Electronics > Beginners

Piezoelectric hydrophone low noise amplifier design questions

<< < (3/10) > >>

David Hess:

--- Quote from: Sparker on April 22, 2018, 03:30:56 am ---Now I think I will stick with transformer input design, to adjust its winding for noise if needed, as you suggested. Also, if I connect the input transformer between the + input of the CFB Opamp and the ground, does anything prevent me from also using the CFB opamp for the first stage?
--- End quote ---

Input noise still counts and I think voltage feedback operational amplifiers have an advantage here but I have not studied the current feedback operational amplifier option in detail.  To put it another way, nobody that I know of is using current feedback operational amplifiers (or transconductance amplifiers, see below) for low noise input stages.

Giving it a little bit of thought, this makes perfect sense.  Current feedback amplifiers are closely related to transconductance amplifiers (1) where the resistance in series with the emitter sets the transconductance which is effectively the gain.  That same resistance however in series with the emitter also adds directly to the input noise; as far as the transistor is concerned, noise in series with the emitter is the same as noise in series with the base and since the emitter operates at much higher current, a small amount of resistance there has a large effect on noise.

Fast voltage feedback operational amplifiers, often called video amplifiers, also usually include a resistor in series with the emitters of the input stage to lower transconductance which counter intuitively, allows for higher speed.  (2) But this *also* makes them noisy.  The very old example of this is the LM318 but modern video amplifiers do the same thing and you can often find this used in audio amplifiers where the increased gain-bandwidth product yields lower distortion, but again at the expense of higher input noise.

So no, I do not recommend using a current feedback operational amplifier or operational transconductance amplifier for the low noise input stage.

Low noise hydrophone amplifiers usually use low noise FET input voltage feedback amplifiers (and no input transformer) however this is because the low input bias current of an FET extends low frequency operation which you do not need.  But a low noise FET input voltage feedback operational amplifier would still be a good choice at least to start off with.

(1) Not necessarily operational transconductance amplifiers which have an input stage identical to a voltage feedback operational amplifier but a quick inspection shows they are pretty noisy also although I do not understand why.  Some old voltage feedback operational amplifier provide access to their transconductance output through their external compensation pin so they obviously have the same noise specifications when used either way.

(2) This is called transconductance reduction or Gm reduction.  Low noise amplifiers which need transconductance reduction do it in a different way which does not require adding resistance in series with the emitters.

Marco:
Why not use a transformer? It's probably be going to be an expensive one, since you want really high permeability to keep the number of windings on the secondary low ... but it does allow you to drop noise a bit and do common mode rejection without needing a differential amplifier.

Sparker:
Sorry for not answering for a few days!

You guys are right about the noise, I obviously forgot about the input current noise influence when I made my assumption of using the CFB op-amp as input stage :palm:
Inspecting this spec for a voltage feedback CMOS device and current feedback BJT device made it clear for me:
CMOS device: AD8651, Current noise density: 4 fA/sqrt(Hz)
BJT device: AD812, Input current noise, + input: 1.5 pA/sqrt(Hz), or bigger by about 3 orders!

Now, by using the AD8651 CMOS input opamp at input and two more stages of AD812 CFB opamps, with a Sallen-Key filter around the second stage, I was able to assemble this circuit in LTSpice(please see attachment). Input transformer is omitted in the model but I will install it in the real circuit.
I was going to add digitally controllable gain, and for that I chose MCP41010 10kOhm digital potentiometer, which is emulated in the circuit by R11. It is the only one I could find with low parasitic capacities (C_A, C_B and C_wiper in the circuit). This digital potentiometer can only tolerate voltages between its GND and Vcc, so I decided I will power it with the analog rails, since they have 5 Volts between them, and shift the logic levels down from the MCU levels (0...3.3V) to (-2.5V...2.5V) with a little BJT circuit(omitted in this model).

I'd be grateful if you have a glance at the schematic, maybe you can see some obvious faults which I don't see.  :)

David Hess:
Do not forget that without an input transformer, the input amplifier requires a ground return for its input bias current because the transducer is open at DC.

Marco:
OPA140 has 5x lower current noise than the AD8651, but that doesn't really matter at the 10-100 Ohm impedance range the transducer probably falls in near resonance ... at least not without a transformer with gain (with 10+ voltage gain with a transformer it might start mattering). AD812 looks a bit expensive, unless you need the output current capability. Why use a digital potentiometer instead of a PGA (PGA112 for instance). That way you don't need that 400x amplification on that last stage either, that's pushing it a bit.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod