PT 100 gremlins

sparkmeredith

Hi there -

I just got off the phone with a Magnum tech and am feeling more confused than before. 

I recently installed two 12V off grid PV systems in northern coastal British Columbia. The systems are identical in every way except the way the panels are mounted. One array is roof mount, the other tree mount.

The systems consist of

12  x Surrette 2V - 1450
Magnum MS2812 on an e-panel
Magnum PT 100 charge controller
Magnum ARC
6 HES-265-60PV panels wired 3 in series into a Midnite combiner box with two MNPV 15A breakers
The panel specs:
265W
VOC - 38.3
Isc - 9.1A
Vmp - 30.8
Imp - 8.61

The installations happened in April  of this year at latitude 54 in a coastal environment, not to cold but lots of moisture around. The new equipment was kept under cover in plywood crates over the winter but not in a heated shop. 

Both of these installs experienced almost identical failures at the PT 100. On the first day of full sun they both displayed F13 (arc fault). On one installation I was able to clear the arc fault code until eventually I got an F11 code and the charge controller ceased to work (Magnum says the two codes are unrelated and the tech i was speaking with has only seen the F11 twice). On the other installation I got one F13, cleared it and very shortly afterwards got the F11 which i could not clear and this controller also quit working.  In each case the F11 fault was not present when there was little sun on the panels. Instead I was seeing the open circuit voltage (around 105V), zero for kW , actual battery voltage and zero amps in. I've been told there is no parts replacement on the PT 100 so i'll have to send them back to Magnum. Incidentally, after 36 hours of use one of the MS2812 inverters went into internal NTC fault which could not be cleared so i'm looking at replacing the FET board on that piece of equipment myself.

I have installed quite a few other PT100's (but none in 12V systems) without any problems. We went through the first system very carefully to rule out a DC side Arc fault. We'll bring an infrared thermometer out next time if that's the appropriate course of action but there is clearly something else going on here. Could it be as simple an issue as moisture?

Thanks for all and any help with this. The systems are installed at very remote kids camps and the operators are hoping for the installs to be up and running by June. It is very difficult to access these sites (4 hours boat ride one way) so when I head back out I need to be sure to be able to fix the problem. 

sparkmeredith

I'll add that the reason we went with the 12V system is because the client had ordered 12 huge non dual walled batteries from Surrette already (208 lbs each). We eventually realized that these were way to difficult to get into these sites so I bought them for another client and replaced them with 24 (12 per site) of the smaller more manageable dual wall batteries. 

Estragon

265w x 6panels ÷12v charging = ~132a vs 100a controller capacity. Not that you're likely to see STC rated output in that climate, but...

F11 appears to be an overcurrent code? Saltwater fog/spray can be quite conductive. I wonder if a combination of storage conditions and running the controller at/near limit is an issue?

BB.

Do you have a definition from them on what an "arc fault" is?

The two typical definitions are either a short between + or - and safety ground. Many times, there is a ~1 amp fuse or breaker between 12 volt/solar panel return (negative lead) and earth ground (safety ground in building or vehicle chassis). If there is a short between + and earth ground (and sometimes a short between - return lead and earth ground), will pop fuse/breaker/current sense in controller and fault.

Note for this "ground fault detect" system (sometimes, falsely called an Arc Fault system by Code Folks), a short anywhere between Hot (and Return) in the DC power system and earth ground will cause this fault (solar array, battery, or load wiring).

Or, is really an ARC fault (arcing between Hot and Return wires, Hot to earth ground, or even a failing connector/connection that is loose or pulling apart) (creates audio frequency interference that the electronics detect, and shuts down charge controller to stop the arcs)?

Note that for this type of system, it is possible for brush type motors (universal AC/DC motors--Typically used for small DC motors for RV Water pumps, some small DC fans, DC motor tools, etc.) or brushed generators to arc on the commutator and create this electrical noise too and possibly ARC Fault the controller (not unheard of for AC Arc Fault breakers to fault in a home main panel to branch circuit AF breaker that "false trip" when somebody is vacuuming).

