0Volts coming from panels

SilverB

I have had my offgrid system running my house for about 3 years now without an issue. I have 9 Eclipsall 300W panels wired in 3 sets of 3panels wired in series. The other day my breaker for my Charge controller (Flex 60) accidentally got shut off, when it was switched back on it blew immediately. I waited till night to switch it back on again, the next day 0V incoming into charge controller, I pulled the fuses (15A/600V/fast acting midget type) and replaced them, still 0V. I tested the voltage of the panels at the combiner box and it still showed 0V, if I remove the fuses and check the voltage at the panels I am getting the correct Voc of 44.8 +-.2.

It seems like as soon as the charge controller is hooked up the voltage drops to 0? I also have a midnight solar lightning arrester on the combiner box, it is no longer lit up? Any tips on what I should check next? fried charge controller? Can a lightning arrester absorb all power flowing through a system if it is blown?

The rest of the system is comprised of 16 Rolls S550 (wired in two banks of 8 running 48V) with a magnum energy 240Vinverter and mini panel. Any help would be appreciated as I am currently relying on my generator for 100% of my power.

BB.

Does not make a lot of sense... Solar panels "go bad" from several different reasons. The closest failure I can see would be if, somehow, the battery was connected "backwards" to the panels (the charge controller was bypassed, or possibly, connected backwards to the battery bank itself). The high current from the battery bank would burnout traces/cells in the panels and blow the circuit breaker.

Otherwise, you have to disconnect each panel and measure Voc (voltage open circuit) and Isc (current short circuit) and see what you find.

Possible other issues are failed crimps or connectors between panels.

-Bill

SilverB

Nothing was changed on the system from the day it was installed 3 years ago, all polarities are correct as they have been. The charge controller has a breaker between it and the battery bank, I have fuses between the panels and the charge controller as well as lightning arresters.  The panels were in full noon day sun when the charge controller breaker was switched back on (and immediately tripped), is it possible the panels exceeded the CC's rating and fried something internally on the charge controller?

BB.

What is the array size/configuration (i.e., 3 x 30 Vmp 300 Watt panels in series * 3 parallel strings)? What is your battery bank voltage/AH rating/battery type?

In general, the maximum breaker for your Outback Flex 60 (?) would be 60 amps * 1.25 NEC derating for breakers+wiring = 75 amps and round up to 80 amps rated breaker/wiring.

MPPT charge controller generally will not output more current than they are rated for (i.e., 60 amp controller will safely/normal operation limit output current a maximum of 60 amps). In general, a breaker will require many minutes to hours if you put 60-80 amps through a 60 amp breaker.

To trip immediately, I would expect to see 2-3x or more rated current through the breaker. The only way that could happen on a typical solar charging system would be if the charge controller output (diodes, FETs, Transistors, filter capacitors) failed shorted. Can certainly happen--And if it did, the controller needs repairs or replacement. Such an internal short may also bring your array voltage to near zero volts. You can use a DC Current Clamp DMM to see if there is current flowing through your array (a good thing if feeding a shorted controller--At least your panels are still, most likely, OK).

http://www.sears.com/craftsman-digital-clamp-on-ammeter/p-03482369000P ($60 or so, good enough for our needs)

I do not know enough details about your controller to make any detailed guesses about the possible/probably failures--You need to call your installer or Outback support for more help.

If you want to check your array--The "best way" would be to connect the array around the charge controller--Directly to the circuit breaker. Use a DC Current Clamp meter to monitor the array current (should be within 80-100% of rated Imp-array when directly feeding battery bank around noon). If not, you can use the DC current Clamp meter to see which strings are supplying power and which are not.

Connecting your array directly to the battery bank for a short time (through a switch like a Circuit Breaker) should not hurt anything. Do not leave connected or you can (will) over charge your battery bank.

Other warnings--Do not connect array "backwards" to battery bank (will destroy your panels). Do not unplug the solar panel connectors while current is flowing--You probably will damage the small connectors/pins.

