SURT Smart UPSs

[MarkC_apc]

I have a few questions after study of the SURT operation, in particular the direct tie of Battery negative to the 240 VAC out "neutral", especially when in operation on 240 VAC "split phase" power common in US:

1. Are there any issues connecting the split phases of L1, L2/N, and ground (no neutral) to the incoming AC power - where L2 is connected to the L2/N (brown?) wire? (when the UPS is configured at 240 VAC)
2. If there are no issues with this, how is the 240 single phase (out) produced from the split phase incoming when on bypass?  Are both phases 
3. Are there potential issues with using the 240 VAC output from the SURT directly tied to a household circuit (that by code has the neutral that is "earth" grounded at the main panel)?   This assumes a transfer switch has disconnected the grid supply.
4. To make such a tie to the household circuits - is it imperative to use a SURT003 iso-transformer that produces split phase or can any transformer that makes 120 VAC be used?

My main concern is how the L2/N connection (at 120 VAC) is processed to become/or isolated to neutral on bypass so that it does not remain at the 120 VAC?  This is easy to understand when the incoming power is single phase - but difficult for me to determine the safety of such connection schemes on split phase power.

Any information is appreciated.

[MarkC_apc]

Anyone - Anyone??  Any tech support from APC that can answer this?  

[voidstar_apc]

Hopefully someone can provide a detailed answer on your questions.

Your main concern seems to be how L2 is isolated during bypass so it isn't "hot". My understanding is your premise is false... in split phase, both phases will be hot. And appliances designed to plug into a 240V source expects that and will not expose the user to either phase. Current flows from one hot phase to the other. That contrasts with the 120V world with a single hot phase and current returning through a neutral that is sometimes exposed, ie, a light socket. Because it can be touched, the neutral needs to be at ground potential on both 120V and split phase systems. My guess is if the 120V light socket were designed today, the neutral would've been protected rather than exposed.

[MarkC_apc]

Voidstar - I agree with your analysis of the use of "neutral" in US household circuit designs - that it can carry current especially since "unbalanced" split phase circuits are essentially always built in to the household circuit installations,  I believe this is why it is mandatory that the "neutral" be bonded at the main panel - to prevent significant voltages being built up at these "vulnerable" points where older designs actually expose the neutral (the socket).  

My question is about the specific use in the SURT design where it appears that a split phase input is somehow converted back to single phase 240 VAC output on bypass.  I'm having difficulty understanding how two 120 VAC lines that are in opposite phases in voltage can be combined into one single phase 240 VAC voltage (to neutral - not UPS ground) at the output.  To obtain the two 120 VAC phases requires a center tapped transformer at the "pole" with windings being tapped going "opposite" current flows to neutral to get the opposing voltage phases (my interpretation!).  

I'm worried that using the UPS output directly to a "household" system for 240 single phase users (small 240 VAC air conditioner for example) may create a problem with neutral/ground connection. 

[voidstar_apc]

The two phases (L1 and L2) are not 240VAC relative to neutral. Rather, the phases are 240VAC relative to each other. They're still 120VAC relative to neutral, but 240V equipment is using the two hots rather than one hot and a neutral. The only special thing on the UPS side is making sure the inverter produces two 120VAC outputs 180 degrees apart while on battery.

If you're familiar with audio engineering, bridging a stereo amplifier to get a higher voltage mono output is similar.

As for how the two 120VAC phases combine to form 240VAC, well, the instantaneous voltage at L1 is (ideally) a sine wave. And the voltage at L2 is -L1 (because 180 degrees out of phase). So when L1 is at it's peak, L2 is at it's nadir, and vice-versa. So if we measure the instantaneous voltage between L1 and L2, that would be V_L1 - V_L2. Previously we said V_L2 = -V_L1, so substitute that in and get: V_L1 - (-V_L1) = 2*V_L1.In the figure below, I've plotted L1 in blue and L2 in green. They're both 120VAC, just opposite in phase. The voltage difference between the two is in red... the 240VAC.

[wayne_i_apc]

Hi Mark,

Can I ask you to please back up and give more specifics about the model you are referring to? I have a SURT2000XLI, which is the "international" version, using IEC power connections, but it is here in the US. Thus, for European power, it has L1 (brown), N (blue) and Earth (ground in the US!) (green / yellow), and expects 230 VAC (nominal) from L1 to N. To "properly" connect to a split phase 240 VAC (US) system would require 4 wires, L1 (black), L2 (red), N (white) and G (green). So, does the SURT you are investigating have 3 or 4 wire connections? In other words, is your SURT designed for split phase 240 VAC operation? Since your question started by referring to 240 VAC, I am unclear what you mean by referring to L2/N and your mention of connecting to 120 VAC, and how that would affect bypass operation. As I currently understand it, if there is an L2 connection, it should be the red wire, not neutral, resulting in 240 VAC relative to L1, and 120 VAC relative to N.

