Following the discussion linked below (battery isolator switches can cause problems), I started measuring the temperature of the original Victron Battery Switch ON/OFF 275A using a temperature sensor via a GX device and VRM. The sensor was screwed onto the cable lug screw.
When the two SmartSolar MPPT RS 48/200 units start delivering power, you can observe a 1:1 correlation with the temperature rising significantly.
Over the past few days, temperatures of over 60 degrees and even over 80 degrees have been recorded almost daily.
The battery supply cable itself remains almost cold (monitored using an additional temperature sensor).
At over 80 degrees, the (calculated) charging current per switch was a maximum of 60A (60A * 4 strings = 240A total current).
The system is still a long way from the 275A continuous current per isolator switch… and in any case, only 100A * 4 strings = 400A would theoretically be possible.
Up to what temperature can the battery isolator switch be used before it melts?
the problem with isolators is that they ALWAYS have a significantly higher resistance than the cable, regardless of whether they operate electronically or mechanically. The only isolators I know of where the resistance can probably be considered negligible are NH fuse switch-disconnectors, because they have completely different contact surfaces from high-power relays! Especially when they are designed for fuses rated up to 500 A!
I think the ones I installed now cost around 60 euros each.
And fuses also always get significantly hotter; after all, they are supposed to blow in the event of a short circuit and are therefore a built-in weak point in the connection, unfortunately with a significantly higher resistance!
I’m a little concerned about how quickly the Victron battery disconnect switch heats up under relatively light loads.
The data sheet specifies 275 A continuous current, 455 A for 5 minutes…..
I also think that 83°C wasn’t the end of it yet (the sun had gone in or the batteries were fully charged), so I can slowly make sense of the following post.
I know, I’m aware of various problems with different fuses that got so hot they unsoldered themselves because they hadn’t been soldered properly. In one case, an 80 A fuse failed at just 30 A, and the corresponding fuse holders weren’t suitable for the temperatures involved either and simply melted. One holder was particularly disastrous, as the fuse holder was pressed against the metal block to which the cable was connected solely by the plastic. That was bound to go wrong!
You know, when I look at the data for contactors, I’m always surprised that you’re only allowed to connect fluorescent lamps with a fraction of the rated power to a 63 A contactor. However, I suspect this is due to the installed chokes and the voltages that occur when switching off. If that were the deciding factor, I’d actually have to install 240 A contactors. But since they only switch infrequently and the whole house is connected downstream of them, the 63 A contactors should easily be sufficient. At any rate, none of them has caused any problems so far. Only one Hager 40 A contactor once got stuck, which could have caused a nasty short circuit. Since then, the contactors have always been interlocked with each other as well, or it is monitored to make sure that one has really switched off before the second contactor switches on.
Sometimes technology just isn’t that simple. Or would you assume that it’s possible to blow a 35 A Neozed fuse through a 16 A circuit breaker?
And when it comes to Neozed fuses in particular, there are ceramic holders as well as plastic ones. I chose the latter because of their insulation. Well, at a continuous load of 50 A, the plastic started to melt, with the usual consequences: an interruption in the circuit!
I’m still trying to decide what to use as a disconnector for the DC lines, after the aforementioned disconnector melted at a maximum of 80 A (if it ever even reached that) after about 2.5 years.
What do you think of DC fuses (60 V, 125 A) for mounting on a DIN rail? Are they any good? Most of the products on offer are, of course, from Chinese manufacturers, which makes me doubt their quality again.
Otherwise, it looks like I’ll end up using NH or Neozed fuses with ceramic bodies. Thanks to @d_ferdi for sharing your experience with them.
I would go with NH fuses. In my case, because of the maximum currents, I ended up using NH1 fuses. I also found the appropriate DC fuse switch-disconnector rated for up to 1,000 V. That is admittedly a bit over the top for 48 V, but it is not a problem. I chose them because of the large contact area and the very good contact connection, and I have not found anything comparable that can also be used as a disconnector at the same time!
I also use one of those Victron 275A switches, and at ~75A I haven’t noticed any significant heating. The connected cables are each 50mm². I measured it several times with an infrared camera under full load and once the temperature had stabilised, when the system was new.
When I bought the switch back then, there was also repeated talk of “cheap” copies. They look identical and are sometimes even sold as Victron products, but apparently the quality isn’t as good. Someone on this forum even did a comparison once and found some significant differences (if I find the post, I’ll link to it here).
Anyway, that’s why I wanted to ask: is it an original Victron product? Did it arrive in sealed blister packaging with the Victron logo?
That’s unusually warm… Are all four affected in the same way, or just one? I’d probably unscrew one of them first and have a look… I can’t imagine that this would affect all the switches equally…
As I said, my switch only gets warm to the touch, even at 75 A… I can take a measurement with the IR camera later (the sun is shining nicely at the moment, so there should be enough power)…
But if all four switches are affected, I’d definitely report this to Victron. If you/we can rule out incorrect handling, it would certainly be valuable for Victron to receive this kind of feedback. Ideally, they would even get your switches for analysis.
@DirkW
This may be a really stupid question, but is the switch actually that hot when you touch it?
A sensor like that isn’t necessarily beyond all doubt either…
Thank you for your justified questions.
I only operate the system, and it’s a 300 km drive there for each round trip. The owners can’t carry out these checks themselves, so I have to do them. Therefore, I can’t currently answer many of your questions. I would also have to borrow an IR camera.
The four disconnectors were operated 5–10 times without a load when the system was commissioned 1¼ years ago. Since it’s a large off-grid system, nobody does anything with them, because otherwise you end up sitting in the dark. I remeasured the currents in the four battery cables with a clamp meter the other day, and they were almost identical at the time of measurement. The cable lugs are from Klauke and were crimped with the appropriate Klauke crimping tool, so that shouldn’t be the issue either. The screws are tightened securely.
I clamped one of the GX temperature sensors at the back onto one of two screws — it has an M10 cable lug — and secured another one with cable ties about 50 cm away on the same battery cable.
The four Pylontech cable sets have been ordered, so unfortunately I’ll be modifying the system over the next few days.
But I’ve just measured again at my end: it was “unfortunately” only around 45–55A, but for more than an hour. The switch doesn’t get noticeably warmer than its surroundings. The 50mm² cable to the switch was around 23°C, with the cellar at 18°C.
Edit: Switching without a load is always a bit of a tricky subject… Are you sure about that?
As soon as the MultiPlus units or the MPPTs are connected, and the battery has been disconnected for more than a few minutes, the input electrolytic capacitors are discharged and draw several kA when the battery is switched on. That then causes the aforementioned damage, which can no longer be rectified without reworking them.
The Pylontech batteries have a soft-start function to charge the input electrolytic capacitors in the MP2; otherwise, they go into fault mode if it isn’t used.
There’s also only 1 MP 8 K connected to 8* US5000s.
As I’ll be getting my two new batteries in the next few days (and the US5000 is being removed), I’ll be redesigning the DC wiring anyway, and after doing some research I’ve decided on the following battery disconnect switch:
Did you retighten the connections with a torque wrench?
I didn’t even know that the Victron Switch had a temperature sensor.
I only have a small installation and switch off my three Multi units via Venus OS; besides, you have to take the battery bank out of service anyway if you need to work on it once a year.
However, this only works if you close the switches first and then switch on the batteries.
If the batteries are already on and you then close the switches, you get the corresponding sparks from the inrush current at the first switch.
They don’t; the sensors were connected separately.
