Hi.

So I'm trying to design a power supply in MultiSim, but since some of the parts I want to use are not in the libraries I had a go with LTspice. I'm having an issue in LTspice where I can't get the differential amplifier output voltage equal to the voltage dropped accorss the sense resistor. In MultiSim it seems fine.

In the MultiSim image, xMM1 shows 1V dropped accross a 1Ω sense resistor - R5.
xMM2 shows the output of the diff amp U3 is 1V which is what I was expecting.

However, when I tried to simulate something similar in LTspice, I can't get the diff amp to work as expected.

 

In both simulations your diff amp have a positive feedback instead off negative one. So, this circuit will never work in real life.

Also do not forget that the voltage at the opamp input cannot be larger than Vcc, in reality, the input voltage needs to be lower then Vcc. And form what I see in your circuit you have Vcc = 5V for the opamp and 18V at the opamp input, not good . The magic smoke for sure will escape.

 

First some comments and questions.

LTspice is correct. The circuit does not work as you are expecting.

I don't know why Multisim is giving different results than LTspice.

Why are you using a LT1259 in the LTspice simulation? This is a _very_ fast current feedback op-amp. It does not work the same as the 741 which is a traditional voltage feedback op-amp. Try a LT1013 instead. (You may have confused current feedback amplifiers with current sense amplifiers).


Now for your main problem. You are trying to put about 18 volts into the inputs of U3. This is far beyound the allowed common mode voltage allowed on the inputs.

There are several solutions to this problem.

The first is to use voltage dividers from the 36 volt signals on the 1 ohm resistor. The voltage dividers must get the voltages near ground for U3 to work.

Another solution is to power the op-amp, U3, from voltages centered around the 36 volts and then use a voltage divider on the output of U3 to get its voltage small enough to properly work with the input of U1A which also needs its inputs near ground.

A third option is to us an op-amp that allows voltages to be near its positive supply. This amplifier has to be able to run on 36 volts of total power supply voltage.

There are op-amps specifically intended for applications such as yours. Linear Tech makes some. Look for current sense amplifiers in the Linear Tech catalog.

 

First some comments and questions.

LTspice is correct. The circuit does not work as you are expecting.......

A third option is to us an op-amp that allows voltages to be near its positive supply. This amplifier has to be able to run on 36 volts of total power supply voltage.

There are op-amps specifically intended for applications such as yours. Linear Tech makes some. Look for current sense amplifiers in the Linear Tech catalog.

Ok so I chose another opamp and made sure there was negative feedback and the circuit started to work in LTspice.

Then I found the LT6106.
I am trying to set this up for a 0 - 30V PSU with limiting from 5mA - 4A in 5mA steps.

The LT6160 output will be connected to a 10bit ADC.
LT6160 Output 0.488mV per 5mA through Rsense.


I think I have managed to do that staying just within the max ratings for the LT6160. It's a pretty simple IC to work with I think but with my level of experience I don't trust myself not to have Fu** something up.

 

Please upload your simulation files including MultiSim and LTspice then our members can simulate the files to test the result, you can compress the files to as *.zip.

 

Hi.

OK, thank you. LTspice file attached.