Your array voltage is probably around 24 to near 100 VDC--As always, be careful. Those higher voltages can easily hurt or kill you.

-Bill

Photowhit

SilverB said:

I tested the voltage of the panels at the combiner box and it still showed 0V, if I remove the fuses and check the voltage at the panels I am getting the correct Voc of 44.8 +-.2.

Okay, this line doesn't make any sense to me. If you remove the fuses and check at the combiner box what do you get? Do all the strings show voltage?

Did you check at the charge controller with another meter? If it doesn't show voltage, I would 'walk' the voltage back to the charge controller. Check across (each side) each breaker and fuse. You'll find something dead along the way.

SilverB

Yes 3x300w panels in series x 3 parallel strings, 2banks of 8 S550 rolls surrette giving 48V @ 856AH (20hr rating). Ill also add that when this happened it was around -30*C, and in full sun, so the panels were probably making well over their rated output, because of the cold.
I dont have a clamp type DMM, but I may invest in one if stuff like this happens again.

When the panels were charging normally I would see input voltages on my CC between 110v-138v (depending on sun and temperature)

Photowhit

It might have been that the VOC was too high when you flipped the breaker back on...

If you have voltage at the panels, you should see voltage at the combiner box, with out the fuse in it, just with a hand held meter. Lightning arrestor would be on the other side, if you have one at the combiner box.

SilverB

Yes if I take the fuses out of the combiner box, I get 100% rated reading for Voc, 44.8 (per panel) As soon as charge controller is connected voltage drops to 0

SilverB

Photowhit said:

SilverB said:

I tested the voltage of the panels at the combiner box and it still showed 0V, if I remove the fuses and check the voltage at the panels I am getting the correct Voc of 44.8 +-.2.

Okay, this line doesn't make any sense to me. If you remove the fuses and check at the combiner box what do you get? Do all the strings show voltage?

Did you check at the charge controller with another meter? If it doesn't show voltage, I would 'walk' the voltage back to the charge controller. Check across (each side) each breaker and fuse. You'll find something dead along the way.

Yes I checked at the charge controller at the PV input (shows 0v), When I check voltages at combiner box with fuses out, yes I get around 134v showing from all 3 strings. Other than the fuses the only thing between the panels and the CC, is the wire itself (4AWG run over about 50ft)

Estragon

A couple of thoughts...
- Have you tried reading voltage at the pv-in terminals on the charge controller with the charge controller connected, but off (i.e. breaker to battery off)?

- Have you looked at the controller logs/stats?  Anything odd there (thinking max lifetime Voc in particular).  Also check settings (in case they've reset to defaults etc.)?

I'm assuming when you say the breaker to the controller was accidentally turned off, you mean the pv breaker, not the battery breaker?  If it was the battery breaker, you might try opening both breakers, then closing the battery breaker first, wait for the controller to fully boot, then close the pv breaker.

SilverB

I havent checked out the log/stats I will try that tonight, the breaker that was accidentally turned off was the breaker between the CC and the battery bank. There is no breaker between the panels and controller, just fast acting fuses. If I put the fuses in and turn on the battery breaker, the CC comes on reading 0V, if I turn the battery breaker off the CC shuts down.

dennis461

SilverB said:

..... tested the voltage of the panels at the combiner box and it still showed 0V, if I remove the fuses and check the voltage at the panels I am getting the correct Voc of 44.8 +-.2......

You have a short circuit on the downstream side of those fuses , you did not state where the fuses are nor a combiner box model(unless I missed it.)

I will assume the fuses are in the combiner box
so  the short circuit is in the combiner box or cable running to charge controller-inverter.

You could disconnect everything during darkness to be safe.
Try running one string only to inverter, bypassing the combiner and existing cable 

Alternately,  lift the panel leads from charge controller AT THE CONTROLLER measure panel voltage at the leads, if it is GOOD, then combiner and cabling is OK.

