0
\$\begingroup\$

I am currently designing a amplification signal chain (with 3 op-amps) for a signal that comes from a hydrophone with around 1nV amplitude and with a bandwidth of 10 to 100 kHz. For that I am using one op-amp as a buffer and the two after together to a 40 x 30 gain. However, when I run the simulation I get the 40 times gain on the first op-amps that amplifies but the second one de-amplifies the signal. To troubleshoot I connected (to simulate the output of the 40 times gain op-amp) a 1nV x 40 voltage source ac analysis from 10Hz to 100 kHz to the 30 times op-amp to simulate the output of the 40 times gain op-amp and I obtained the 30 times gain at the output of the 30 times gain op-amp. This made me realise that the problem has something to do with the interface between the two amplifiers with gain 40 and 30, but I can't tell or even imagine what would be the problem. I am going to put the circuit schematic so that you guys can have a visual representation.

enter image description here

\$\endgroup\$
2
  • 4
    \$\begingroup\$ In a low noise application, there's never any point to have the input device as a unity gain buffer, if you can configure it to have gain, which will effectively eliminate the noise contribution of all succeeding stages. Post the simulation .asc file as code (ctrl-K), so we can run your simulation exactly as you have it, without the time and potential differences of us entering it from the schematic. \$\endgroup\$ Commented Jul 23 at 15:04
  • 1
    \$\begingroup\$ The AD8421 is an instrumentation amplifier, not an op amp, and I don't see the in amps' REF pins connected to anything. They should be connected to your signal ground. \$\endgroup\$ Commented Jul 23 at 15:06

2 Answers 2

Reset to default
3
\$\begingroup\$

You have a number of issues.

  • The reference (ref) pin of the instrumentation amplifiers needs to be tied to a low impedance source. For your circuit, ground the reference pins.
  • You are DC coupling the amplifiers. With high gain stages this will create a DC offset that could push the output to the rails. Capacitively coupling the stages will reduce such issues. Be sure to place bias resistors from the inputs to ground.
  • Why are you placing a 1 megohm resistor in series with the input? This will increase the noise due to the thermal (Johnson) noise of the resistor. The noise of the resistor is approximately \$ 0.13 \sqrt{R} \;\;in \;nV/\sqrt{Hz} \$, or about 130 \$nV/\sqrt{Hz}\$.
  • To get the best noise response from the first opamp, you need gain, anywhere from 10 to 20 dB.
  • Trying to amplify a 1 nV signal is not realistic, especially in the stated bandwidth of about 100 kHz. The noise of the amplifier and hydrophone will be \$ \sqrt{100kHz} \$ times the input noise of the amplifier, or about 316 x 4 nV = 1200 nV of noise within the bandwidth (ignoring the noise from the input series resistor). The 4 nV is the amplifier noise which will be added to the thermal and environmental noise from the hydrophone. This will mask the 1 nV signal. If you are using an FFT to analyze the data, you'll need a FFT with lots of bins to resolve sub-1 Hz.
  • You should have input protection if you are connecting to a hydrophone. A hydrophone element can magically charge up to many hundreds of volts just sitting around due to thermal stresses. This may be enough to damage the input to the amplifier when the hydrophone is connected to the amplifier. Back-biased diodes can be used to make an input protection circuit. Also, hard thumps to the hydrophone can generate voltage spikes. Back-biased diodes will have an impedance around 1 to 5 megohms which may cause issues if the hydrophone simple capacitance (capacitance at 10x below resonance) is low (under 15 nF).
\$\endgroup\$
1
  • \$\begingroup\$ Thanks, it was the dc coupling that caused the problem. Well i should have mentioned that i wanted around 0.8nV around 100kHz but at the lower frequencies the input would be bigger around 1.6microV. With this implementation i get around 10 microVolts/sqrt(Hz) of output noise at 100khz with 1200. this means if i can amplify the signal at least a bit more than 10microVolts/sqrt i can have a SNR bigger than 1 at 100k.Do you think there is a alternative just like this one where i can amplify the signal to be bigger than the noise or should i try something different? Maybe using chopper amplifiers? \$\endgroup\$ Commented Jul 25 at 22:24
2
\$\begingroup\$

a signal that comes from a hydrophone with around 1nV amplitude and with a bandwidth of 10 to 100 kHz

Considering that the LTC6268 buffer op-amp has an input referred noise of 4.3 nV/\$\sqrt{Hz}\$ and your signal has a bandwidth of up to 100 kHz, the effective RMS noise seen at the input (that your signal has to overcome) is nearly 1.4 μV RMS. This is massively larger than your input amplitude (stated to be 1 nV) so, how can you expect this design to ever work?

Apart from that, the middle stage instrumentation amplifier has a lower value equivalent input noise compared to the front-end op-amp so, the op-amp buffer is pointless (even if the noise is still far too high to make a reasonable amplifier for your application).

when I run the simulation I get the 40 times gain on the first op-amps that amplifies but the second one de-amplifies the signal

In addition, the reference inputs on your instrumentation amplifiers need to be tied to 0 volts. I expect that if you looked at the DC output from the middle gain stage it will be swamping the final stage thus preventing any reasonable amplification.

\$\endgroup\$
5
  • \$\begingroup\$ As noted, your amp's noise is way above the signal level. For a 1nV signal in a 100kHz bandwidth, you would need an amp with less than 3pV/rt-Hz noise, which would require some type of cryogenically cooled amplifier. \$\endgroup\$ Commented Jul 23 at 16:39
  • \$\begingroup\$ @CarlRutschow it's not my (your) amp so, who are you directing the comment towards? And, who is "you"? \$\endgroup\$ Commented Jul 23 at 20:05
  • 1
    \$\begingroup\$ @CarlRutschow - Hi, As commented above by Andy aka, who are you wanting to address in your comment? By commenting on this answer, you're implicitly addressing its author, since no-one else will be automatically notified of your comment. So if you intended to address the OP, for example, then commenting here is the wrong place to do that. || If you tried to use the @username feature, I could explain why that wouldn't work then. However in a previous similar situation (June 14), my offer of help wasn't accepted :( If that decision not to accept help changes, please let me know. TY \$\endgroup\$ Commented Jul 23 at 20:46
  • \$\begingroup\$ My apology. I answered here because @Andy aka made a comment about the amp's noise, but obviously I used the word "your" and "you" incorrectly. Still getting used to the arcane rules here. \$\endgroup\$ Commented Jul 23 at 20:54
  • \$\begingroup\$ Proper use of words like you or used can hardly be described as arcane. \$\endgroup\$ Commented Jul 30 at 17:31

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.