I’m absolutely in favour of technology, I just don’t understand why knowing the “internal temperature of the Multiplus” matters except from the geek “I want to know everything” point of view – the fan assistant allows a user to do everything that can be done to help keep the unit cool.
I believe there are several internal temperature sensors (transformer, power switching stage – so which one is “temperature”?), and the MP uses these together with things like current draw and operating mode to decide when to turn the fans on and at what speed – and in my experience this happens at quite a low temperature, roughly when the ambient gets much above 25C so well before any temperature derating. At this point the internal fans turn on at low speed – which has quite a big effect compared to no fans at all and just relying on convection for cooling – and this is also when the fan assistant kicks in, which can be used to turn external cooling fans on.
Once this has happened then it’s correct that at high power levels the internal fan speed ramps up but the MP (or Quattro in my case) still carries on heating up internally, and this does eventually lead to power output reduction starting after about half an hour – I see this as reduced charging current into the batteries. But I don’t see how knowing exactly what the temperature is (which one?) helps, there’s nothing more that can be done to keep it any cooler apart from redesigning the unit – and Victron have already been doing this, the temperature derating is less bad for the MPII than the MP/Quattro (and better still for the new MPII models like the 6.5k), but in the end it is what it is.
Thanks but was looking for the efficiency curves though. I’ve seen some of those posted before but lost the reference/link. Some information/knowledge is still quite scattered over various forums and repos and hard to find.
I’d really like a replacement for the Quattro 48/10000 with less temperature derating, charging current really falls off as it heats up – but unfortunately my electrical cupboard (in the stern of a narrowboat) is not tall enough to fit the newer MPII in…
Plus an external transfer switch, since I use both generator and shoreline. I really wish Victron would update the bigger Quattros, but I’ve been hoping that for a couple of years now…
For charging we use a secondary (hf) charge system with 95% or higher efficiency, controlled by a Node-RED flow to follow the charge requests made by DESS. That was (relatively) easy to accomplice. Not so much so for boosting discharge / feed-in power though, that’s a whole other can of worms. But even with just the additional charge power, our trade only system is much better able to catch the dayly low price hours (at 9kW flat on a single phase 40A mains) while the high price hours are usually spread more widely where we can feed-in at max 4.4kW. It doesn’t even matter too much that the feed-in power starts dropping after a few hours (depending on ambient temperatures and additional cooling) as long as DESS hits the highest prices right.
I don’t want to trip over any forum rules by going into the details publicly on Victron’s own forum but for questions I can be reached on victron@av-e.nl
No! My inverter & battery room is too small. In the meanwhile it has two diffrent air conditioners to keep the Multiplus in operation with power specified in datasheet. 6 Multiplus started with a 12000 BTU machine. After increasing to 9 Multiplus it was too small in summertime. If room temperature is over 30 deg celsius, the heatsink seems to have 60 deg and available power goes down very fast.
Recently I added 18000 BTU machine. Both air conditioners are now able to keep the room at constant 20 deg plusminus 4 where the Multiplus datasheet is specified. For this meassurement, I use the Pylontech battery internal temperature sensor. Although the Pylons are only a minor source of heat, they do have internal temperature sensor available for CAN protocoll. Multis are a strong source of heat, after them the RS450. Both do not have any temperature signal in their protocol.
