SW5548 not recognizing ?good? power on AC1

ronrosenfeld

After 14 years off the grid, we finally brought grid power to our property. I didn't really want to, but there was a "perfect storm" of reasons to do so. So I am now dealing with an issue that, of course, never presented a problem in the past. My setup:

• SW5548 x 2 Series stacked Updated to Software revision 4.10 when that became available.
• 1.98 KW PV Panels
• Outback MX-60
• Bergey Excel-R with VCS-10 Controller
• 12 Surrette 8CS25P 8V batteries in 48V bank
• 12kw Kohler propane generator (auto start controlled by the inverter)

There were no problems with the inverters when running off-grid.
Yesterday we "went live" with the grid, and on one inverter, the AC1 green light is blinking almost continuously. Occasionally, it will become solid green for 5-10 seconds (sometimes a bit longer). The other inverter seems to work as expected. When AC1 is blinking, I cannot sell to the grid on that inverter; it will not charge the batteries from the grid; etc.

Measuring the voltage at AC1 on each inverter shows about a 4V difference, with the lower voltage on the dysfunctional inverter -- but I have not seen it lower than 119.5VAC.

Any thoughts as to what might be the problem? I did try a factory reset, but that did not help anything. Could there be an adjustment on the "third" menu that might fix the problem?

Thanks for any insight.

westbranch

Re: SW5548 not recognizing ?good? power on AC1

Does it have to 'verify /qualify' the Volts, Phase or frequency of the grid? If so i the grid 'flakey'?
Are all grid related settings on each inverter the same?

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

As far as I know, Volts and Frequency. Grid-related settings on both inverters is the same.
I have not tried measuring the voltage with the inverters turned off. I don't have any instruments to measure frequency.
I don't know if the grid is "flakey". But it would be odd for that two show up on only one side, wouldn't it?

Shawn

Re: SW5548 not recognizing ?good? power on AC1

westbranch wrote: »

Does it have to 'verify /qualify' the Volts, Phase or frequency of the grid? If so i the grid 'flakey'?
Are all grid related settings on each inverter the same?

Edited to remove bad advice.
Please see later comment for better advice
Thanks to all that provided better troubleshooting ideas.

-Shawn

westbranch

Re: SW5548 not recognizing ?good? power on AC1

No Killl-a-watt meter?

I assume you have 2 power lines coming in on your grid feed (@ 120V) and 1 goes to each inverter and to each side of the Distribution box..??

ADD: good idea Shawn. that would show if there is a problem on one phase, as long as the inverter settings are matched.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

I don't know what you mean by "rolling the phases"

BB.

Re: SW5548 not recognizing ?good? power on AC1

Be very careful of "rolling wiring" on the input to the inverters... Are they "stacked" output? (i.e., in parallel for more current or in series for 120/240 VAC connections)? If they have "stacked outputs", rolling Lx phases could be unpleasant (just guessing--But would worry me a lot).

Also, I assume they have a common AC neutral on the AC outputs--Make sure that the AC input neutral and AC output neutral are property wired (did not swap Lx with Neutral).

-Bill

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

The main power to the house is single phase 240 VAC with two hots and a neutral, which I am told is standard residential power in the US. There is 120V between hot and neutral; and 240V between the two hot wires. One hot is connected to one inverter, and the other to the other inverter.

There is a primary line from the street to near our house (7200V) which is stepped down by a transformer to our secondary. There are no other connections to the transformer other than our home.

The inverter output is synchronized so that 240V appliances see a proper waveform. With regard to your question about the criteria for connecting to the grid, this is the data:

Attachment not found.

Blackcherry04

Re: SW5548 not recognizing ?good? power on AC1

It sounds like it may be qualifying the power , but the relay may be sticking or has flaky contacts and it cannot complete the transfer. When it's trying to qualify the grid, give the inverter a good rap / jolt with a rubber mallet or a piece of 2x4 on a corner , sometimes that will shake one loose. On the bench they will pick them up about a inch and drop them on the surface.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

BB. wrote: »

Be very careful of "rolling wiring" on the input to the inverters... Are they "stacked" output? (i.e., in parallel for more current or in series for 120/240 VAC connections)? If they have "stacked outputs", rolling Lx phases could be unpleasant (just guessing--But would worry me a lot).

