Fried conext sw4024

jtdiesel65
jtdiesel65 Solar Expert Posts: 242 ✭✭✭
Was testing the charging on two parallel sw4024. I noticed the master was charging, but the slave was not. The slave appeared to be inverting and providing a couple amps of power while the master was in absorb. The battery was full, so if I could understand there is no reason for the slave to charge along with the master. Though, I don't see in the documentation that this is how it works.

Anyway, the second I switched the generator start to manualoff, the master made a awful noise and vibrated. And then did it again but worse when the generator actually shutoff. Then it threw an over current fault. There was really no reason for this as there would have only been a light on. The slave didn't have any problem. So I hit all the breakers off. Then powered back up. Both inverters came online and the master was inverting. So then, I reconfigured and fired each up individually as a stand alone unit. I went thru generator manualon, charge for a bit, and then manualoff on what was the slave (now configured as a master) with no problem. On the original master, I went thru manualon, charged for a bit, but then during generator shutdown it behaved the same way except with smoke. And now it's dead. No inverter hum and with a constant output voltage under fault.

All I can think is it's a bad inverter. Anybody see this happen before?

Also does anybody know in a parallel configuration if the AC/gen input neg bus needs to isolated from the output neg bus? The sw4024 manual has a page that says it does, but the wiring diagram shows both are on the same bus. So the manual is telling me two things. My old sw4024 used the same bus.
«1

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Fried conext sw4024

    Ouch... Best wishes to getting this fixed quick.

    In general, the warning is to have a single point ground for DC to Ground Bus and AC Neutral to Ground bus (usually the same ground bus) of TSW inverters.

    You don't want, for example, a heavy DC current to find its way back in the (typically smaller gauge wiring) in the AC side of the system.

    There have been a few reports where the AC Neutral to Earth Bond was made in the genset, and "circulating" AC and/or DC currents got back to the genset and blow the AC Voltage Regulator (Icarus/Tony?).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    Thanks. We'll see how it goes.


    Maybe I don't quite follow you or I didn't explain it well.

    This is the negative not the ground bus. You have separate negative buses for DC and AC sides. On the AC side, the manual states that it wants to have separate negative buses: one for the gen input and one for AC output (output from the inverter) (page 1-34 from http://lib.store.yahoo.net/lib/wind-sun/Conext-SW-120-Install-Guide.pdf ) So that would be a total of three separate negative buses: one DC and two AC. However, the manual also has a wiring diagram (page 3-8 ) that only shows the DC bus and a common AC bus which is common to both the AC input and AC outputs.

    So the manual says two different things. And to add to that confusion, the AC connect box for the SW systems only comes with one negative bus and the breaker kit for stacked SW's doesn't provide an additional bus.

    I'm not sure why the AC input and AC output neutrals need to be separate.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    Thinking about this even more, it doesn't make sense to have separate buses for AC in and AC out. The Schneider Electric Conext SW AC Circuit Breaker Panel and the AC Breaker Kit for Stacked Conext SW have the breakers for a bypass where you can run the house directly on the generator bypassing the inverter. There is no breaker for neutral so you wouldn't be able to have bypass capability unless both negatives were on the same bus.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Fried conext sw4024

    I am a bit confused--Negative bus or floating vs grounded neutral bus?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    Isolated neutral not the ground. Here's the page from the manual:
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    and here's the wiring diagram that shows they use the same bus:
    Attachment not found.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Fried conext sw4024

    I cannot tell from the second image (forum probably dropped the resolution too far--fuzzy).

    I am guessing that they are intending for the Neutral to be bonded to earth in one location. In off grid inverters, many will bond inside the inverter. With two or more inverters, they don't want you bonding it two inverters (ground loop posibility). I think they just want the neutral bonding done outside the inverter (the inverter is limited to 60 amps AC. If you bond internally, it would be too much current).

    Another question is the AC Transfer switch. A good number of inverters will transfer Hot and Neutrals. Some will just transfer Hots. If you transfer Hot+Neutral, then you can have two different neutral bondings. One remote (such as the AC bonding at the RV park AC panel and the second/off grid at the RV Trailer bonding).

    But if you have AC Generator Support (where larger gensets are factory default with Neutral bonding inside the genset) and generator support (where the AC Inverter adds additional Battery energy to AC Generator output for heavy loads)--That adds another place to look for neutral bonds and ground loops.