Those are my two guesses/questions...

Arc Fault detection can be troublesome to use/debug/fix. The Ground Fault system I think is actually "dangerous" (for various reasons). The true Arc Fault detection system--Works but can be a pain until the cause(s) is(are) found.

Good Luck.
-Bill

Estragon

According to the manual, the arc-fault detects a "series type" arc fault in which resistance is unexpectedly high in the pv circuit, typically from damaged or corroded wiring/connectors.

BTW, were the connections at beakers and controllers, etc re-torqued a day or so after installation?

BB.

Wonder if partial shading of solar panel(s) could cause this fault (one array in "tree mounted"... Does the array array have vent stack shading, or other shading issues such at antenna guy wires, etc.). Solar cells go "high resistance" if they are shaded.

-Bill

sparkmeredith

Yeah we did re-torque all of the connections at the combiner box, e-panel and in the charge controller. I have really come to dislike the DC breakers (CBI?) that Midnite provides. I never feel like i've made a great connection with those things, but regardless i'm sure everything was torqued. And I was diligent in making up the home runs with the MC4 connectors, no nicks, snapped tight....

My understanding from Magnum is that the PV array is sized fine for the charge controller and that the PT 100 is self limiting for amperage, so we ruled over current out as a possible gremlin.

I'm wondering more and more  if moisture is the issue. The failure on the inverter ("internal NTC") is likely a result of "stuffing" on the FET board.... a little bit of salty moisture that creates an increasing resistance. If that were the case then the failure on the charge controller (F11 - internal power fault) , on equipment stored in the same environment (under a shop roof in a plywood crate over the winter... and my shop is literally on the beach in BC looking at alaska - dry but not heated), could also be due to a bit of corrosive moisture. I've have installed plenty of Magnum Inverters in this environment and have only had to replace a couple of boards. 

The system is small enough that I could get an infrared thermometer to search for the arc fault on the DC side without much difficulty. Does anyone have experience with this? On a sunny day would I be likely to find the "hotspot" of an arc fault? Also , we determined that the arc fault was not a result of anything happening on the AC side of the system. 

Estragon

That would be a nasty design. A bird/leaf/whatever sits in front of a panel for a while, sets off the AF, and the controller needs a hard reset?

BB.

I would also use a dc current clamp meter.

For example, zero the meter then clamp the + and - power leads from the solar array. Of no shorts to ground then the sum of the plus and minus leads would be zero amps (next to the area, next to the charger controller input, other DC load and charger pairs, etc.). A short to ground would pull more current from one power lead vs the other.

What might be easier to see.... A FLIR camera (infra red) would be faster than an IR temperature gun. (iOS or Android):

https://www.amazon.com/Seek-Thermal-Compact-Imager-Android/dp/B00NYWAHHM

I will have to read more about their attention fault system. I think I need to understand better to debug.

Bill

sparkmeredith

I did consider the shading issue as there is slight shading on the top row of cells on two of the panels in the tree mount installation. We were about 8 inches shy in our spacing (three racks of two). The other site (with the same fault conditions) has no shading at all.

I spoke with an distributor/installer this evening who had a client with this same issue (F11). He claimed that there had been a recall on a run of PT100's sold last year (I installed six last year , with two complete failures after working for about 12 hours) and that he sent one faulty unit  back to the manufacturer to be "flash programmed" which solved the fault. If this were the case it is strange that this hasn't been mentioned in my three calls to Magnum about this issue. But it is also strange that they have not been able to give me any insight as to why this may be happening. Usually these guys are awesome at troubleshooting.

Still confused. And thanks Bill I have ordered the FLIR for iOs. Seems like it will come in handy.

BB.

Let us know how the IR Smart Phone Camera works for you...

-Bill

Dave Angelini

Have you tried or used Magnums web portal for monitoring? I looked at it last year and it seemed to be missing alot of the basics that Outback and Schneider have. Any comment appreciated?