SolarEdge Auto-Transformer W/ Sunny Island 6048

jakobw

Hi all,

First major post here, been a lurker for a long time. Have a question for any of you that have experience with SolarEdge systems and specifically their SEAUTO-TX-5000 Auto-Transformer.

I'm needing an Auto-Transformer to go between a single 120V Sunny Island 6048 and a 220V Sunny Boy 6000US. Typically I would look to Midnite or Outback for an auto-transformer but I'm curious if the SEAUTO-TX-5000 would be a drop-in replacement as it's significantly cheaper.

Unfortunately SolarEdge is fairly light on specs and when I asked them about a configuration like this (120V-240V step-up/phase duplication) they simply said they couldn't help or provide insight (I was just looking for a yes or no answer) as it would be a non-warrantied configuration. I figured that would be the case, no fault to them. They are just protecting themselves.

So I'm looking for anyone who has installed or used these on a SolarEdge installation and if there is any reason they couldn't be used on another system or in the configuration I'm looking at. Based on their manual and a couple of YouTube videos (yes, I know), they simply have a 3 wire hookup (120V-0-120V by the looks of it) and a 2-wire twisted pair from a builtin temp sensor.

Based on what I see/read I think it should drop-in without issue but perhaps I'm missing something. Any insight would be super here. Thanks a bunch.

Here are a couple of relevant PDFs:

https://d3g1qce46u5dao.cloudfront.ne...sheet_na_1.pdf
https://www.solaredge.com/sites/defa...tasheet_na.pdf

BB.

Do you have a link to your transformer?

Auto transformers are not isolated, and frequently have odd taps to adjust voltages.

Bill

jakobw

Hey Bill,

Thanks for your reply. Unfortunately the best I can do is the PDF's I linked to above and for example these 2 videos from SolarEdgePV:

https://www.youtube.com/watch?v=ru0aNmrpvyo

https://youtu.be/r1nlaMzc7ww?t=355

Solaredge has essentially a passing comment about it on this page:

https://www.solaredge.com/us/products/StorEdge-products-for-backup-applications#/

I have no need for isolation for this specific installation and I'm wondering if the taps I'm seeing in the second video are multiple voltage taps like you mentioned however if this is a pre-wired end product I'm not sure what need they would have for multiple taps but I could be wrong.

I appreciate any insight you can provide. Thanks!

BB.

Here is our forum Host's website for the Storedge system:

https://www.solar-electric.com/solaredge-seauto-tx-5000-auto-tranformer.html

And here is a Application/Configuration note (not much more about the Auto Transformer):

https://www.solar-electric.com/lib/wind-sun/se_storedge_applications_connection_configuration_guide.pdf

And here are a couple AutoTransformers... One from Outback (two different models) and a SolarEdge:

https://www.solar-electric.com/search/?q=autotransformer

At least in the USA, the SolarEdge is the less expensive unit. The Solaredge unit is also a higher rated unit, and has thermal sensor too... I don't see any reason to not use the SE unit from that point of view.

Our host, NAWS (Northern Arizona Wind & Sun) does have engineers to consult... Perhaps they can answer your question about physical connections/issues.

The SE AT has a maximum limit of 25 Amps output per lead... So, that would work out to ~3,000 Watts/VA maximum @ 120 VAC.

I am a little confused (ok, a lot confused) about how you would use/justify the use of a 6,000 Watt GT inverter on a circuit that can only do 3,000 Watts max on the 120 VAC L1/N circuit through the Auto Transformer (given that the Sunny Island is a 120 VAC only inverter).

I am certainly way out of my depth here regarding the use of 5,000 Watt/VA (at 240 VAC) auto transformer...

-Bill

jakobw

Thanks Bill!

I may have made a mistake on my calculations but based on for example Midnite's autotransformer document, being a non-isolated transformer, power on L1 from the Sunny Boy would pass directly to L1 of the Sunny Island. Power on L2 from the Sunny Boy would pass through the transformer, change phase 180 degrees, and also provide power to L1 with a matching waveform. So assuming a 6000W load on the L1, there would only be 3000W passing through the SE AT.

