bidirectional dc current sensing with 358 op amp

[Tunelabguy]

OK, let's look at the numbers. The inverter supplies up to 50 Amps. The shunt is 0.001 Ohms. So the maximum shunt voltage is 50 mv. As someone pointed out before, if you want 0.1% accuracy, you need an input offset voltage of less than 50 uv, which is really hard to do in the presence of high power circuits. Also, it is very hard to get a 0.001 Ohm resistor accurate to within 0.1%. That means knowing the resistance to within 0.000001 Ohms. It would take some very careful 4-wire connections to the resistor to accomplish this feat.

A more dependable solution is to raise the value of the shunt resistor. Of course it will have to be a power resistor now, maybe even with a heat sink. But at least you can measure it, control it, connect with it, and do something easy with the voltage generated. A shunt resistor of 0.01 Ohms would dissipate 25 watts, with a maximum shunt voltage of 0.5 volts. If your inverter is supplying power at 230 or 120 volts, it most likely can afford to sacrifice half a volt at full output.

 

[FvM]

0.1% was a suggestion in the discussion, not a specification by the original poster.

Although not completely impossible, it reduces the selection of available shunts to a few very expensive special manufacturers. 1 mOhm/1 or 2 % would take the specification to general purpose range, but still requires a precision OP. Relying on LM358 (typical 2 mV, maximal 5 mV Offset) involves either very poor accuracy, high power dissipation, or at least the necessity of individual offset adjustment. Sensing of low DC voltages is not really difficult in presence of power electronics if some precautions are taken. 4-wire shunt connection is obligatory of course.

 

[sahu]

0.1% was a suggestion in the discussion, not a specification by the original poster.

Although not completely impossible, it reduces the selection of available shunts to a few very expensive special manufacturers. 1 mOhm/1 or 2 % would take the specification to general purpose range, but still requires a precision OP. Relying on LM358 (typical 2 mV, maximal 5 mV Offset) involves either very poor accuracy, high power dissipation, or at least the necessity of individual offset adjustment. Sensing of low DC voltages is not really difficult in presence of power electronics if some precautions are taken. 4-wire shunt connection is obligatory of course.

my big essu is not very poor accuracy .that is , how can find it with 1 adc using opamp 358 .

 

[FvM]

What I would do:
- use a single OP differential amplifier with 4 resistors, e.g. gain = 40
- state-of-the-art four-wire connection to the shunt
- set zero to mid supply (connect reference resistor to 2.5 V)
- perform zero adjustment in software, store in EEPROM

 

[sahu]

What I would do:
- use a single OP differential amplifier with 4 resistors, e.g. gain = 40
- state-of-the-art four-wire connection to the shunt
- set zero to mid supply (connect reference resistor to 2.5 V)
- perform zero adjustment in software, store in EEPROM

pl give me here code example .

 

[naseerak]

Why don,t you use 75mv shunts available in local market.

- - - Updated - - -

Use inverting input for the negative voltage from the shunt and positive input for the positive input from shunt. OR it with diode to give it to a single ADC.