Further, since the SURT is a double conversion topology, the output will be at the rated voltage independent of the input voltage, in other words, for a 240 VAC SURT, over an input voltage range of something like 100 to 280 VAC the output will be 240 VAC. Bypass operation is NOT possible if the input voltage is outside a limited range (210 to 260 VAC?), and certainly would not be allowed if the input is only 120 VAC.

All kinds of magic can be done with transformers, but you must consider the safety aspects of the ground and neutral connections for both the people using the equipment, and the equipment itself. It is possible that improper use of a transformer could lead to damage of the UPS.

[MarkC_apc]

Wayne;

The model is a SURT6000XLIX547.  It is GE branded, but definitely an APC unit that was used in the US.  The incoming SURT wiring box is labeled "L1" (Brown), "L2/N" (Blue) and green ground - that connect to the main board.  I had been told that to use on 240 VAC (US split phase), to wire L1 to L1, L2/N to L2 and ground to ground.  The neutral would not be used.  I do have the unit configured at 240 VAC. 

This SURT does not have any L14-30R outlets (I believe there is such a SURT, but quite rare?).  All are L6's.  So again, my concern is if this unit is claimed to be usable in the US with split phase incoming AC, can any of the outputs (which are 240 VAC single phase - L1 to neutral) be used in a US household circuit (where the neutral and ground are bonded)?  Or is an isolation/center tapped transformer required? (as it certainly would be if the 240 VAC split phase is needed).  Another possibility is that the SURT has circuitry that "converts" the incoming split phase to single phase when on bypass that would essentially isolate "L2/N" from the SURT's output neutral.

It would be nice to have a detailed wiring schematic of at least the entire power level wiring of these SURTS!

Hope this clarifies and does not confuse further!

Mark

[MarkC_apc]

Thanks - great illustration of many times a confusing concept of US split phase power!  

[voidstar_apc]

It sounds like your UPS is 240V only (not split-phase). So it will run on the 240V split-phase but the output is 240V only -- no neutral.

[MarkC_apc]

Voidstar;

Definitely it is the 240 VAC out - L6-20/30R's!  However, does the output remain split phase? It appears to be single phase - which was the origin of my question about how is this handled during "bypass" operation?

Would you have any issues using it for 240 VAC, split phase input to a "household"  (US - off-grid) sub-panel with a SURT003 iso/split phase transformer?  I know that it works fine with local plug-in users (L14-30R connection on back of SURT003) as the local users are "isolated" from earth ground - although their "shell" is grounded to the SURT003 case.  The question is if the neutral/ground bonding within the sub-panel would cause any issues (with the SURT6000XLI)?  This may answer the original question of using the 240 VAC directly out of the SURT6000XLI (no SURT003) with 240 VAC users (L1/L2 connected with ground,but no neutral).  An example of this "user" is a small 240 VAC air conditioner that has a 3 prong (dryer) type plug or the locking type (L6-30P).

Apologize this is taking so much blog space, but it seems I'm still not in full understanding of what seems to me to be "potentially" a safety issue.

[wayne_i_apc]

Mark, I for one, appreciate this discussion, because I am in a very similar situation. I use my SURT under very limited situations-- usually only power it up when there is a utility power failure to get my network up, run some lights and keep the refrigerator cold (beware the defrost cycle!). To do this I use a isolation / configurable step-up / step-down transformer to give me a choice of split phase 120 / 240 with a neutral connection (series connected secondary 1 (L1) to center tap (N) to secondary 2 (L2)) or I can parallel the secondaries to give me 2x current at 120. I have run my 240 water well pump directly off the SURT with NO ill effects (so far!) because my transformer doesn't have the VA rating needed for the pump. My well pump is 2-wire, so it there are no concerns about neutral or ground safety. On the other hand, the output of my isolation transformer remains fully isolated from any and all metallic connections to my regular home wiring, including ground.

I would be careful regarding your 240 VAC air conditioner with it's 3 prong "dryer type" plug. I have an older dryer that has a 240 VAC 3 prong (L1, L2, G) plug that uses the ground connection to run the timer on 120 VAC, which would NOT work on the output of a SURT since the output of the SURT has no neutral connection therefore cannot be "bonded" or make a neutral to ground connection.

Regarding the split phase versus single phase differences, there are three cases to consider-- 1) on utility, 2) on battery and 3) on bypass. I'm guessing right now, but I think--

1) While on utility power, the UPS will connect it's input neutral to it's output neutral, meaning in the case of the SURT connected to US split phase 240 VAC, it will connect input L2 to output L2 while leaving it's output L1 isolated from input L1. There will be NO neutral output connection available to provide split phase 120 VAC from the UPS though the L2 output measure 120 VAC to utility neutral, but ONLY while on utility power.

