I need to get an constant 3ma AC current through the Load of 7k to 18k range( used for impedance cardiology ) . I've tried using the standard Howland Current Pump circuit as given below with I/P voltage of 1.8Vpp or 0.636 Vrms , 68kHz , and the O/P current is only < 1.1 mA with voltage across load 0.5V .

with R1=R2=R3=R4 = 220ohm

Please say how to correct this error , or any so .
Thanks in advance , attached the multi-sim copy of circuit .

 

http://www.ti.com/lit/an/snoa474a/snoa474a.pdf

but be aware that the max voltage, current and impedance are related.

Picbuster

 

3mA through 7kΩ requires 21V and through 18kΩ requires 54V.
How do you expect to generate that much voltage?

 

3mA through 7kΩ requires 21V and through 18kΩ requires 54V.
How do you expect to generate that much voltage?

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Then atleast for 2k load the desired Voltage should be made , but its not the case .

The Op-amp O/P Voltage swing for >= 2k load is 14V , and according to my circuit the V across load is Vout / 2 , i.e the swing is about 7V .

Now , for 2k load and 3mA AC current only 6V drop is made out .

 

You op amp appears to be slew-rate limiting at that frequency.
You need one with a higher slew-rate rating.

Also note that half the output current is going through the other set of 220Ω resistors.

 

Here's the LTspice simulation of a modified Howland circuit with a high speed op amp that may work better for you.
It adds two resistors, R5 & R6, to allow larger feedback resistor values without limiting the current to the load.

Note that the output current stays constant at about 3.3mA peak while the output voltage increases due to the simulated increase in Rld load resistance with time.
This continues until the resistance exceeds about 4.2kΩ at 170μs, at which point the op amp has reached its output voltage limit and starts to clip.

 

Here's the LTspice simulation of a modified Howland circuit with a high speed op amp that may work better for you.
It adds two resistors, R5 & R6, to allow larger feedback resistor values without limiting the current to the load.

Note that the output current stays constant at about 3.3mA peak while the output voltage increases due to the simulated increase in Rld load resistance with time.
This continues until the resistance exceeds about 4.2kΩ at 170μs, at which point the op amp has reached its output voltage limit and starts to clip.

View attachment 128520

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------


It works fine ! Thanks for your time and effort

 

Here's the LTspice simulation of a modified Howland circuit with a high speed op amp that may work better for you.
It adds two resistors, R5 & R6, to allow larger feedback resistor values without limiting the current to the load.

Note that the output current stays constant at about 3.3mA peak while the output voltage increases due to the simulated increase in Rld load resistance with time.
This continues until the resistance exceeds about 4.2kΩ at 170μs, at which point the op amp has reached its output voltage limit and starts to clip.

View attachment 128520

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Our new circuit and its calculations made on it , gives us

- the 3mV Vp is made successfully till 4K load and as load increases the voltage drop increases beyond the voltage swing limit , so current clipping is observed beyond 4k .

If I need to drive the current through 8k load , do I need to use op-amp with high o/p voltage swing , any choices . Thanks !

 

One I found is the LTC6090 which is rated for 140V operation but, for some reason, the simulated slew-rate at 68kHz is only about 4V/μs although it's rated for 21V/μs.
Now sure why that is, except perhaps there's an error in the Spice model.

You need ≥14V/μs slew-rate to generate a 68kHz, 3mArms sinewave current into 8kΩ.

Edit: Apex Microtechnology makes some high voltage op amps but they are not cheap (what's your budget?).

 

Here's the LTspice simulation of the circuit for up to an 18kΩ load @ 3mArms using an Apex PA340 high voltage op amp (about U$15 @ Digikey).
The gain is set to ≈1mA/V
The amp starts to clip at about a 17kΩ load @ 3mArms.