Now that explanation is not entirely accurate/complete due to the fact the transformer will be providing load balancing at all times so, for example, a 2000W load on L1 would mean 1000W on both L1 and L2 of the Sunny Boy with 1000W passing through the SE AT.

Again I could be mistaken about this but that's what I'm currently thinking. Any thoughts on if this sounds correct?

I may see about speaking with NAWS engineers to confirm, appreciate the pointer there.

Going to have a look at those documents....

jakobw

Oh one additional comment. The reason I'm using SB 6000US inverters is that I already have several of them. They were taken out of a previous installation. Now just trying to reuse them to build low-cost AC-Coupled off-grid systems for a few rural homes with Sunny Island 6048's.

BB.

For SI to 120 VAC loads, the AT "does not do anything".

Where current (power) will flow is if you have, for example, your GT Inverter connected to L1/L2, and you pull power from L1/Neutral, to either backfeed the SI to recharge the battery bank, and/or to supply your 120 VAC loads on L1/N. And if you have 120 VAC loads on L2/N circuit too. As long as L1/N/L2 current flows are never more than 25 Amps each, you should be OK.

Note that only L1 and L2 need a double pole breaker on the Transformer leads (the type used on 120/240 VAC split phase North American branch circuits). You do not need a breaker on the Neutral Line, as N=L1-L2 currents.

The SE Inverter configurations show that they placed the double pole breakers on the L1/L2 output to the AC loads (not on the Auto Transformer directly.

Probably need to think about the A or B location of the 25 amp breakers... The 6048 SI Inverter is capable of something like 50 Amps max continuous output, and I would be suggesting something like 50a*1.25NEC-derate=~63amps (to avoid false trips if you really did have loads in the 40-50 amp range on the 120 VAC circuit. Following the Sunny Island Inverter manual/installation guides as always to ensure SI is "safe".

If you had a 25 amp single pole breaker from the 50 Amp SI Inverter output, that 25 amp branch circuit goes to the Auto Transformer, And L1+L2 auto transformer to loads/GT inverter 240 VAC output... That limits the SI Inverter to 25 Amps @ 120 VAC maximum on the AT branch circuit (combination of 3,000 Watt output "on battery" and a maximum of 3,000 Watts charging from the GT inverter(s) (25 amps @ 120 VAC).

You do have the possibility of SI Inverter 3,000 Watts + another 3,000 Watts (or more) from the GT inverters on the 240 VAC L1+L2 circuits (max of 25 amp difference between L1 and L2 for Auto Transformer limit)... But that is only when there is enough sun (mid-day), and the SI Inverter is not drawing 120 VAC power for charging.

Are you planning on having "oversized daytime only loads" (like water pumping with a 240 VAC pump)? The configuration between the different inverters and externally supplied power--Gets complicated. The Sunny Island is certainly very configurable... I don't know about the details or limitations.

-Bill

jakobw

Thanks for that explanation! That's basically what I had in mind however I am wondering about 1 single point where I think I may be miscalculating and would like to check.

Assume 6000W available PV and the batteries are full. If I should pull 6000W (50A) on L1 at 120V, it will initially come from the SI however as the SB ramps up, the load will be transferred. Unless I've made a mistake, a 50A/120V Load on L1 will pull 25A per leg (L1&L2) on the SB. 25A will be passing through the AT to L2 and 25A will be NOT be passing through the AT but be directly loading L1 on the SB.

However, where I would get into trouble is if I tried to load 6000W/50A on L2 as this would initially all be passing through the AT to the SI until the SB ramps up. 25A would be the max load here.

Am I thinking correctly here?

For the first installation, all the house loads will be connected to L1. No exceptions, the entire house is wired to a single 120V breaker box.

With regards to oversized daytime loads, the only consideration is a small 220V A/C (not installed yet). Assuming my calculations above are correct, when we go to buy the A/C I'll calculate if a single AT will be large enough to run the A/C and if not, perhaps add a second AT in parallel similar to how Outback recommends paralleling the X240.

Any reason paralleling them would be a bad idea?

Thanks!

BB.