Also, I assume they have a common AC neutral on the AC outputs--Make sure that the AC input neutral and AC output neutral are property wired (did not swap Lx with Neutral).

-Bill

Since I don't know what "rolling" means, I have no worries. They are "series stacked" to provide 120/240 VAC. Trace provided a cable which synchronizes the output.
My electrician is coming by tomorrow morning, and I will make sure he double checks the wiring. He's a competent and licensed master electrician, but I guess anyone can make a mistake. Also the 4V difference between the two AC1 (that's the input from the grid to the inverter) terminals is unexplained at this time. I'll probably recheck that with the inverters off.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

Blackcherry04 wrote: »

It sounds like it may be qualifying the power , but the relay may be sticking or has flaky contacts and it cannot complete the transfer. When it's trying to qualify the grid, give the inverter a good rap / jolt with a rubber mallet or a piece of 2x4 on a corner , sometimes that will shake one loose. On the bench they will pick them up about a inch and drop them on the surface.

Thanks for the suggestion. Unfortunately, it didn't work. Maybe I didn't rap it hard enough? Not sure exactly where to hit it, but I rapped on the AC side of the inverter.

BB.

Re: SW5548 not recognizing ?good? power on AC1

Rolling would be taking the L1 (black) wire and replacing it with the L2 (red) wire... But because your inverters are "stacked", you cannot do just change one inverter from L1 to L2--That would mess up the 240 VAC output connection.

You could try moving the L1 from inverter A to inverter B, and L2 from Inverter B to Inverter A, and see if the problem follows the wiring or stays with the inverter.

However--Working with AC power and stacked inverters is not a trivial job... If you are not sure what you are doing, don't do it. You could get a very expensive mistake (and I do not know enough about your products to tell you more than swapping the AC input power between inverters and see what happens--Any other wiring changes would not be for me to suggest from behind a keyboard).

-Bill

Blackcherry04

Re: SW5548 not recognizing ?good? power on AC1

ronrosenfeld wrote: »

Thanks for the suggestion. Unfortunately, it didn't work. Maybe I didn't rap it hard enough? Not sure exactly where to hit it, but I rapped on the AC side of the inverter.

It's kind of hard to do if it's mounted. You should be able to hear the relay if it makes the transfer. When I do them , I just give them a thump on the corner edge.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

Blackcherry04 wrote: »

It's kind of hard to do if it's mounted. You should be able to hear the relay if it makes the transfer. When I do them , I just give them a thump on the corner edge.

They are wall mounted. But wait -- when you wrote

like it may be qualifying the power

, when, exactly, is that happening? Is that when the green light goes solid for a few seconds? If that is the case, I need to wait for that to recur -- with this SW revision, once it drops off, there's a five minute delay before it tries again.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

BB. wrote: »

Rolling would be taking the L1 (black) wire and replacing it with the L2 (red) wire... But because your inverters are "stacked", you cannot do just change one inverter from L1 to L2--That would mess up the 240 VAC output connection.

That's not something I would do

You could try moving the L1 from inverter A to inverter B, and L2 from Inverter B to Inverter A, and see if the problem follows the wiring or stays with the inverter. However--Working with AC power and stacked inverters is not a trivial job... If you are not sure what you are doing, don't do it.

That's something we might try tomorrow. My electrician will be here tomorrow to do some other work for me, and we'll get into more troubleshooting if preliminary testing doesn't come up with any clues. I don't feel comfortable myself moving the wires around, but he's quite competent. I do like blackcherry04's sticking relay theory, though.

Blackcherry04

Re: SW5548 not recognizing ?good? power on AC1

ronrosenfeld wrote: »

They are wall mounted. But wait -- when you wrote , when, exactly, is that happening? Is that when the green light goes solid for a few seconds? If that is the case, I need to wait for that to recur -- with this SW revision, once it drops off, there's a five minute delay before it tries again.

When it first sees " Power " it qualifies it, then it will sync to it, then it makes the transfer. The 5 minute delay is a UL1741 requirement to grid tie. So, it could be doing everything right, but it's not accepting the power from the grid and once it faults , then the 5 minute delay kicks in.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

When it first sees " Power " it qualifies it, then it will sync to it, then it makes the transfer

I waited until the light went solid. I did not hear any relay click. I pounded with my mallet, but no joy, and after maybe 10-20 seconds, the green light started flashing again. Does this indicate a relay problem? Or a power problem? Or a need for continued troubleshooting?