You may have a shorted out charge controller

Estragon

SilverB said:

I havent checked out the log/stats I will try that tonight, the breaker that was accidentally turned off was the breaker between the CC and the battery bank. There is no breaker between the panels and controller, just fast acting fuses. If I put the fuses in and turn on the battery breaker, the CC comes on reading 0V, if I turn the battery breaker off the CC shuts down.

You should power the controller up with batteries (only) first and let the controller boot up fully, then add the pv power.  Breakers are handier for this, but fuses get the job done too. 

I'm wondering if by dropping the stable DC from the batteries, the controller has gone into an unwanted state (like a factory reset sort of thing).  Rebooting the controller with only battery power only may restore to proper settings (but check all to be sure before powering with PV).

It could be a wiring thing, but the timing just seems pretty co-incidental to me.

SilverB

The cables are 4AWG TEK (underground rated) buried 3 ft down in solid clay, inside conduit for the last 10ft at each end of the run, I really doubt there is a short there, considering the size of cable should handle way more V/A than the panels could ever create.  The combiner is a Midnight solar box, couldnt tell you the model (not sure why that would matter its an empty metal box with fuse holders in it) But the fuses are 600V fast acting fuses rated at 5amp (made by littlefuse)

The combiner box is at the panels, again the location shouldnt really matter, its going to do the same job whether its 5 ft from the panels or 50ft from the panels.

dennis461

SilverB said:

The cables are 4AWG TEK (underground rated) buried 3 ft down in solid clay, inside conduit for the last 10ft at each end of the run, I really doubt there is a short there, considering the size of cable should handle way more V/A than the panels could ever create.  The combiner is a Midnight solar box, couldnt tell you the model (not sure why that would matter its an empty metal box with fuse holders in it) But the fuses are 600V fast acting fuses rated at 5amp (made by littlefuse)

The combiner box is at the panels, again the location shouldnt really matter, its going to do the same job whether its 5 ft from the panels or 50ft from the panels.

Trouble shooting should never contain a statement like..."... I really doubt there is a short there,..."
This kind of talk has got me into hot water many times:-(.

I suspect the worst happened, the battery breaker should never be closed in last on the el-cheapo Charge Controllers (CC).
I know you have a Flex 60 which is not el-cheapo, and the Flex 60 manual does not warn about this, but, it is no coincidence that your trouble started after the battery breaker was closed in to the CC.

So, If it was my system, I would hang a voltmeter at the combiner box and open the circuit at the CC.
Do you have a DC disconnect between combiner box and CC?

If you are lucky ($$), you have a shorted cable, unlucky means the CC died.

SilverB

I dont have a disconnect, just the fuses in the combiner. 

And I know I shouldnt assume things, but how would a WAY overrated TEK cable in frozen ground (not under a driveway) get damaged? With how quickly the breaker blew and the fact that .5 second before that everything was working fine, I dont think there would have been enough time to heat up the cable to the point of shorting.

At this point Im hoping its the charge controller, Direct burial TEK cable isnt cheap, especially 4AWG and about 70ft worth with all the bends and such, plus excavation, not to mention the ground is frozen rock solid about 3 feet down, and it doesnt look like its going to be thawing anytime soon. The charge controller I can switch out myself in about 10min.

dennis461

OK, I see your dilemma, cable can fail from electricity or environment.
I've found buried cable failures caused by small knick in insulation during the burying process, with insulation damage, or a sharp stone touching the cable, many years of good service is ended in a split second when the voltage gets through to the ground (dirt).
The systems I dealt with were single conductors with very thick insulation designed for direct burial.

If your cable itself is in FROZEN ground, it was not buried deep enough, and can ultimately crack open.
Hopefully yours is under the frozen ground, and is able to flex due to ground movement (ever so slightly) without insulation cracking.

