Tian power BMS is reading amperage incorrectly

[aren]

According to my inverter, my two batteries are producing ~650W of power right now. At 52 volts, that's 12.5 amps, or 6.25 amps each.

But the logs are showing 648 Amps per battery. That can't be right.

2024-11-05 08:31:07.719 | DEBUG | Thread-10  | o.ModbusRTUTransport:266   | Sent: 01 04 10 00 00 17 B4 C4
2024-11-05 08:31:07.789 | DEBUG | Thread-10  | o.ModbusRTUTransport:443   | Response: 01 04 2E 14 57 FD 44 1A CD 00 C5 00 D2 00 00 00 00 1E 00 02 8C 03 E8 29 16 00 0C 27 10 0C B8 0C B5 00 00 00 C8 00 BE 00 DC 00 00 15 E0 27 10 00 00 22 E9
2024-11-05 08:31:07.789 | INFO  | Thread-10  | verter.BmsToInverter:320   |
BMS     SOC       V       A      CellMinV        CellMaxV       CellDiff
#1       65.2   52.0    648.3   3.253(#0)       3.256(#0)       0.003
#2       64.2   52.0    648.4   3.253(#0)       3.256(#0)       0.003

2024-11-05 08:31:07.789 | INFO  | Thread-10  | verter.BmsToInverter:390   | BMS alarms:
        NONE
2024-11-05 08:31:07.789 | INFO  | Thread-10  | verter.BmsToInverter:268   | Reading BMS #2 TIAN_MODBUS on /dev/ttyAMA0...
2024-11-05 08:31:07.839 | DEBUG | Thread-10  | o.ModbusRTUTransport:266   | Sent: 02 04 10 00 00 17 B4 F7
2024-11-05 08:31:07.911 | DEBUG | Thread-10  | o.ModbusRTUTransport:443   | Response: 02 04 2E 14 57 FD 4D 1A 11 00 C8 00 D2 00 00 00 00 1E 00 02 82 03 E8 28 98 00 24 27 10 0C B9 0C B3 00 00 00 C8 00 C8 00 DC 00 00 15 E0 27 10 00 00 4F CA
2024-11-05 08:31:07.911 | INFO  | Thread-10  | verter.BmsToInverter:320   |
BMS     SOC       V       A      CellMinV        CellMaxV       CellDiff
#1       65.2   52.0    648.3   3.253(#0)       3.256(#0)       0.003
#2       64.2   52.0    648.4   3.251(#0)       3.257(#0)       0.006

[aren]

Ah here's the issue. The bytes are FD 4D. The code is interpreting them as unsigned short, which is 64845. However, the TianPower BMS docs show this is a signed short, so it should be -691 (or -6.91A).

This makes sense, because the cells are producing power so the current would be negative.

[ai-republic]

Hi @aren,
could you attach the Tian ModBus spec again. I can't seem to find it - thanks!
I'm wondering if the registers are actaully all 2 bytes long (u_int_16 or s_int_16) and not 4 byte (u_int_32 or s_int_32). I've got a lot of specs that are basically the same as the Tian and all use 2 byte registers which are also really the ModBus standard.
But I remember my surprise when I implemented it that they were stated as each register holding 4 bytes

[aren]

See #71

[ai-republic]

Thanks, I found that in my specs collection but wasn't sure which manufacturer that belongs to 😉
Here it actually states the registers are holding 2 bytes - hmm, where did I see that with the 4 bytes 🤔

[ai-republic]

I changed it according to the spec. Could you do a Clean install and check what the Tian BMS binding comes up with?

[aren]

It's way off now.

2024-11-13 19:44:07.899 | DEBUG | Thread-10  | o.ModbusRTUTransport:443   | Response: 02 04 2E 14 82 00 00 15 A4 00 AF 00 C8 00 00 00 00 1C 00 03 D3 02 30 16 19 00 1A 27 10 0C D3 0C CF 00 00 00 B4 00 AA 00 C8 00 00 15 E0 27 10 00 00 07 AA
2024-11-13 19:44:07.900 | INFO  | Thread-10  | verter.BmsToInverter:320   |
BMS	SOC	  V  	  A  	 CellMinV 	 CellMaxV	CellDiff
#1	 0.0	0.0	52.5	0.0(#0) 	0.0(#0)	0.0
#2	 0.0	0.0	52.5	0.0(#0) 	0.0(#0)	0.0

2024-11-13 19:44:07.900 | INFO  | Thread-10  | verter.BmsToInverter:388   | BMS alarms:
	BMS #1:	WARNING -> PACK_TEMPERATURE_HIGH
	BMS #1:	WARNING -> PACK_TEMPERATURE_LOW
	BMS #1:	WARNING -> DISCHARGE_CURRENT_HIGH
	BMS #1:	WARNING -> CHARGE_MODULE_TEMPERATURE_HIGH
	BMS #2:	ALARM -> FAILURE_SENSOR_PACK_TEMPERATURE
	BMS #2:	WARNING -> ENCASING_TEMPERATURE_HIGH
	BMS #2:	WARNING -> PACK_TEMPERATURE_LOW
	BMS #2:	WARNING -> DISCHARGE_CURRENT_HIGH
	BMS #2:	WARNING -> CHARGE_MODULE_TEMPERATURE_HIGH
	BMS #2:	ALARM -> FAILURE_COMMUNICATION_INTERNAL

[ai-republic]

oh 😮 - let me change it back and try the s_int_16

[ai-republic]

@aren please try now with a Clean install