I want to be very clear... The SB Inverter is 120 VAC... And in the USA, we would say there is an L1 and a N (Neutral) output). Neutral is typically ground bonded (grounding rod, cold water pipe, etc.). L1 (and L2, etc.) are "Hot" with respect to ground/neutral...

So, from the SB Inverter, you would have L1 and Neutral, 50 Amps maximum or 6,000 Watts (P=V*I=120 Volts * 50 Amps = 3,000 Watts).

The 240 VAC GT inverters have an L1 and L2 output, 240 VAC between them. And usually no Neutral output (there may be a neutral connection for a 120 VAC to Neutral "sanity check", but there is not "split phase" transformer output to Neutral). So, if you had 6,000 Watts of solar, that is 6,000/240 VAC = 25 Amps in L1 and L2.

The Auto Transformer has three power connections. L1, L2, and Neutral. The SB Inverter connects to L1 and Neutral (at 120 VAC), and the GT inverter (high current connections) connect to L1 and L2 (at 240 VAC).

And the transformers are constant power devices... Power In = Power Out... And Power=Voltage*Current.

So, if you dump 6,000 Watts @ 240 VAC into the Auto Transformer, that is 25 amps on L1+L2. If you try to pull 6,000 Watts out, then that is 50 Amps on L1+N (6,000 Watts / 120 VAC = 50 Amps) going to the SB inverter/120 VAC "main panel". But the Auto Transformer is only rated for 25 amps per leg maximum.

I have to be a little careful here... Your GT inverters are rated at 6000 Watts, and the Auto Transformer is only rated at 5,000 Watts total/full load (or 5,000w/120=20.8 amps across L1+L2). Or 3,000 Watts (I think) across L1 and Neutral (120 VAC).

Note that L1 and L2 lines are just arbitrary designations... L1 is "no different" than L2 functionally... Just trying to keep using the same designations in the discussion to avoid even more confusing (like saying L1 and L2 on the SB Inverter AC output--I would call them L1 and Neutral @ 120 VAC... Where as L1 and L2 are 240 VAC across them... And L1 to Neutral, or L2 to Neutral is 120 VAC. And Neutral being the "center tap" of a transformer (as used in North America).

The only way you can get 5,000 Watts "through" the Auto Transformer (not 6,000 Watts), is to connect a 5,000 Watt source across L1 and L2 (240 VAC). And connect your 3,000 Watt (maximum) 120 VAC loads across L1 and Neutral for Branch Circuit 1, and a second Circuit 2 across L2 & Neutral--Also 3 kWatt maximum @ 120 VAC.

If you have a 240 VAC load and a GT Inverter @ 240 VAC source, then there is (near) zero current flowing through the auto transformer. L1&L2 current from GT Inverter to 240 AC load (such as your inverter).

If you have no GT inverter power (dark, etc.), then the SB inverter would supply 1,200 Watts to L1&N, (1200w/120v=10amps) and the Transformer would supply 1,200 Watts to the AC load through L1&L2 connections (1200w/240v=5amps through L1&L2).

And since the Auto Transformer is limited to a maximum of 25 Amps per leg, the input power the transformer can manage is 25a*120v=3,00w from the SB Inverter.

https://circuitglobe.com/what-is-an-auto-transformer.html

This drawing shows a "stepdown" confuguration, but Transformers are bi-directional, so it can also be a step up too. Your GT Inverter is on the left side (120/240 VAC L1/N/L2) and the right side is just 120 VAC to your SB Inverter (L1/N).



-Bill

jakobw

Thanks Bill. Appreciate that, very clear. I was/am still coming up to speed on transformer calculations and today I see where I was coming up short.

After some further research I'm wondering if SolarEdge is undersizing their specs a bit. I say that because I think they are basically matching the transformer specs to the max output of the inverter it's designed to go with.

This may not mean much but it weighs more than the FW-X240:

SE AT 29.7 lbs
FW-X240 28.4 lbs

I may get one just to do some testing and see how the temp rises under load.