Frozen ground is only a small challenge, a backhoe cracks open the top much as you would break ice off a lake.
But I digress.

After dark, disconnect the solar input right at the CC, then test voltage at the disconnected cable in the morning.

Does your cable have an aluminum jacket (TECK as opposed to TEK)?

Estragon

SilverB said:

I dont have a disconnect, just the fuses in the combiner. 

And I know I shouldnt assume things, but how would a WAY overrated TEK cable in frozen ground (not under a driveway) get damaged?
<snip>

In frozen ground (especially clay) you get pockets of water which expand and cause heaving.  You would be surprised at the power of such heaving - it can move entire buildings.  If you have rocky clay, even worse.  The rocks turn into shear presses which could easily make short work of even a 4/0 cable.

I'm still thinking controller issue, but I wouldn't rule anything out just yet.

SilverB

Sorry yes TECK, it is aluminum coated, code in Ontario calls for 18" down, I went 3 feet down, which should be below the frost line if we have good amounts of snow on the ground. to be honest Ive never heard of frozen ground affecting cables, almost every buried cable in Ontario is only 24" down at the most and Ive never come across it.

Photowhit

I don't use Outback charge controllers and I suspect things might have gone south there...

I don't know, but suspect you could have lost your settings and might need to simply reset them, I think they are designed to be remotely setup with a mate, so either the disconnect or the inrush might have messed things up.

I will suggest you ask at the Outback forums, and when I went to find a link, I found you've already gone there.

I think you are in good hands!

For others...

http://www.outbackpower.com/forum/

Might go back and hook things up as if commissioning a new system.

http://www.outbackpower.com/downloads/documents/Make_the_Power/flexmax_6080/owner_manual.pdf

Estragon

In my part of Ontario, frost can get to 6-7' down. I doubt this is the problem though, just not ruling it out.

mcgivor

It would seem if everything was working fine up to the point the breaker was turned off, then on, that the cable  would unlikely be the problem. What I would do is lift the solar input to the CC, after it is powered down and verify there is power, thereby eliminating it as a possible cause, such an internal short. Also check the arrestor or temporarily remove it for testing  purposes. Probably a good idea to partially  cover the array with blankets to reduce the chance of a mishap if working in full sun.

My thinking is being that it was cold and sunny at the time, the sudden collapse or reintroduction of the, load may have caused damage to the CC as the VOC would probably have been pretty high, or hopefully it's just the arrestor. 

Vic

 Hi SilverB,

At first blush,   had assumed that probably the String Voc into the FM60 had probably been exceeded,   as your configuration is strings of three 72 cell PVs in series.   At 25 C,  the STC Voc is 46.1 V per PV.  If the PVs were at --30.00 C this would yield around 147 V.   This IS perilously close to the absolute Maximum Voc rating of the FM CCs.

Am with mcgvor,   that the CC could well be toast ...

In your climate,  personally,   would never,   ever,  run a system this close to the absolute maximum,  Warranty Voiding input V limit of 150.00 volts.

Normally,  for these MPPT CCs,  if the input PV terminal show 0.00 volts with the battery breaker closed and the input open,   the FETs,  and perhaps other electronics connected to the PV input are toast.   If good PVs,  in good sun,   are connected to the input of this CC,  and the input terminals still read 0.0 volts,  then it would seem that the CC is really toast.

There should be a Log entry for the highest recorded Voc for this CC.   You might want to look at that,  to see what this value is.

Just a drive-by.   Good Luck,   Vic

mike95490

fuses (15A/600V/fast acting midget type)

Do they make midget 600VDC fuses ?   AC I can understand, but 600VDC is usually a longer fuse with sand to quench the arc.
If it came in a midnight combiner box, I guess it's legit. but that worries me.

Photowhit

mike95490 said:

fuses (15A/600V/fast acting midget type)

Do they make midget 600VDC fuses ?   AC I can understand, but 600VDC is usually a longer fuse with sand to quench the arc.
If it came in a midnight combiner box, I guess it's legit. but that worries me.