Shawn

Re: SW5548 not recognizing ?good? power on AC1

I guess I should have suggested that you verify the proper phase wiring before suggesting that you change it. I assume that with stacked inverters that both inverters need to have the same phase on L1 and the other phase on L2. That could be the reason that one of the inverters is unable to bring the grid power online.
-Shawn

Blackcherry04

Re: SW5548 not recognizing ?good? power on AC1

ronrosenfeld wrote: »

I waited until the light went solid. I did not hear any relay click. I pounded with my mallet, but no joy, and after maybe 10-20 seconds, the green light started flashing again. Does this indicate a relay problem? Or a power problem? Or a need for continued troubleshooting?

It could be, the contact points have a life and get pitted, not to say thats it. Generally since everything else seems to be working, it would be a good chance it is. Maybe boB will come along and put his $.02, he certainly has the knowledge, I am just guessing.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

It could be, the contact points have a life and get pitted, not to say thats it. Generally since everything else seems to be working, it would be a good chance it is. Maybe boB will come along and put his $.02, he certainly has the knowledge, I am just guessing.

This morning we swapped the hot lines between the two inverters, and the problem stayed with the inverter.

Saw something curious. Monitoring AC1 (Grid) voltage on the inverters, with both inverters in FLT mode, using the built-in voltmeter in the inverters, the voltage would smoothly go from about 66-123 and then back down again. On both inverters! And they didn't sync. Measuring the voltage with a DVM (Fluke), at the connection to the inverters, the voltage was stable at 120. After watching this for a bit, I put the inverters into SELL mode, and the voltmeter reading eventually settled down, and the #2 inverter saw "good grid power" and started to sell back. The #1 inverter never saw good grid power, with its built-in meter showing a steady voltage at 119.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

My present thinking is that the AC board is bad. I suppose it could be the relay on the board, or the board itself. If that relay is only used for AC1, then it's never been activated before a few days ago, so perhaps it, or its associated circuitry, has always been bad.

The reasoning: It is my understanding that the steady green light (which appears from time-to-time but only stays on for a few seconds), means the Control Board has determined that the power on AC1 is good. It then sends a signal to the AC Board. Since the relay on AC board never switches, either that signal is not getting there, or there's a problem with the relay.

Unless someone comes up with a better idea, the plan is to swap the board from an old, but never-used SW5548 that I have sitting on the shelf.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

Blackcherry04 wrote: »

It could be, the contact points have a life and get pitted, not to say thats it. Generally since everything else seems to be working, it would be a good chance it is. Maybe boB will come along and put his $.02, he certainly has the knowledge, I am just guessing.

Another thought -- Is that relay only used for switching to AC1? If so, it's never been used so pitting would be unlikely. Sure wish I had a schematic of the unit. My thinking now is bad AC board.

ronrosenfeld

Re: SW5548 not recognizing ?good? power on AC1

ronrosenfeld wrote: »

My present thinking is that the AC board is bad. I suppose it could be the relay on the board, or the board itself. If that relay is only used for AC1, then it's never been activated before a few days ago, so perhaps it, or its associated circuitry, has always been bad.

The reasoning: It is my understanding that the steady green light (which appears from time-to-time but only stays on for a few seconds), means the Control Board has determined that the power on AC1 is good. It then sends a signal to the AC Board. Since the relay on AC board never switches, either that signal is not getting there, or there's a problem with the relay.

Unless someone comes up with a better idea, the plan is to swap the board from an old, but never-used SW5548 that I have sitting on the shelf.

Well, the problem has been solved, thanks to the input of Barry at ePowerNow, a repair service and former Trace warranty outfit. I had phoned wanting to have my AC board repaired, once I had swapped it. However, he spent quite a bit of time on the phone with me talking me through some further troubleshooting, all of which continued to point to the AC1 relay. However, he then pointed me to a small jumper that is next to the AC1 red LED relay light near the AC1 In connection. Removing that jumper disables the relay for Utility testing. The original installation had resulted in those wires getting mashed. Once we pulled out the AC leads, it was apparent that the jumper was missing. I replaced the jumper and everything is working normally. Both inverters sync to the AC Line, and yesterday (it was pretty windy), I was seeing 4-5kW going back to the power company.

A surprisingly simple fix, but without someone to point me in that direction, I would have replaced the whole board instead.