@KSTR: I disagree. This is an I to V circuit, and does not resemble the circuit you posted. It converts the input current to voltage.
The OP: the circuit should work, but I would be concerned the Vos of the op-amps. Shorting them may cause Ib offset to flow, causing a DC offset on each of the outputs. So a tiny resistor in front of the inverting input may be a good ide to limit that.
The OP: the circuit should work, but I would be concerned the Vos of the op-amps. Shorting them may cause Ib offset to flow, causing a DC offset on each of the outputs. So a tiny resistor in front of the inverting input may be a good ide to limit that.
I suppose that would depend on the DAC jagwap maybe? My experience has been that current output DACs are always sourcing, or sinking, a set amount of current and thereby inherently cause an in-built offset at the output of the IV opamps.
I guess you'd need very good matching between the two IV resistors.
I guess you'd need very good matching between the two IV resistors.
Probably because things are ideal in a simulation, without offset.
I just simulated a similar circuit, using OPA1612. V1 = V2 = GND. Rgain = 0 (shorted). Voltage generator connected to the (shorted) inverting inputs via a 1k resistor, so like a current source.
With no offset it works OK.
But with 1mV DC between the two inverting inputs I get -4.27V on one op-amp output and +4.28V on the other op-amp output.
With ideal op-amps it becomes even worse. I got +/- 8.8TV on the outputs. 😱😀
So, I agree with KSTR. It won't work.
What is the intention behind using two op-amps by the way?
Also that the current out of the DAC: How is that split between the two opamps? How to ensure it is equal?
The non inverting inputs are effectively shorted to ground. so any current will flow into both, but not defined equally. So again, an small input resistor may help.
Yes. Accuphase used to use such topology. It increases the SNR to some degree.
You will have to watch for few things (eg. current sharing), but in general it should work.
You will have to watch for few things (eg. current sharing), but in general it should work.
From the opamp's perspective it is the same circuit. Effectively both non-inverting pins are connected together and both inverting pins are also connected together.
This only works in simulation with 100% identical units where there will exist no difference signal that is subject to the infinite gain (actually, the open-loop gain). The moment you have any difference signal from non-matching offsets (in the simplest case) or any other dynamic mismatch, the infinite gain will make the outputs go slam into the rails.
It doesn't work and that's one of the first things you learn when working with opamps, you cannot simple parallel them by connecting their inputs to each other.
The only way this circuit can work halfway is with some series resistor in each of the inverting inputs which must reduce the differential gain to low enough levels so that offset and other mismatches don't have any dramatic impact. But then it stops being a classic I/V with zero input impedance...
Because they will be exactly identical in the simulator, same input offset voltage. Real opamps will not have the same offset, nor the same tempco of offset, so any attempt to parallel opamps directly at the inputs will latch immediately to the rails or worse due to the small offset being multiplied by the open loop gain of the opamp (typically in the range 100,000 to 10,000,000) - so even 0.1mV difference in offset will blow up 10 to 1000V at the output.
However you can parallel closed-loop opamp circuits happily, so long as you recombine the outputs with enough resistance to allow for variation in offsets (and gain variation due to resistor tolerances).
