, which is required for the inverter anyhow. (this design is symmetrical, which is different than what is in there now which is forward only)
3. MPPT that boost/buck as the second stage to the 340VDC bus internally. I.e. you use a half bridge to MPPT with shift and read, and then whatever the output of those maximized watts are, you buck/boost to 340VDC. This would actually be significantly more efficient with high voltage strings than MPPT to 48VDC AND once it's at 340VDC if you're running off of solar there is no further conversion and any excess after the inverter is taken care of will go out to the battery using #2.
4. Dual 120VAC outputs capable of taking 100% of the wattage from the internal DC bus as needed to either leg. (and international markets would just be 240VAC doing the same).
5. Output 340VDC (you can hack heat pumps and anything else that advertises "DC Inverter" to bypass the rectifier (or use the rectifier harmlessly) and feed this in without having to have huge inverters and save yourself a TON of in efficiency. Same with hot water heaters),
6. Output 120VDC (ELV) for things like high dryers etc. (The great problem children of design, but ELV is unregulated up to 120VDC with less than 5% ripple)
7. Output 48VDC for USB C 240V plugs. (I'd expect virtually all devices in the next 5 years like TVs etc. to only come with a USB C input because then they don't need UL certification at all, just FCC and any wall wart will work. If you switch the wall to be coming from 48VDC anyhow with USBC ports then you don't need the inefficient wall wart, the inverter nor the rectification. USB C will output 3.3, 5, 9, 12, 18, 20, 24, 28, 36, and 48VDC so you have everything that anything else DC would need so everything with a wall wart would be covered and work.
8. Build an on wall DC/DC cart charger for CCS Type 2 that boosts/bucks the 340VDC to whatever the car DC voltage request is at 98+% efficiency thus eliminating the entire need for inverter and then rectification just to charge a car making it 15-25% more efficient which is a big deal on solar off grid.

If you did this, you'd be VASTLY more efficient and need VASTLY less inverter power which would save a TON on the cost. (i.e. half/full bridges are always VASTLY cheaper and more efficient to build so even if you do need to convert voltages with DC, it's always cheaper than the copper in a transformer to do it purely AC and more efficient and easier than SPWM inverters.

This is a FET design in a pulsed environment so using GanFETS (600-900VDC max) for most of the design, and SICFets for the MPPT (1800VDC max) works really well and isn't very expensive because the pulsed continuous load rating on these FETs is in the 100s of amps per FET on a 6MM copper board trace.

Honestly thinking about designing one of these as a modular design in a rack... The only part that gets expensive is the AC/DC mains input because of the UL and FCC certification to do it right. Given ENNOID-BMS, this becomes super compelling in design because we can easily scale our batteries to higher voltage with this design and reach better efficiencies on the bus over time as we can afford more batteries, while minimizing inverter power required.

YouSolar makes a system similar to that that is stackable, but is is not relatively inexpensive DIY.

Edit: Also Sorry for the off topic, but https://www.dc.systems/ is also pretty neat.