Lack of continuous temperature measurement below the freezing point

I have the batteries, inverter, and chargers in two cabinets, each heated with 400 W heaters, and the cabinets are located in an unheated hall. Currently, when the SOC is low and a frosty night is approaching, it’s better to shut the system down for the night and keep some energy for the morning to heat the cabinets, rather than allow the batteries to discharge to a very low level.

I’ve noticed that below 0 °C, Batrium shows a temperature of –40 °C and begins to display correctly only once the temperature goes above freezing.
When this first happened, I began to suspect the NTC sensors, but now it looks to me like an incomplete implementation of the NTC sensor characteristics in the firmware. But, even if a 3950 NTC sensor becomes inaccurate below zero, it’s still better to have a reading than to have no reading at all. Maybe it would be a good idea to select an NTC with a different Beta values from the dropdown list?

The lack of continuous temperature measurement is a major inconvenience when operating a photovoltaic system in a colder climate.
What’s more, the Toolkit software allows setting any temperature below zero in the Thermal Control tab, but the sensor only provides feedback down to 0 °C — below that, it jumps straight to –40 °C.
At least that’s how it looks in my case :thinking:

I’m rather annoyed by Batrium’s temperature readings. The system is considered a high-end BMS, yet in this area it behaves like a cheap Chinese one.
The actual battery temperature is around –1 to 0 °C, but… a temperature of –40 °C is being sent to the Victron system, the Toolkit software shows +1 °C, and in the BMI the min and max are both +1 °C — yet the average is somehow –40 °C. Meanwhile, the temperature of each individual cell is shown as –5 °C.



I dont get this issue.. thou not sure how it works out 30deg from those readings.

and in BMI:

Hi wmichale,

I’ll run some tests in the freezer to confirm this for you, it’s true that B3950s get less accurate below 0, but -40 is the same reading we get when it’s not plugged in at all :thinking:

Expect a response in a couple of days

Hi wmichale,

Ran some tests today, and couldn’t reproduce that strange temperature behaviour

This could be due to a weak freezer in the office, but I think the test was worth it anyway.

Could you please email info@batrium.com with your WatchMonCORE serial number, and make sure it is provisioned, and we’ll take a closer look at your logs over there.

This is a situation from this morning when temperature of the battery dropped to 0 — both the Victron and BMI screenshots cover the same timeframe.


I’m glad the matter was confirmed and that I didn’t create unnecessary drama :wink:

1 Like