I have attached a drawing of a 4-20mA current loop circuit. This has been used by my colleagues in the past, however I am having some trouble with oscillations on my output.

I'll give a quick explanation of the circuit.

• The circuit is designed using negative feedback to control a transistor.
• Let's say floating ground is at 0V for discussion purposes. Therefore earth ground will have a negative voltage.
• Based on ideal opamp priciples, connecting the negative input to 0V should create a virtual short and also connect the positive input to 0V.
• Setting Vsignal will generate a small current across Rset. (Vsignal - NULL_NODE) / Rset = Iset
• Assuming infinite input impedance, all the current will flow across Rbias.
• The voltage drop across Rbias will equal the voltage drop from floating ground across Rsense. (NULL_NODE - VOUT) / Rbias = -VOUT / Rbias
• Iout = Ibias + Ilimit + Iquiescent, where I quiescent is the current sunk into the floating ground.

Anyways, I've made all sorts of attempts to stop the circuit from ringing. I've added bypass to components, added filter caps to the feedback, moved my transistor from the 12VDC rail to 24VDC (which helped significantly). I've changed the opamp and transistor to different components and have changed some ceramics over to tantalum capacitors. I am just wondering if anyone out there sees anything grossly wrong with the circuit or has any suggestions for improving things. I don't have a strong analog background and am not sure where to go from here.

On a side note, the circuit itself is on a 2 layer PCB that is roughly 2"x2". The opamp circuit has a tight feedback loop.

When applying a 3.0V to Vsignal, I'll get a >2V output oscillation in the 1-3kHz frequency. I am able to change the frequency and the shape of the oscillations, however they are still significantly large.

Please let me know what additional information is necessary to understand what is going on. Thanks in advance.

 

Your circuit is not correct. There is no current sense feedback. The feedback resistor is in the wrong place. You don't need Rbase.

 

I'm not sure I understand what you mean by no current sense feedback? Where should the feedback resistor be located? Rbase was added to limit the base current. I'm not sure if it is really necessary or not.

 

Your sense resistor is effectively grounded on both ends and won't develop a voltage across itself with no current flowing.

Limiting the base current will limit the total current through the transistor to roughly 10X the base current. Limiting the load on the op-amp may or may not be necessary.

 

Here is a circuit for a current source:

 

In response to Wayne, I believe I understand you on Rbase. I've removed that resistor in my trials but the only effect it has had was changing the frequency of my oscillations.

As to your other comment about the Rsense being grounded at both sides...if the transistor is sinking say 20mA into the floating ground node, there will be a 0.602V drop across Rsense. Therefore Vout will be at -0.602V. This will be the same drop across Rbias. In this circuit, the floating ground reference changes based on the load current. Therefore, if the current is 20mA, floating ground will be at 499 * 20mA ~ 10V and if the current is 4mA, floating ground will be at 2V in reference to earth ground. I'm not sure if I'm explaining this clearly.

MrChips, can you explain how I could modify the circuit you provided to work with a floating ground, where the load is referenced to earth ground?

Thanks for everyone's help so far.

 

I don't know what you mean by "floating ground".

 

Don't you think that R1 in the original schematic should be samller to ensure an Iload of 5 mA? 2k2 I mean.