Hydrophone amplifier and sampling

[MarkT]

The 5.1 nV/sqrt(Hz) voltage noise is acceptable for most LF application. Obviously, the good old time with sub-nV/sqrt(Hz) FETs are gone.

Well there's still:
https://audioxpress.com/news/linear-...n-channel-jfet

[WMXZ]

Well there's still:
https://audioxpress.com/news/linear-...n-channel-jfet

And it seems that https://shop.micross.com/ sells these good old stuff online (e.g. DPAD protection diodes)

Worthwhile to see http://www.linearsystems.com/lsdata/...9_App_Note.pdf

Anyhow, I'm happy to see that Linear reintroduced these components again a year ago.

[KOC62]

The OP has an interesting project going that I'd like to learn from.


In an AudioXpress (April 2007) edition, Dennis Colin had a project that used a couple of LSK389 in parallel whose output was fed to an AD797AN opamp. He claims around a 1nV/sqrt(hz) referred input RMS noise performance. Later (Sept. 2007) he published an RIAA preamp using a similar (very low noise) topology.
Another diy'er claimed that Colin's design could be improved by using a very low noise DC power supply as the circuit hasn't got super high power supply noise rejection. I'm not sure if these 'tweaks' are noticeable in actual use, as some of these ideas end up in specmanship debates.


Peter.

[julianblanco]

Im really loving the discussion going on. Thanks so much all! Im learning a ton. Going to be building it on the bench this afternoon, will post some pictures and results soon.

[WMXZ]

In an AudioXpress (April 2007) edition, Dennis Colin had a project that used a couple of LSK389 in parallel whose output was fed to an AD797AN opamp.

For the record
https://audioxpress.com/assets/uploa.../colin2993.pdf

[lightningmar]

First version I built was some air wired hot garbage. This is the second that I just threw on a breadboard. Signal is weak and i think I am picking up a lot of noise from my cheap bi-polar voltage reuglator. Need to replace it and put a low pass filter before for the HF noise I seem to be picking up (maybe the supply maybe something else). Going to try to make more progress on it tomorrow hopefully.

[WMXZ]

IMHO, you will never get a realistic performance picture without a solid soldered setup (best is pcb or veroboard)
Then measurements should be made in a tin-box to shield electric fields that couple into wires.
A fully differential setup will also help against external interference.
Obviously, Only 1 power (either PC or external (e.g. battery)) is beneficial to reduce risk of 50 HZ ground loop.