eCircuit  Center


About SPICE | SPICE Basics | Running SPICE | CIRCUIT COLLECTION | SPICE Commands | SPICE Demos and Downloads
About Us | Contact Us | Home



I-to-V Amplifier - Stability



                             OPITOV.CIR                Download the SPICE file

If you said that the easy application of a trans-impedance amp is too good to be true - you'd be right! Optimizing current-to-voltage amplifiers can be one of the most challenging aspects of op amp design. Why? The main culprit is the sensor's own capacitance CS.

Capacitance CS brings two ill effects into your circuit.

1. It effectively places a short circuit across IS at high frequencies, limiting the circuit's bandwidth.

2. CS tends to make circuit unstable. How so? Notice as the signal runs along the feedback path from node 2 to node 1, components RF and CS create a low-pass filter. And filters of the low-pass kind contribute negative phase to the feedback loop, making the circuit less stable. (See Op Amp Feedback Analysis.) Instability, rears its ugly head in your circuit as overshoot, ringing or even oscillations.

 CIRCUIT INSIGHT   Insert CS=100pF across IS by removing the * at the beginning of the CS statement (also make values RS=500k and RF=100k). Run a simulation and plot the Transient Response of the output V(2). What happened to the output voltage? How much overshoot and ringing does CS cause?

Now, let's see how CS puts a bump in an otherwise flat frequency response. Plot the AC Response of the output VM(2). How big is the peak in the response versus frequency?



Restoring stability requires counteracting the undesirable low-pass filter RF and CS in the feedback path from output to input. You can do this with a high-pass filter in the same feedback path. Just add capacitor CF. High-pass filter, CF and RS, adds positive phase to the loop pushing the circuit toward stability. But, as you know, circuit design is an arena of compromise. And typically, the first casualty of stability is bandwidth.

 HANDS-ON DESIGN   Place CF=2PF across RF by removing the * at the beginning of the CF statement. Run a simulation. Check out the Transient Response of output V(2). How much of the overshoot and ringing have been reduced? Increase CF until the overshoot and ringing are gone. But as you do this, notice what's happening to the rise time.

Take a look at the AC Response VM(2). Yes, the peak is being reduced, but, so is the overall bandwidth! That's because CF and RF create a low-pass filter in the forward signal gain path from IS to Vo.

Tuning your circuit for optimum response can be a two step process:

1. Increase CF until the overshoot and ringing are acceptable for your application.

2. Check the rise time. If it meets your requirements, you're done. If its too slow, pick a faster op amp and repeat steps 1 and 2.



Download the file or copy this netlist into a text file with the *.cir extention.

IS	0	1	AC 1 PWL(0US 0UA  1US 10UA  20US 10UA  21US 0  40US 0)
*RS	1	0	500K
*CS	1	0	100PF
RF	1	2	100K
*CF	1	2	2PF
XOP1	0 1	2	OPAMP1
* connections:      non-inverting input
*                   |   inverting input
*                   |   |   output
*                   |   |   |
.SUBCKT OPAMP1      1   2   6
RIN	1	2	10MEG
* gain bandwidth product = DCGAIN x POLE1 = 10MHZ
EGAIN	3 0	1 2	100K
R1	3	4	1K
C1	4	0	1.5915UF
EBUFFER	5 0	4 0	1
ROUT	5	6	10
.TRAN 	0.2US  	40US
.AC 	DEC 	10 100 10MEG



2002-2010 eCircuit Center