    Yep, I am confused...

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: Fried conext sw4024

    They may be isolated. However, when you connect them together, they're no longer isolated.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    The second image is page 3-8 from http://lib.store.yahoo.net/lib/wind-sun/Conext-SW-120-Install-Guide.pdf

    I'm confused too. I'm not sure which way they want it. Also not sure why they want different buses only with stacked inverters and not also single stand alone inverters. I do know that the AC breaker panel and AC stacking kit do not support the multiple buses as outlined on page 1-34.

    What is also strange is the ac switchgear install manual doesn't provide for a neutral line from the inverter to the switchgear. It has a neutral line on the input side.
  • Blackcherry04
    Blackcherry04 Solar Expert Posts: 2,490 ✭✭✭
    Re: Fried conext sw4024

    From what I can remember your old Trace sw4024 did have a straight through neutrals and Xantrex has always had Isolated, in the units I have installed. Without a separate neutral you got a fault from the sub panel.
    .
    .
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024
    From what I can remember your old Trace sw4024 did have a straight through neutrals and Xantrex has always had Isolated, in the units I have installed. Without a separate neutral you got a fault from the sub panel.
    .
    .

    That's the only thing I can think of. But, the manual isn't clear which way it should be connected. It says both ways. It's confusing. The AC distribution center doesn't even show a neutral wire from AC out which implies it's straight thru neutrals (no reason to connect the neutral twice). Seems like if separate neutrals are a problem, then there should have been an issue right from when I started the generator. And when I connected the slave as a master, I should have seen the same problem. But I didn't. The other thing is the AC center has a physical bypass lockout and it doesn't have breakers for neutrals. There is no way the bypass lockout can work if the neutrals are on separate buses.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    Although I don't have an answer regarding neutrals, I called Schneider Electric told them the problem and how it was wired, and they told me they thought the unit was defective and took all the info required by the warranty. I bought it from WindSun and was told I have to work through them to get the replacement.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    Update:

    On my sixth inverter from Schneider. Still have the same problem where when genset power is applied only the master enters charge mode. This only occurs randomly. More frequently, both enter charge mode and everything works fine. When only the master enters charge mode, the scp displays a very large load and when you kill the generator, the master will fry. It's almost like the master is providing the slave with the juice to charge the battery....like in some cyclic pathway perhaps where some set of relays isn't firing correctly. We widened up the in input tolerances and it didn't make a bit of difference. Changed out network cables to new cat6. Tried rerouting the network cables in case there was interference. No luck. I'm really at the point that there is a flaw in the product.

    You can see what I'm dealing with here:

    https://www.youtube.com/watch?v=0aQwUY6K8cw
    https://www.youtube.com/watch?v=Ld_X7vLA62E

    In this state, the only way to prevent the master from frying is to shut the DC off to the unit while shutting off the generator in breakers.

    Any ideas would be most appreciated.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    And here is what happens when I switch the generator off


    http://www.youtube.com/watch?v=8DNZ8aYu9ec
  • halfcrazy
    halfcrazy Solar Expert Posts: 720 ✭✭✭
    Re: Fried conext sw4024

    Really hard to tell but it almost sounds like one relay may be switching before the other? Although it also looks like one is feeding the other in a semi perpetual motion thing as well. Hopefully Schneider will get to the bottom of it. Sounds like it is repeatable so that's a good thing from a troubleshooting stand point.
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Fried conext sw4024
    halfcrazy wrote: »
    Sounds like it is repeatable so that's a good thing from a troubleshooting stand point.

    That's the understatement of the day! He's on his sixth inverter! --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    The thing is, it's intermittent. I also get an f72 AC out association mismatched fault when I start the units, which goes away in a second, and then the inverters invert as normal. I'm not sure if that behavior is normal on startup of master/slave configured inverters. I also had the fault randomly occur once after being installed and inverting for 3 straight days. It's not a good thing since the power goes out for a second when it happens. I've been told this is due to the units either being configed with EURO Freq or that the AC_IN assignment on both units are not the same. However, both settings are what they should be: EURO disabled on both and both with GEN1. So why it happens is a mystery.