I use "KLK" midget fuses, yes they are rated at 500 volts DC.

KLKD are rated at 600 VDC and SPF are rated at 1000 volts DC.

http://www.littelfuse.com/products/fuses/industrial-power-fuses/midget-10x38mm-fuses.aspx

SilverB

Yes they are definitely 600VDC/VAC, fuses usually dont care if the power is DC or AC, they are specific to PV systems, they are non indicating ceramic fuses (KLKD15 if you want to check them out), tested the fuses and they are definitely good. This weekend while Im home during the day I will unhook the PV input wiring and test it to see if the PV power is getting to the CC. If it isnt then I will start disconnecting my panels to test individually. I checked the log data in the CC and it shows maximum voltage of 146.8V, but it shows that on the day this happened, so it may have even been higher, but that is where the CC stopped logging. If it does turn out to be the CC, I will probably switch to a Magnum PT-100 which will handle up to 240V input and will work along with everything else in my system.

BB.

Fuses, breakers, switches, etc. really do care about AC vs DC power. DC voltages/currents are much more difficult to "break"--DC tends to sustain arcs much more than AC.

https://www.youtube.com/watch?v=Zez2r1RPpWY

-Bill

dennis461

Getting a bit off track, but since none of his fuses cleared, I am re-thinking short circuit.
My math says  one string should supply around 8.3 ADC with 15A fuses..

Perhaps there is a high resistance connection somewhere, which allows a voltmeter to show 120+ VDC, which then drops to zero when trying to pass current.

Vic

dennis461 said:

Getting a bit off track, but since none of his fuses cleared, I am re-thinking short circuit.
My math says  one string should supply around 8.3 ADC with 15A fuses..

Perhaps there is a high resistance connection somewhere, which allows a voltmeter to show 120+ VDC, which then drops to zero when trying to pass current.

Perhaps,   but the FM CC injects a small amount of current into the PV in circuit,  connected to the PV input terminal.

If there was a high-resistance connection to the PVs,  this current would seemingly cause some positive voltage to show as Vin.

At night,  with the PV input breaker/fuse closed/intact this voltage should be present when the CC is not damaged.   With  only a few PV strings,  this sleeping/resting voltage reading will usually be between 1/2 to 2/3 of the battery voltage at that particular time.  With many PVs/strings of PVs,  this voltage will usually be reduced,  by PV leakage current.

Shorting the PV in terminal to ground,  with the battery connected to the output of this (and many other) MPPT CCs will usually damage the CC  --  blown FETs.   With FETs blown,   it is my understanding that the Vin terminal and Vin voltage will usually show 0.0 volts,  and be a fairly low impedance.

However,   have never (yet) blown up a CC,   so this is not from direct observation.

YMMV,   FWIW,   and so on,   Vic

SilverB

I forgot to mention, yesterday when I got home (5:30pm still plenty of sunlight)  I went to pull all of my fuses in the combiner box, when opening the last fuse holder my finger slipped off, opening the fuse holder only part way and it was shooting a good size arc from the fuse holder to the top of the fuse (once I got the holder open fully the arc immediately stopped) This makes me think the panels are still making plenty of Voltage & current, otherwise they wouldnt have been able to shoot an arc of that size.

Photowhit

SilverB said:

I forgot to mention, yesterday when I got home (5:30pm still plenty of sunlight)  I went to pull all of my fuses in the combiner box, when opening the last fuse holder my finger slipped off, opening the fuse holder only part way and it was shooting a good size arc from the fuse holder to the top of the fuse (once I got the holder open fully the arc immediately stopped) This makes me think the panels are still making plenty of Voltage & current, otherwise they wouldnt have been able to shoot an arc of that size.

Also indicates that the current has some place to go, a short circuit or a charge controller that is allowing current to pass.