Victron disconnect quirk causing temperature sensor failure?

Saipro

I've been on my semi-off-grid PV coupled RE setup for a few years now. Began with merely inverters and batteries about 2 decades ago then added PV about 5 years back. I have a dual-system running (one for critical loads and the other for general loads).

Previous year, I changed both inverters to Victron (MultiPlus 48V 5kVA 230V and Phoenix VE Direct 48V 375VA 230V). Both still work. Last year, the Phoenix developed a peculiar problem while I was away. The fans would spin at max revs all the time, whether on or off. Initially, I felt it had to do with the rising summer temperatures but blasting it with my blower didn't help. Did all forms of troubleshooting including various disconnect/reconnect tricks, firmware reload, starting with a weak battery, etc to no avail. With a month to go on the warranty expiration, I reached out to Victron and after a few weeks of trading emails and remote troubleshooting, a replacement unit was sent.

Now for the amazing part, I returned from another trip to discover the MultiPlus was giving the same problem. What were the odds? I sat down with the missus and narrowed down the common denominators. One thing stood out: both were in use when the missus would switch them off (load still running) when leaving the house. Sequence being power down inverter -> switch off loads -> leave the house. On returning, switch on the inverter -> power on the loads. I should state at this point that the Phoenix is never loaded above 60% and the MultiPlus never above 30% (unless I'm around, doing some stuff in the workshop). It would appear the Victrons are intolerant of powering down with loads on, however small the loads (for the MultiPlus, 400W max out of 5kVA capacity!). My Chinese MustPower inverters did the same job (under the same conditions of use by the missus) for many years thus my being unaware of the load trip issue.

Two questions (1) Victron CC have a reputation for surviving any kind of battery line cut during full charge; why would the inverters fail under a load trip {unless I'm mistaken on the coincidence and the fault is really due to something else}. (2) Has anyone else experienced this and what might be the remedy? (I've given out the moderately noisy Phoenix to someone to tinker with but put the MultiPlus in storage as the noise of the fans at full revs are incredible - and I hope to fix it someday)

I'm back to my trusty MustPower 4kVA for now but would like to return to the fully programmable Victron MultiPlus. All suggestions (asides using a hammer) would be welcome.

Estragon

I almost always have (small - eg. a radio and couple lights) loads on when turning inverters on/off.  It confirms the state change for me.  My Outbacks and Morningstar seem fine with it.

Don't know about Victron, but I kinda doubt that's the problem unless there's something like big motor loads running?  I could see it being a problem if the inverter was running hot, and then shut down with no fan to move the heat.

Have the fans always been so noisy?  I wonder if bearings are going.  That could be noisy, as well as making heat and taking more power to run.  Spiders etc also like to take up residence in inconvenient spots.

Saipro

Fans are noisy only because they run at full revs. Checked bearings and they seem to have no issue. On 3 different occasions, the fans on the MultiPlus revved down to normal and the over-temperature warning light went off. On all 3 occasions, some ironing was being done (a 1kW pressing iron) and the iron would have been going on for more than 30 minutes. On all occasions, battery voltages dipped a bit more than usual (batteries never discharge past 82% thus are nearly always full).

Or could the original problem be related to shallow discharge cycles (typically 88% - 92% on most days, 82% on extreme days and a single incidence of 79% as reported by WizBang Jnr.)

Estragon

Interesting.  You'd expect the inverter would be running warmer when more heavily loaded and/or with lower DC input voltage, which should increase fan speed.  Instead, it apparently does the opposite.

Makes me wonder if there's a problem with the sensor or related circuitry.  Presumably, absent "sane" temperature input from the sensor, the default would be to run the fan on full.  Maybe running the unit warmer gets the sensor into a sanity-checking range, in which the fan runs at a speed appropriate to that temp?

It seems to me many temp sensors are pretty cheap (<$1) and can sell in lots of thousands per order.  Victron may use the same sensor part across product lines, and maybe got a bad batch or faulty integration/installation?

Saipro

Estragon said:

Interesting.  You'd expect the inverter would be running warmer when more heavily loaded and/or with lower DC input voltage, which should increase fan speed.  Instead, it apparently does the opposite.

Makes me wonder if there's a problem with the sensor or related circuitry.  Presumably, absent "sane" temperature input from the sensor, the default would be to run the fan on full.  Maybe running the unit warmer gets the sensor into a sanity-checking range, in which the fan runs at a speed appropriate to that temp?

It seems to me many temp sensors are pretty cheap (<$1) and can sell in lots of thousands per order.  Victron may use the same sensor part across product lines, and maybe got a bad batch or faulty integration/installation?

Your logic makes tons of sense.

Saipro

I eventually found someone who seems to have a related issue here

Saipro

Problem fixed. It was a hardware problem. Fixed by service center. Issue resolved.