AC Coupling

There is no way that the DAC will produce a perfect sine wave with no DC residual, so the current source output tries to force the cap to pass the DC residual by saturating at one peak of the sine wave. Are you sure you need the blocking capacitor or a current source output, instead of a voltage output?

That depends on whether the INA and opamp are trying to produce an approximation of a sinusoidal current (both swings) or a raised (all in one direction) DC current with a sinusoidal variation from zero to the peak to peak value of the sinusoid. If the second case, then, yes, the amplifier should stay saturated. Since it, apparently, is not, I assumed that the O.P. is amplifying a bidirectional signal that is nearly absent a DC bias, so what it clips off removes the net DC so the capacitor can pass all of what is left of an average cycle.

I reckon the coupling cap is causing enough phase shift that the configuration may be marginally stable. Note that there's an extra op-amp in the feedback loop - this alone makes it 'touchier'.

You're seeing 'grass' on the waveform ? Probably oscillation due to stability issues.

Increase the coupling cap size and see if it goes away. I assume it's not there with no coupling cap at all ? You also want to use a decent quality film cap in this application.

Graham

Good point. But wouldn't that mean that the op-amp output would be saturated

• or - with no signal in such an event ?

Graham

You are using a current source which is *not* the right circuit for what you want to do. There are other ways to achieve a symmetrical signal without DC-component. You connect your amplifier +in(3) to the output of the DAC and the -in(2) to the reference voltage(5V). pin5 goes with the O/P(6) and Pin1 is gnd to gives the output potential. Then you do not need a capacitor and the O/P level is +/-5V with a 0...10V output range of the DAC. If you want a precise offset trimm, you can add the pot as in Fig.2 If your DAC has a smaller O/P range, you better amplify at the O/P of the differential amp with another amplifier, because the mached resistors in the 2133 should be used.

Hello Ban,

Would you adivce me that why its not a right current source to use!

Thanks John

John P> > John P>

Hello Ban,

Thanks for ur reply! I am > john wrote:

A current source looses a lot of the available supply voltage for its circuitry and gives little compliance. It will also saturate when no load is connected, and then hang some milliseconds on one rail. This produces the effect you describe. If you really need an output current it will be better to produce an unipolar current as your DAC does and subtract a constant current(Imax/2) from the output, so the O/P can be bipolar. A servo amp can regulate the constant current without any capacitor to any desired mean value, which includes 0.

is

Not necessarily. You can play the same game with a bipolar output current source for the DAC and for the nulling current source. This will avoid the Vsupply.Idc permanent power dissipation.

Also note that this _is_ equivalent to bypassing the DAC current source output with an almost perfect inductor, which of course gets ride of the DC offset component.

John, I see that your output capacitor was set to 40nF so your frequency might be not too low to use a real inductor. All that of course in case you really need a current source output.

Why don't you advise us on the purpose of the circuit instead of asking someone to fix your kluge of an hunch.

Hello,

OPA131 is creating the problem. for some reason it is getting the

-15volts at its non-inverting input and its output is connected to the Reference pin ( pin 1) of the INA133 which adds this -15volts to the output. So when there is no output from INA2133, the opa131 has negative 15 volts at its non-inverting input. Please advice, how to get rid of this

-15volts!

Thanks John

It might be you who is creating the problem.

Do you still have a capacitor between the current output and load? If so, you are asking for the impossible. The INA and opamp try their best to pass current through the load, but the capacitor cannot pass even a trace of DC, so you have an immovable object working against your unstoppable force. But the force ends when the INA hits its negative power rail, and the immovable object wins.