Using XT1 as an emergency over-temperature sensor in LiFePO4 battery enclosure

[Helles]

Hi Dacian,

I have a 24 V / 560 Ah LiFePO4 house battery on a cruising boat, consisting of 16 × 3.2 V / 280 Ah prismatic cells in a 2P8S configuration, controlled by an SBMS0.

I am currently improving the fire/thermal safety of the battery installation. The battery will be surrounded by non-combustible thermal/fire barriers, with approximately 50–100 mm free space above the cell terminals and a controlled ventilation path from this upper space to outside the living area.

I am considering installing the recommended 10 kΩ B3950 NTC connected to XT1 in the upper gas collection area or at the entrance to the ventilation duct, rather than attaching it directly to a cell.

The purpose would be to detect an abnormal temperature rise or hot gas release from a cell and have the SBMS0 immediately disable both charging and discharging through the EXTIO controls.

I have a few questions:

1. Do you see any technical problem with using XT1 in this way rather than as a cell-surface temperature sensor?
2. Would setting Max EXT Temp to 60 °C be reasonable for this purpose, assuming normal temperatures at that location remain well below this?
3. My understanding is that after an XT1 over-temperature fault the SBMS0 automatically enables charging/discharging again after the temperature improves by approximately 100 counts (~3.5 °C). Is that correct with the current SBMS0 firmware?
4. For this particular application, I would prefer the over-temperature event to latch OFF, requiring manual intervention/reset before charging or discharging can resume. Is there an existing SBMS0 setting or EXTIO configuration that can provide such a latched shutdown?
5. If not, what would you recommend as the simplest way of preventing automatic restart after an XT1 high-temperature event?

The objective is not to use XT1 as the primary protection against thermal runaway, but as an additional early shutdown layer: if abnormal heat or hot gases reach the enclosure exhaust area, all charging sources and loads should be switched off before the situation develops further.

Thanks,
Helles

[Dacian Todea (electrodacus)]

Helles,

1. There is no issue to measure air temperature inside the battery box but that could have much higher variation than battery temperature especially that you have ventilation.

2. You will be mostly measuring the ambient temperature if you have forced air ventilation. Not sure how useful is that and if ambient can ever get close to 60C

3. The temperature limits with the new firmware have a 3C hysteresis so yes charge or discharge set limit temperature will automatically enable if temperature improves by 3°C.

4. You can use type 5 to isolate the battery and remain so until you manually reset the tripped circuit breaker.

5. XT1 high temperature even is unlikely to catch the type of issues you may be thinking about. If a cell fails internally stopping charge and discharge will not stop the cell from heating.

If you have bad connection between cells and that is the reason for the heating then yes it will help but you do not need latch for that. You will notice that your loads / charging is turned OFF and you will investigate the reason find the problem and correct the bad connection thus solve the overheating.

But if the battery box is ventilated then is unlikely you will be able to detect this since air is always exchanged and very unlikely to ever get to 60C from a bad cell connection heating.

You can test Cell's internal resistance using the diagnostic tool and connect that during initial installation and commissioning and you can regularly perform that test (say once or twice a year) to know that connections are still good.

This same test will also measure the cells internal resistance and if those are OK cell is not heating due to charge or discharge current.

LiFePO4 risk of dendrite growth is very low not comparable with other type of Lithium cells.

The increase chance of that will be due to:

- Charging below freezing 0°C

- High C-rate charging or discharging.

- Deep over-discharging (below 2V)

All of this will be prevented by the SBMS0 if correctly installed and setup.

The thermal runway is very unlikely with LiFePO4

It can happen due to physical damage, external heat or fire over 200°C or overcharging due to charger failure and no type 5 second level protection to trip a breaker and isolate the battery.

I will say adding a type 5 protection will be the most important thing for battery safety to protect in case of equipment failure like one of the chargers not stopping the charge after battery fully charged.

Temperature sensor will be good mostly to prevent charging below freezing and maybe on high ambient temperature where limit should be below 50°C but for battery longevity is should be below 35°C (of course you do not want to set 35°C high limit but you can monitor that and if you can limit the amount of time battery is above that temperature.)

If cells connections are good and C-rate is in normal limits below 150A for your battery then battery temperature will always be at most 3 to 5°C above ambient.

[Helles]

Subject: SBMS0 Type 5 – Blue Sea 7702 vs. 7717 battery disconnect

Hi Dacian,

I would like to follow up on your recommendation to use Type 5 to isolate the battery and keep it isolated until the tripped circuit is manually reset.

I am considering two 24 V Blue Sea ML-RBS battery switches for the main battery disconnect:

Blue Sea 7702 – bistable

• Requires a momentary pulse to switch OPEN or CLOSED.
• After the pulse is removed, it remains in the selected state.
• Therefore, after a Type 5 trip opens it, it could remain OPEN even if the Type 5 output later resets automatically. A separate deliberate CLOSE command would be required.

Blue Sea 7717 – Auto-Release

• Requires a continuous control signal to remain CLOSED.
• Loss of the control signal causes it to OPEN.
• My concern is that if Type 5 resets automatically after the fault condition disappears, the control signal might return and the 7717 could close again automatically.

Which of these two switching behaviours would you recommend for use with SBMS0 Type 5?

Would the 7702 bistable/pulse-operated version be the better match for the manual-reset safety behaviour you described, or would you recommend the 7717 Auto-Release with some additional latching/reset circuitry?

Also, can the Type 5 output directly provide the required control signal, or should it operate the Blue Sea switch through an interface relay?

Thanks,
Hellmut

[Dacian Todea (electrodacus)]

Hi Hellmut,

Here is the response clearly organized with the help of Gemini

1.Normal State (Running)

• BMS signal is active +24V on the Orange wire.

• The 7702 sits CLOSED.

2. BMS Detects Fault

• BMS output drops from +24V to 0V.

• The voltage transition on the Orange wire triggers the internal latch inside the 7702, snapping it OPEN.

• Power to your system is cut instantly.

3. BMS Signal Recovers

• Fault clears; BMS output goes back from 0V to +24V.

• The 7702 remains OPEN. The Blue Sea 7702 only responds to rising edge signals on its Brown wire to close. A rising voltage on the Orange wire is ignored for closing operations.

• The system stays safe and disconnected indefinitely regardless of what the BMS is doing.

4. Manual Re-Engagement

• An operator walks up to the control panel and presses a physical NO (Normally Open) Momentary Pushbutton.

• The button feeds a momentary +24V pulse to the Brown wire.

• The 7702 snaps CLOSED and restores power.

BMS Compatibility Check

• Signal Input: The control input pins on the Blue Sea 7702 pull less than 10 mA continuous sensing current when state changes.

• BMS Output Limit: Since your BMS provides max 50mA, it will trigger the 7702 switch input reliably without overloading or frying the BMS output stage.