AMPLIFIER WITH SINGLE POWER SUPPLY

Bisni,

Please see the attached zip file containing a power point and TINA files.  It describes two different amplifier topologies that should work for your application.  The basic design method is also given.

Best regards,

Art

 ac comp-inv-and-non.zip

 

I tried your design and I got the output successfully.I need more clarification.The frequency of my signal is very low,approximately
.318 hertz.So can I use the same design for amplifying the signal by changing the capacitor value C1 n the design?If yes,What will be the value of the capacitor you suggest?


Thanks & Regards,
BISNI

Bisni,
Your issue has nothing to do with the OPA350. In order to effectively amplify such a low frequency signal you should use an extraordinary RC time constant. On the contrary, your capacitor and resistor values do not follow that requirement. For instance (I assume that you use the "ac coupled-inverting" example that Art provided), with C1=200uF and R2=200 Ohm, the gain drops to g=8 at 0.32Hz. This figure is quite lower than your target (g=100). So you have actually built a high pass filter with a corner frequency fc = 4Hz. This explains why you get so low an output signal (please also account for the capacitor tolerance) at 0.32Hz and accurate gain at much higher frequencies.

Your options with the "ac coupled-inverting" (which is preferrable to the non-inverting alternative, since it uses only one large capacitance) are the following two:
a) further increase the capacitance (in the order of 20000uF, not a practical value it seems)
b) increase R2. E.g., you may use R2=20K, R1=2M, C1= 220u. In that case, you must also increase R3, R4 to about 40k in order to minimize the effect of the input bias current. Needless to say, the resulting circuit will have an extremely long startup time as Art has already noted.

I suggest you to carefully read the slideshow provided by Art Kay. After understanding the limitations of the ac coupled amplifiers, you may experiment with capacitor and resistor values in TINA, in order to obtain the desirable gain. Please always remember that the theoretical gain at 0.32Hz must be close to 100 (otherwise the gain will be strongly depended on the tolerance, temperature coefficient and the non-linear voltage-capacitance characteristic of the capacitor).

Art has attached two alternatives:
a) the "ac coupled.tsc" circuit
b) the "ac coupled-inverting.tsc" circuit

I guess you have built the circuit a. This circuit will work too, however it has the drawback of using two RC time constants (i.e. two capacitors and the associated resistors). One RC is located at the non inverting input (R3//R4 x Cin), the other RC (R2 x C2)is located at the feedback network. Both of these RC time constants affect low frequency gain, thus you need to use a large R x C value for *both* of them.

If you insist on using this circuit, suggested values that would work for you might be, for instance: R1=1M (990K if gain must be accurately set to 100), R2=10K, C2= 470u, R3=R4=2M, Cin=10u. However, these large resistors may introduce leakage, offset current and noise issues. This (along with the slow startup) is the price to pay for such a very low corner frequency.

Whicever of the two circuits you finally choos, I would *strongly* suggest you to simulate the circuit in TINA (the zip contains both versions, you wave only to change component values and adjust the simulated frequency range to start at, say, 0.32Hz). This will help you to understand the operating principle. Later on, you may also look for better alternatives like a two-stage approach or, even better, a dc-coupled amplifier.

Good luck!