# Interacting current sources

Suppose I have two current sources as depicted. One current source nominally delivers a current of frequency 1000 Hz and the other nominally delivers a current of frequency 1010 Hz. The nominal current amplitudes are equal. Also R1 = R2 and L1 = L2.

If the current sources were ideal then the coupling between inductors L1 and L2 would be irrelevant to the circuit behavior. But if the current sources are real (not ideal) then I assume that the coupling between the inductors would matter with respect to the circuit behavior.

What sort of departures from the ideal case would one expect to see for real current sources? I realize that this question depends upon the type of current source (ie. the circuitry that tries as best it can to maintain ideal behavior) so anything you can tell me about the expected behavior for the different types would be gratefully received.

• Without dot notation this cannot be judged. Commented Jun 13, 2017 at 18:13

The main relevant parameter for a constant current generator is the compliance voltage - this is the voltage at the load for which the current will depart from the intended current by some defined error amount.

In your question one of the generators will create a voltage across the load of the other generator that will cause its current to vary if the voltage is large enough.

What is the reason behind your question?

Current sources, especially AC ones, are unusual and VERY rarely used in practice. I have only once used an AC current source in 40+ years of electronic design.

The nearest thing to an AC constant current source in common use is to feed an AC voltage source through a resistor that is high in relation to the load. This was the technique normally used to drive the recording head in analogue tape recorders. The inductance of the head would cause the recording current to drop at higher frequencies if that was not done. The resistor value was chosen to be high relative to the inductive reactance of the head at the highest frequency needed (~10-20kHz for audio).

• Kevin. Current sources are used in magnetic resonance imaging to control the imaging gradients and in transcranial electromagnetic stimulation. They would also be advantageous in transcranial magnetic stimulation. Commented Jun 13, 2017 at 19:26
• I am most interested in whether new frequencies (eg the difference frequency) would emerge as a result of the coupling and non-ideal behavior of the sources. I didn't make this a part of my initial question because I didn't want to limit the scope of the responses. Commented Jun 13, 2017 at 19:31
• I understand that one circuit could cause the other to exceed compliance voltage. But this situation is a bit more complicated due to the fact that this load varies in time, isn't it? Commented Jun 13, 2017 at 19:35
• Yes - there almost certainly will be a dynamic component, or the compliance voltage will be frequency dependent. But it would depend upon the particular implementation. You are asking the question as if there are off-the-shelf designs you can buy. In general the applications are rare enough that they are designed with specific uses in mind - such as in MRIs. Commented Jun 13, 2017 at 20:41
• Thanks, Kevin. Sorry if I gave the impression that I was looking for an off-the-shelf solution. I am not (at least not yet). Presently I am simply trying to understand what sort of coupling problems might arise when using real current sources because, yes, my interest is in one such rare implementation. Commented Jun 13, 2017 at 21:00