Best PV panel cables and connectors to use for replacing existing?

QrizB

ThisIsWeird said:

Isc - it's ok to short the panel? No resistance load? I guess only for long enough to take a reading?

A solar panel isn't a battery, shorting it won't do any harm and Isc is only slightly higher than the usual operating current. You'll see this when you find the spec of your panels!

ThisIsWeird

Brilliant! Thanks.

Isc is 5.62A for these Suntech Pluto 200W jobbies. Voc 45.7V

I presume anywhere near these sorts of Isc figures will require full sunshine? Sparks ahoy :-)

QrizB

ThisIsWeird said:

Brilliant! Thanks.

Isc is 5.62A for these Suntech Pluto 200W jobbies. Voc 45.7V

I presume anywhere near these sorts of Isc figures will require full sunshine? Sparks ahoy :-)

Yes, Isc is quoted in full sun (1000W/m2) and generally scales with insolation - so you'll see 60% of that in 60% of the illumination.

If you have a clamp-type DC ammeter (something like this) you can cover the panel to suppress the output, short-circuit the connectors then remove the cover and put the clamp around the cable. This way you avoid any sparks

(Note that not all clamp-type ammeters will measure DC current. Check before use.)

ThisIsWeird

No, only a 'probe' type. I'll cover the panel first, tho', connect it and expose myself. The reflection from my white bod will provide the amps.

Thanks again.

ThisIsWeird

Job completed on one array.

All panels provided 40+V, and over 5A in sunshine, and ~3+A in late afternoon.

All panel tails shortened and MC4s fitted. Cables from end panels to inverter checked and have no scuffs or damaged, so kept.

Riso = 0Meg.

What the hell? 

Have to say, I was more than half expecting this, as to have had both arrays showing 0Meg Risos at the same time seemed too unlikely to have been normal leakage faults.

But two inverters - one a pro refurb unit - showing the same?! What the hell!

Heedtheadvice

I meant to post earlier so hope not too late!

If you have a report of low insulation resistance that will normally be from one or both (+ or - ) dc  circuit connections back to 'earth'. Those quite high normal insulation readings (20MOhms) are a combination of all the Insulution resistances in  parallel throughout the circuit.

When you have low readings that can be from one fault or several faults together (parallel insulation resistances again) anywhere in the circuit. So it can be part of the inverter, all the connecting cables and their connectors, isolators, within panel connecting boxes or the innards of the panels.

Best way to locate is  by using  an insulation tester on the whole circuits and then logically all the component parts as above. These testers usually measure in Mohms ranges (a zero can be quite high resistance - not absolute zero but several kOhms -That could be the same for the inverter reported insulation value). The testers must use a high voltage consistent with the circuit operating  characteristics so will be several hundred volts dc for typical solar strings. What can be a high Insulution value at low voltage can break down at high voltage.

Otherwise without a tester it will be trial and error.

You would be advised to check the panels too as well as cables/connectors and the inverter! Any or all can be the cause.

If the value reported varies it is more likely (but not  necessarily) that it is an exterior component. Water ingress, even a slight amount, that gives a path to earth can produce insulation lows in the low Mohms...or less if it is contaminated with 'nasties'

Being anywhere in the circuit causing the problem was the reason I asked why you assumed cables needed changed. (You could be right!)

If you do not have  an insulation tester can you use your inverter (?) progressively starting with no cables/panels connected adding up to a full system. A spare cable to link the circuit may be useful

ThisIsWeird

Thanks Heed.

I don't have access to an insulation tester, alas, nor the knowledge of how to use it on things like this. 

What I have tried are things like fully disconnecting the earth from the inverter and panels, but no difference. I even isolated the whole garage supply (where the panels are) and ran a power cable from the house to the inverter - nada.

I have this recollection of when I replaced my own 'repaired' inverter - which showed the 0.00M Riso - with the pro refurbished one. The new one worked when first connected, and a few minutes later when I was tightening up the wall mounting bolts, I'm sure I heard a small 'pffft', and it just went dead. At some point - I think when I tried rebooting it - it would sometimes trip the RCD.