    I've also seen the two inverters cycle attempting to engage charge mode. When normal charge mode engages, it sounds to me like there are two relays per inverter that fire. When it cycles, it sounds like only one relay per inverter fires, and then releases, then wait a few seconds, and it tries again. There might be up to 5 attempts and then it will either enage charging on both inverters as normal, or only the master will engage charging. You can't force it to cycle like this. Sometimes, both engage on the first attempt. Sometimes, only the master engages on first attempt. I don't know how this is designed to work, but all I can think is that when only the master engages condition happens, maybe the master on-lines the generator a split second before the slave which causes a change in input voltage or freq which causes the slave to drop the qualification. But that's only a guess. And either way, when the system is in this state, there is some weird load issue as seen in the youtubes and shutting down the generator shouldn't fry the master.

    I've also swapped the wiring on both units and the problem followed the inverter hardware. Continuity tests on the wiring pass. Both inverters work fine as stand alone units, with no faults and no problems engaging charge mode (no cycling).
  • Blackcherry04
    Blackcherry04 Solar Expert Posts: 2,490 ✭✭✭
    Re: Fried conext sw4024

    What you mentioned below sounds familiar. I have a SW 3000 that will do the very same thing. It will take 4-5 + tries to ever stick in the charge mode and will drop the generator or Grid Power. The way I see it, it has something to do with the battery voltage. I cut the charger % back , that never worked. I seems like once the bulk voltage starts to come up, it will finally stick.

    I know, why would someone start talking about some other Inverter, take a look at the pictures of them. It looks like the SW3012, SW2012 SW 4024 and the 3000's, they must share something platform wise besides the case, because of the common price points.
    .
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    I figured out how to reproduce this:

    master: inverter and charger enabled, search disabled
    slave: inverter and charger enabled, search enabled

    Outputs from inverter out breakers need to be connected to their
    respective buses. But there doesn't have to be any loads connected to
    those buses.

    1) shut all breakers off
    2) start generator
    3) turn generator breaker for only the master on
    4) wait for master to enter charge mode
    5) turn generator breaker for slave on. At this point, the slave won't enter charge mode( or at least no light will be displayed). Slave inverter out L1 to L2 will measure ~19v, L1-N and L2-N each 9.5 v. Master inverter out L1-L2=240, L1-N,L2-N=120. Input readings on both are L1-L2=240, L1-N,L2-N=120.
    6) Turn on inverter breakers. At this point, the master will issue a sound as if under significant load and load in the SCP will visually go from 0 to ~2000w. Readings will be typical for 240v.
    7) At this point, shutting off the generator breakers will fry the master. To avoid this, shut off inverter out breakers and then the generator in breaker for the slave followed by the generator breaker for the master.

    What is going on is the slave sits in "beingqualified" and doesn't make it to "acpendinggood". Looking at the meters view, the slave indicates
    that it is bulk mode and the AC looks good. So it's a situation, that the inverter's operational status thinks its in bulk when it's not. If I
    do the opposite and throw the slave's breaker first, it moves into "acpendinggood" and then when I turn on the master's breaker, both move
    into charge mode. So I think there is a product flaw.

    BTW, I'm also being told now that the output neutrals need to be on a separate bus than the input neutrals and that neutral is switched in the inverter. But I'm not convinced that they need to be on a separate bus. And at any rate, the problem still occurs regardless of the neutral bus.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Fried conext sw4024
    jtdiesel65 wrote: »
    BTW, I'm also being told now that the output neutrals need to be on a separate bus than the input neutrals and that neutral is switched in the inverter. But I'm not convinced that they need to be on a separate bus. And at any rate, the problem still occurs regardless of the neutral bus.

    This stuff gets amazingly complex very quickly.

    If the Conext measures current in the neutral legs--Then, yes, you probably need to switch/isolate them.

    Or, there is the whole "where is your Neutral/Ground bond taking place" question... But that should not cause the inverter to behave like you are seeing...???

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • halfcrazy
    halfcrazy Solar Expert Posts: 720 ✭✭✭
    Re: Fried conext sw4024

    Yes I see the issue now. So it is a combination what you are doing and software.

    First thing is both inverters need to be in search or not in search. (The software should force this issue) There is no reason not to have them match here.

    Second is all AC breakers that do the AC out and AC in need to stay on at all times. It sounds like you are not actually getting them on and stacked before you have them do work?