2) While on battery power, the UPS will isolate it's input neutral from it's output neutral with it's output L1 still isolated as before. It connect it's output neutral to ground to provide a bond connection. This would maintain the single phase 240 VAC "Euro power" (240 VAC to G) configuration that it was designed to deliver. This way, there will be NO neutral connection available to provide split phase 120 VAC from the UPS.

3) On bypass, the UPS will connect both it's input neutral to it's output neutral and it's input L1 to it's output L1. This means in the case of the SURT connected to US split phase 240 VAC, utility power L1 and L2 pass straight through the UPS referenced to neutral and ground identical to the UPS being totally removed from the circuit. This could be connected as typical US split phase 120 / 240 VAC, but would ONLY work when in bypass! (not very useful for a UPS set up!!!)

AND! Let's not overlook that you've determined that the batteries are sitting at 120 VAC when the SURT is connected to US split phase 240 VAC!

I suspect we've stepped WAY outside the way APC was thinking when they designed this device, I hope they've covered their eyes to this post! Meanwhile, let's work it through and get the confidence to use these UPSs as best we can to solve our needs!

[MarkC_apc]

Wayne;

I agree - any consideration for using these SURTS for any "US household" connections (especially 240 VAC), must be fully vetted - appears you have been very careful (as I feel I have also).  I have a scope and plan on looking at wave forms under various scenarios, on the output of the SURT6000 and the combo with SURT003 (iso transformer).  It'll take some time as the SURTs are not in my workshop.  I do have one that I'm attempting to repair, but in the analysis phase, not the repair phase!

Note that the SURT003 is not configurable, but is a good match to the SURTs 5k and 6k.  It's a simple, center tapped toroidal transformer - that weighs almost 100 pounds!.  

[MarkC_apc]

On 1/6/2017 12:08 PM, Wayne said:

I suspect we've stepped WAY outside the way APC was thinking when they designed this device, I hope they've covered their eyes to this post! Meanwhile, let's work it through and get the confidence to use these UPSs as best we can to solve our needs!

Hi Wayne:

Hoping that you are still interested in these high end UPS's as multifunctional units:

• Traditional UPS for limited backup systems
• "Second life" as inexpensive, powerful, well built inverters for such as PRIUPs  generators (most of the "failed" UPSs simply have new batteries needed for such a use!)
• Home backups for the adventurous - that have to deal with "split phase" panels and systems (with safety/regulations and transfer requirements all designed properly)

As such, the SURT and SRT lines, when used in a "household" system will need to be combined with one of the isolation transformers - that seems settled.  However, I'm still in a quandary about how the neutral from the SURT/SRT into the iso transformer is connected (or not) to the output neutral.  In the SURT003 design, the output neutral is bonded to the case ground (and G on NEMA 14-30R).  But, does this mean that the input neutral is connected directly to the output neutral?  Guess I'll have to check on my SURT003 (I'll do that).  The worry is that the DC "zero voltage" of these UPSs  are apparently direct connected to the AC system neutral.  This appears to be true for the +192 VDC SURT/SRTs up to 6000 VA and for the +/- 192 VDC SURT/SRTs 8000's and above.  I wonder if this requires more protection of this DC voltage everywhere in the AC "neutral" wiring system (and bonded ground).  Maybe DC rated breakers?  Actually, the APC DC battery cages all have DC rated fuses, however, a "home" systems might require special design attention?  

At this point, this is simply "musing", however I'd like to have a backup system to charge my Nissan Leaf using a solar/battery backup/UPS system and it will need to replicate the split phase with neutral and ground as if the charge controller is "seeing" grid power!  So far, without the iso transformer, the charge controller will not start up due to "no ground".  

Let me know if still monitoring this blog - The second life of these units - IMO - has some real "potential"  

[MarkC_apc · Accepted solution]

Wanted to update the PriUPS system that I've recently tested:

System consists of a SURT6000 and SURT003 isotransformer being energized at 200 - 250 VDC by my 2012 Prius.

Ran system about 2 hours on wattages varying from 1200 to over 4800 (110% rated with UPS alarming!!) on Heater strips with PF of 1.

Temp of SURT6000 went up about 10 oF above ambient - no signs of being stressed.

Set number of batteries at 10, however even with DC voltages consistently well above the 168 VDC cutoff, the low battery alarm went off in about an hour.  No shutdown problems.  Silencing the alarm is required!

The SURT300 has so many outlets, it's hard to imagine not having enough to plug in the usual required backup items - refrigerators, TVs, heater system fans (assuming gas fired), etc. etc.  with extension cords.

Using the isotransformer, it is very likely that a backup subpanel system could be installed for a more permanent house backup system.  This would require grounding  both SURTs, using the 4 wire L14-30R to power the subpanel (L1,L2,N, Ground to subpanel) - HOWEVER I HAVE NOT CONFIRMED THIS and do it at your own risk!!

I have some data logs if requested.

Mark