It's been warm (bludy hot) and sunny, everything outside is bone dry and damned hot, the panels all tested for good open V and current generated, the connectors all replaced and fully smeared in sili grease to lube the O rings, the cables all visually checked - nary a scuff - and the Riso reading of 0.00 is just nuts. AND it's the same on both arrays. I just cannot see any way that the panels or cabling is leaking. The actual PV panels should have zero contact with their ali frames, and - even if they did - surely disconnecting the earth cable to the mounting rails would prevent the inverter from detecting such leakage? 

My gut overwhelmingly tells me summat has popped inside the inverter - a capacitor most likely - and that's doing all the literal leaking! Just bludy incredible if it happened to both Auroras at virtually the same time. But then, they do develop the EO31 faults with astonishing predictability.

I wonder if the manufacturer will answer a request for info - ie which failed internal components could produce this 0.00 result?

Failing that, my plan it to buy a new dual-inverter, such as the Solis - they are good, aren't they? Possibly a Growatt? Or even the absurdly ill-reputed Sunny Roo just as a test - why are they so bad?!

Heedtheadvice

Your evidence does point towards a problem on the vicinity of the inverter(s).

A few thoughts...

Does the inverter do a RIso test even with the panel input disconnected? If so that could be conclusive proof of it being a local issue. Perhaps with the DC isolator switch off? 

Next with the DC cables totally disconnected from the inverter. That would show if there was a problem with the inverter itself. That could be low insulation or a fault in the circuitry. It could also be that the RIso measuring circuit is faulty and giving incorrect readings.

I would suggest that diagnosing those last two causes are beyond the average DIYer and you would probably need an engineer for that or as you wrote the manufacturer. Some do give excellent support even out of warranty but you might be best if it is an aged inverter of going for a new one!

Panels working with voltage and current output is really a red herring!

RCD tripping though is not and if that happens with the DC disconnected it is almost certainly an inverter issue!!

ThisIsWeird

Heedtheadvice said:

Your evidence does point towards a problem on the vicinity of the inverter(s).

A few thoughts...

Does the inverter do a RIso test even with the panel input disconnected? If so that could be conclusive proof of it being a local issue. Perhaps with the DC isolator switch off? 

Next with the DC cables totally disconnected from the inverter. That would show if there was a problem with the inverter itself. That could be low insulation or a fault in the circuitry. It could also be that the RIso measuring circuit is faulty and giving incorrect readings.

I would suggest that diagnosing those last two causes are beyond the average DIYer and you would probably need an engineer for that or as you wrote the manufacturer. Some do give excellent support even out of warranty but you might be best if it is an aged inverter of going for a new one!

Panels working with voltage and current output is really a red herring!

RCD tripping though is not and if that happens with the DC disconnected it is almost certainly an inverter issue!!

Thanks, HtA.

It's an Aurora Power One, and it only fires up when it receives a DC voltage from the panel, so I cannot get it to do a Riso test without the panels being switched on in daylight.

Yes, diagnosing internal inverter issues would be beyond me, tho' I might open my own self-'repaired' one again and do a visual check for a swollen capacitor or similar.

I'm going to buy a new inverter, and if that works I can return the pro-refurb'd one as it's still within their 3 month warranty period, and I did actually report the issue to them within a couple of day of purchase in any case.

The refurb'd one definitely did go 'pfft' when it tripped the RCD. Anyhoo, hopefully I'll know soon enough :-)

Cheers.

ThisIsWeird

Well...

Unbelievably, two inverters - my own, on which I'd just carried out an E031 repair - and a professionally refurbished one (for the same issue) simultaneously developed an additional fault - the zero Riso.

My own inverter, after replacing the relays, gave the Riso: 0.00M ohm warning on both arrays when I refitted it, and I obviously blamed my own kludsiness during the repair for having caused an additional fault.

The pro repaired one worked for around 3 minutes, and then returned the exact same error. This blew my mind.

Removed half the PV panels, and replaced the original connectors with MC4s - quite a task, even tho' I could work off a raised bank - but there was clearly nothing amiss with the panels or the wiring - all looked good, and gave solid outputs. No difference - zero ohm Riso

NOS - a 2016 model, but never installed - arrived today, and it works perfectly - 20M Riso, and not a single mA of leakage, so my panels are fine as I'd hoped.

What are the bleedin' chances?! Two inverters having developed the same - and seemingly unusual - fault?

They don't call it electrickery for nothing.