    Also when you start the generator it should be turned on to both at the same time.

    Basically any connecting or disconnecting should be done to both inverters simultaneously. What is happening is one is running backwards as if it was AC coupled to the other one. You have a loop going. I suspect this is because of the sequence of breakers etc. I think the way this should work is the breakers that do the individual SW's need to stay on all the time so both inverters see the same thing. What Schneider should do (In my opinion) is write software that looks for this buy montioring the AC ins and AC outs and if it is told it is a dual system verify that. If they see AC on one but not the other they should shut down.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    I can reproduce it by manually switching breakers and that seems to duplicate whatever scenario is occurring in normal operation. Normal operation is gen breakers on, inverter breakers on, and the autostart turns on and off the generator. In normal operation, the problem is intermittent. So there is some timing type of software flaw happening that the manual application of the breakers duplicates.
  • halfcrazy
    halfcrazy Solar Expert Posts: 720 ✭✭✭
    Re: Fried conext sw4024

    Ok. So it is a software issue they are for some reason forgetting they are paralleled. I know Sandra is spending the week in engineering trying to solve this.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    Who is sandra? Schneider software dev?
  • halfcrazy
    halfcrazy Solar Expert Posts: 720 ✭✭✭
    Re: Fried conext sw4024

    Ok big red flag I just read the manual. Do you have the AC in Neutral and AC out Neutral isolated from each other? It appears they are doing some funky voodoo internal measuring current on the neutrals and the manual points out (And Sandra from Schneider confirmed) that on DUAL ONLY stacks the Neutrals MUST be isolated.

    This is just my opinion but I suspect they are measuring current on the neutrals to balance the inverters and if they are common in and out they can not measure current. This could lead to some funky stuff.

    Ryan
  • halfcrazy
    halfcrazy Solar Expert Posts: 720 ✭✭✭
    Re: Fried conext sw4024
    jtdiesel65 wrote: »
    Who is sandra? Schneider software dev?

    Sandra is one of a couple in Level 3 tech support at Schneider
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    Re: Fried conext sw4024
    halfcrazy wrote: »
    Ok big red flag I just read the manual. Do you have the AC in Neutral and AC out Neutral isolated from each other?

    From post #20 in this thread:
    jtdiesel65 wrote: »
    BTW, I'm also being told now that the output neutrals need to be on a separate bus than the input neutrals and that neutral is switched in the inverter. But I'm not convinced that they need to be on a separate bus. And at any rate, the problem still occurs regardless of the neutral bus.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • halfcrazy
    halfcrazy Solar Expert Posts: 720 ✭✭✭
    Re: Fried conext sw4024

    Ok I missed that in the reading. Sorry about that.

    I am at a loss other than to say it is a software bug. Have to wait and see what Schneider comes up with.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    The manual actually has it both ways. It says isolate the neutrals on 1-34, but the wiring diagram in the appendix page 3-8 has both neutrals going to the same bus. Schneider original told me the neutrals were common and page 1-34 was meant for stacked inverters not using the Schneider AC panel. The thing is that feedback matched what the parts list in the AC breaker panel and the Breaker kit for stacking inverters provides. There is no extra bus provided. There is only one neutral bus in the panel and another one is not provided. Extra breakers for the neutrals are not provided either. And the bypass lockout mechanism is for only two breakers, not three (total of 6). Also the kits don't provide the AC Out neutral wires. Originally I was told that the ac out neutral was redundant because the neutrals are bonded, but that I should install the neutral wire for load. So unfortunately, the AC panel and breaker kit aren't all that usable. It's going to be difficult to shoehorn in another bus and the lockout will never be usable. At any rate, I still see the problem regardless of the neutrals. Of course, now that they are saying the neutrals must be isolated, it's entirely possible that the inverters were somehow damaged because the neutrals were connected. Although they both invert fine, they both invert and charge fine as standalones, and the problem is intermittent.
  • jtdiesel65
    jtdiesel65 Solar Expert Posts: 242 ✭✭✭
    Re: Fried conext sw4024

    update:

    Evidently Schneider found a software timing issue and there is a firmware fix in test. I'm told it will fix the problem. Sandra/Schnieder is working on getting me the proper parts to make the AC panel's bypass functionality work while I'm waiting for the firmware.