In an effort to ensure I'm understanding correctly, I've also been digging through the literature and specs of the 6000W Midnite AT that SMA recommends and the Outback PSX-240/FW-X240. I noticed they were using 25A breakers on both lines.

http://www.midnitesolar.com/pdfs/Auto_Transformer_Manual.pdf

https://www.solaris-shop.com/content/PSX-240%20Specs.pdf

I pulled this from the Midnite manual:



After reading that and more about AT amperage calculations, this is what I've arrived at:




Based on my calculations the combined output of the 2 25A 120V windings in series would be 50A between L1 and the neutral tap. However, this doesn't trip either 25A breaker as only 25A is passing through each one. I could be wrong on this though.

Here are a few things I pulled from the SMA Smartformer manual:






It looks like the SMA Smartformer is using 2 120V 3.4kVa windings in series and may not be much larger than the SE AT. I'm just guessing here.

Thanks again for the detailed explanation, several points I hadn't thought of.

BB.

Just for information... Here is a nice Outback Application guide for their auto transformer:

http://www.outbackpower.com/downloads/documents/integration_products/autotransformer/manual.pdf

Note that the windings used for the "Center Tap", are both brought out separately, and tied together at the bus bar. So, in this case, there are only 25 Amps on each lead entering/leaving the auto transformer. And not violating the 25 Amp per leg rule... (I think that is correct, I have not worked very much with Auto Transformers, so I cannot claim that this is an accurate statement or not).

Note that the "50A=I2" current path in the "Center Tap" connection (Neutral, typically "ground bonded neutral") in our discussion.

Note too that there are no breakers in the Center Tap/Neutral leads... In general, you do not want to circuit breaker "neutral" leads, If the neutral connection is opened, your 120 VAC appliances (if any) on the L2 and L1 to Neutral circuits can see anything between zero volts and 240 VAC. This is true of 120/240 VAC split phase power... A bad neutral connection in the main panel from the utilty drop to the main neutral bus bar will cause the same problem (0-240 VAC across any 120 VAC branch circuit).

The reality is that you will probably never use 100% of the 3/6/whatever kWatt ratings for your inverter/hardware (other than starting a well pump, etc. for short periods of time). I like to be conservative, so suggest that wiring/breakers/etc. be designed as if you are using 100% of rated power--Just to be safe. HOWEVER, when working with solar power (like your GT inverters, Battery Charging, etc.), you can expect many hours of running at rated output (i.e., 6 kWatt from a 6 kWatt GT inverter with >6 kWatt Pmp rated array, charging a battery bank from 50% state of charge, etc.)--In those cases, then it is very easy for an off grid battery based power system to run at rated load for an extended period of time (and why I highly suggest 80% or 1.25x NEC deratings. NEC is pretty conservative, but not for 100% of rated power for hours on end). For example, the typical NEC type breaker will usually never trip at 80% of rated current, but is supposed to trip at 100%+ of rated current (may take many minutes to many hours to trip). Don't want "false trips" to happen to you or your system:

• 15 Amp Breaker * 0.80 NEC derating = 12 Amps max continuous circuit loading
• 15 Amp Desired continuous current * 1/0.80 = 18.75 Amp minimum rated branch circuit/wiring/breaker (round up to 20 amps)
• 1/0.80 = 1.25 derating factors

In general for engineering, things that are less than 2x different are "almost the same". Things that are 10x different, that is a "night vs day" difference (and usually can ignore the 1/10th portion for calculations).

Also, be aware that Power equations for Voltage and Current are "squared" (P=V^2/R=I^2*R)... So that if you "double the current", you get 4x the power (or self heating effects in wiring/transformers due to resistance). So, in those cases square root of 2 (=1.414) can be used as the ceiling estimate of "don't care"...

In the case of a 30 lb vs 28 lb transformers--That is pretty much no difference. Of course, the design matters too (type amount of steel, amount of copper, cooling air flow, optional fan, etc.).

It would appear that your usage of 25 Amps max on the L1/L2, and 50 Amps on the Neutral/Center Tap can be correct... As long as the Auto Transformer center tap wiring comes out on "Two Wire" to a bus bar that is rated to be connected to a 50 Amp Center Tape Load. And, of course, the 240 to 120 VAC to Auto Transformer connection is rated for 50 Amps (50 amps from L1 120 VAC side, 25 Amps to Auto Transformer Wiring, and 25 amps to 240 L1 wiring).

