problems tesing howland current source.

[zoulzubazz]

hey guys have this dual op-amp howland current source i designed for my purpose using an xr2206 sugnals generating ic and a dual op-amp. the figure shows the design. the circuit gets its input (Csin in the figure) from the xr2206 and the output is being tapped from pin 3 of the op-amp (VCCS in figure).

**broken link removed**

now to test this circuit i am attaching a shunt resistor in series with a 10kOhm trim pot which is grounded on the other end and observing the voltage across the 100ohm shunt using an oscilloscope also shown (rather crudely) in the figure. ideally the voltage across the shunt should remain constant when the load resistance changes since this is a constant current circuit; which is the case, but when the load is open circuited (essentially opening the current path) the oscilloscope continues shows the same voltage across the shunt? now when the ground-clip of the scope is connected to circuit ground and then the load is changed the voltage across the shunt is observed to be swinging with the change in the trim pot. i dont quiet understand what is going on and would like some help regarding this. thanks very much.

 

[FvM]

1. No cross posts, please.
2. The circuit doesn't work as a current source with C3 in place
3. Exchange the position of current sense shunt and 10k pot, because the OP input is grounded.

 

[crutschow]

Your attachment gives me an error message saying it is an invalid attachment.

 

[zoulzubazz]

thanks guys, apologies FvM cross posting was not my intention, honestly was assuming the post in electronics elementary questions was deleted. i am attaching the picture gain for others to have a look. i am trying to change the circuit like suggested FvM. could you shed some light on why C3 would cause the circuit from behaving as an current source? always thought a cap in the feedback was good and it is explained to add stability to circuit in the following link:

**broken link removed**




thanks very much for your input.

 

[FvM]

A capacitor may be necessary for stability, the value matters however. According to the capacitor value, the TI design works as current source up to a 100 kHz range and yours not even for 1 Hz.

 

[zoulzubazz]

thanks for that FvM, i was using a 100nF cap as C3 unlike what is shown in the figure and found like you said that this made the output current to change linearly with the load, after having switched to 10nF now the output is stable with changing load. what value for C3 would you recommend if the chosen frequency of operation was 400kHz? thanks very much for your valued input.

 

[FvM]

Even with a compensation capacitor as small as a few pF, the output impedance at 400 kHz can't be larger than several kOhm with a 4 MHz GBW OP.