OP-AMP Input Protect

[Suraj143]

Guys I made this circuit to sense current on a motor.The trip point is 500mA. When the motor reaches above 500mA AC current I need an interrupt.The motor varies current from 0 to 3A.

The circuit works fine.The only problem is when the current sense resistor(R3) opened due to failure the remaining circuit blows up.

How to protect it?

 

[Mike odom]

wow, yes, it would, you have 230V running through your opamps!!!!

There shouldn't be much current in your sense amp loop, so I'd suggest putting a 10mA polyfuse before R2, that way, if your sense resistor opens, then the fuse will open before anything gets cooked.
But then, that's the problem, at 600V, they only have 150mA and 160mA hold currents, and at 250V, 80mA is the minimum... but I think 80mA will fry the opamp, especially on the input side. You could use a regular fuse (poly fuses are automatically resettable when the fault goes away). Of course, I just looked on digikey, they may have others...

 

[Misterbenn]

I think a instrumental amp, a set up than can cope with high common mode voltage, might cope a bit better.

Alternatively use a current transformer.

 

[Mike odom]

Yes, I think a current transformer is called for here. Good call.

 

[Suraj143]

This must do without using CTs.

Reg above circuit
If this is sensing the neutral wire then the voltage drop will be very less isn't it?

No more burning OP-Amp side I believe.

 

[Misterbenn]

Nope.

Your measuring a differential voltage that is proportional to current. Doesn't matter if it's in live or natural it will have the same volt drop.

Why can't you use a current transformer? How about a hall effect sensor, that's what I use in motor control.

If you must use op-amps I'd suggest a instrumental amp construction with a fuse and zener clamp on the input.

 

[Misterbenn]

Or as a really dumb solution build your sense resistor from 3 low value resistors in parallel. Each resistor being sized for 150% of the Peak current.

This isn't safer but it is more robust.

 

[Suraj143]

Guys in my original design I gave Live & Neutral like in the 1st schematic.It was the one blowed up after working 3 months continuously by opening the sense resister.

What about swapping the Live & Neutral like in the 2nd schematic.Will it still blow the op amp when the sense resister opens up...!!!

 

[ericgibbs]

Hi
Do you want to keep the motor running when and if the 0.1R shunts resistors fails open circuit.?

If yes, place two 5Amp mains diodes in anti parallel across the shunt resistor.
Or
you could split the 10K input resistor into say two 4k7 in series and connect two signal diodes in anti parallel from the junction to 0V.
The high side 4k7 would have to be rated at ~ 10Watts

 

[ericgibbs]

hi
In your latest circuit #9, if the 0.1 R fails o/c the whole of the OPA and MCU will be at 230Vac LINE supply

 

[ronsimpson]

USE HIGHER WATTAGE RESISTOR.

Many op-amps can withstand 2mA of input current. You have a 10k on the input. At 20 volts across your burned out resistor you hit the 2mA limit. Check your part's data sheet part for maximum input current. You could increase the value of resistors around the first amp.

If you are going to use a fuse, don't put it in the input of the op-amp. The amp will die long before the fuse. Fuse the motor not the amp.

The input of the amp should never be far from ground. You can add input protection diodes. Some people put them from input to the supply voltages. In this case I would not. Place two diodes in parallel (one pointing up, one pointing down) from input of amp to ground (neutral). This way the max voltage to the amp will be 0.7 volts. (well more like 1 volt). Now the 10k resistor will get TOO MUCH CURRENT and burn out. The resistor is now a fuse. Better replace a resistor than a amp.

 

[Suraj143]

Hi eric thanks for the suggestions. It was helpful.

 

[Suraj143]

USE HIGHER WATTAGE RESISTOR.

Many op-amps can withstand 2mA of input current. You have a 10k on the input. At 20 volts across your burned out resistor you hit the 2mA limit. Check your part's data sheet part for maximum input current. You could increase the value of resistors around the first amp.

If you are going to use a fuse, don't put it in the input of the op-amp. The amp will die long before the fuse. Fuse the motor not the amp.

The input of the amp should never be far from ground. You can add input protection diodes. Some people put them from input to the supply voltages. In this case I would not. Place two diodes in parallel (one pointing up, one pointing down) from input of amp to ground (neutral). This way the max voltage to the amp will be 0.7 volts. (well more like 1 volt). Now the 10k resistor will get TOO MUCH CURRENT and burn out. The resistor is now a fuse. Better replace a resistor than a amp.

Very good answer.

I used a cheap carbon 2W resister as in the picture.I will use a ceramic resister.

I inserted two 1N4007 diodes to the input.I doubt whether it is connect to "A" or "B".

 

[ericgibbs]

hi,
The diodes are not connected correctly

Like this image across the 0.1R they must be able to carry the full motor current in case of 0.1R o/c

 

[ericgibbs]

hi,
Connected this way.
This SIM shows the result of a 0.1R o/c after 0.5seconds.
Use at least 5A rated diodes.
E

 

[ronsimpson]

Now there are two options:
1) input of op-amp to protect the amp and let the resistor (10k) open up. My idea.
2) across the 0.1 resistor. In this case you need BIG diodes. They must take the stall current of the motor.

3) Do both. yes do both

 

[Suraj143]

Hi thanks for both of you.

I think due to stall current of the motor the cheap carbon film resisters burned out.It is better to use 0.1R/5W ceramic resisters first.

 

[Suraj143]

The 10K resister will burn if I place the diodes after the resister.If I place the diodes before the 10K then it will never burned.

 

[ronsimpson]

The 10K resister will burn if I place the diodes after the resister.If I place the diodes before the 10K then it will never burned.

Yes, assuming the current never goes above stall current. If the motor burns out or if someone mis-wires or something happens that causes the 20A breaker to pop....lightning happens....I have seen wires milt, PCB traces to fuse open......for a very small price you can force a very bad situation to not kill the amplifier.