And it probably does not matter in a larger installation like yours... But you do have transformer losses... If you assume 20 Watts typical no-load power consumption for the Auto Transformer, and run it 24 hours per day:

• 20 Watts * 24 hours = 480 Watt*Hours per day

That is about 1/2 the amount energy a refrigerator uses, or a good size laptop could consume if running most of the day.

With all of the above caveats, it does look like you could do what you needed.

Of course, this is your system and your home. If anything we talked about here does not look right, ask more questions, do further research. I (we) want you to be safe, and not having the experience/being there, operating at the "edge", etc., I cannot be sure of your installation/assumptions made/etc.--You are the person that needs to be 100% sure of what you do is correct and safe for you and your family.

-Bill

mike95490

You need to be very careful with transformers if you do not understand them well.  

Such as - the bottom winding of the transformer  (next to the label i2-i1) is only rated for 25A. You cannot get 50A out of it,
without cooking the transformer. (Unless the labels are wrong,)  There are no polarity indicator dots that show any easy way to split the winding.    All this shows me is a 25A step down from 240V - 120V  The amps do not double the way it is drawn.

jakobw

Thanks Mike.

I should have been more clear on my math and the diagram. 

Here is my math and the revised diagram. If anyone sees any mistakes, just let me know.




Thanks Bill. I appreciate that thorough explanation! Most of that I knew but a reminder is always super and will help someone else down the line as well. I supposed I should mentioned in the beginning that I'm studying electrical engineering (PV and ESS installations as a secondary profession) but hadn't come up to speed on transformers yet. I now see how simple they are after reviewing the last couple of days. Hence the above math.

BB. said:

Note that the windings used for the "Center Tap", are both brought out separately, and tied together at the bus bar. So, in this case, there are only 25 Amps on each lead entering/leaving the auto transformer. And not violating the 25 Amp per leg rule... (I think that is correct, I have not worked very much with Auto Transformers, so I cannot claim that this is an accurate statement or not).

Yep, exactly. Your 100% right on this or at least that's my opinion.

I'm thinking about just using a 5kVa 480x240 240/120 generic transformer and using the secondary windings in series. Almost cheaper than the SE AT but significantly heavier which is a downside should I order from fleabay. A 3kVa would do it but I don't want to load it that heavily during the charging cycle when the SB is at max output (6000W). If I was to try it, I think I would only use it with a max of around 3KWp like you suggested. 

BB. said:

In the case of a 30 lb vs 28 lb transformers--That is pretty much no difference. Of course, the design matters too (type amount of steel, amount of copper, cooling air flow, optional fan, etc.).

Yep, I was just tossing that out there but in hindsight (and right after I wrote it) I can definitely see that. To many factors to really make an accurate comparison based on weight alone. Here is a couple of screenshots I pulled from one of their installation videos. I'm still pondering their internal wiring of the transformer. I could be wrong but are we looking at 4 windings? 2 series - 2 parallel? If anyone has any further thoughts on this I'd be very curious to hear them.

mike95490

The polarity markings  (dots) on the transformer are critical - do it right and you get 50A output, wrong - you short circuit the works and get lots of smoke.  A 6KVA overload is going up in flames real quick.   Please be real careful, and you can try using a much smaller transformer, maybe a 24V bell transformer, it may have enough power to to drive the 240V input with no loads, so you can use a volt meter to measure +1.2V & -1.2V and make sure when connected, it's really in parallel and you still get 1.2v and not 0V.

jakobw

mike95490 said:

The polarity markings  (dots) on the transformer are critical - do it right and you get 50A output, wrong - you short circuit the works and get lots of smoke.  A 6KVA overload is going up in flames real quick.   Please be real careful, and you can try using a much smaller transformer, maybe a 24V bell transformer, it may have enough power to to drive the 240V input with no loads, so you can use a volt meter to measure +1.2V & -1.2V and make sure when connected, it's really in parallel and you still get 1.2v and not 0V.

Thanks Mike, I may try that.