| Electronics > Beginners |
| Piezoelectric hydrophone low noise amplifier design questions |
| << < (4/10) > >> |
| David Hess:
Exactly how large is the piezoelectric microphone? Specifically, how much capacitance does it have and what will be the length of cable between the microphone and input amplifier? It matters because if operation down to low frequencies is not a requirement, then the capacitance at high frequencies determines the input impedance and the optimum ratio between voltage and current noise. Conceivably a bipolar input operational amplifier with much lower voltage noise is a better choice if the frequency is high and the capacitance is large. The required bias current resistor, which is needed anyway unless a transformer is not used, can be chosen based on the low frequency cutoff and capacitance. Some hydrophone amplifier designs include a second DC servo loop to cancel the input bias current while providing a higher input resistance for extended the low frequency response but that does not apply here. |
| Sparker:
Exact data on the piezoelectric antenna is unknown yet. I'd love to know and give more data, but I can't. I can only say that it will be a one-dimensional array or piezoelectric elements(to have high directivity in orthogonal direction). Something like this one: The part marked as "2" is the antenna, it has a row of elements connected in parallel and it's filled with epoxy. The cable length will be about a meter. --- Quote ---Why use a digital potentiometer instead of a PGA (PGA112 for instance). --- End quote --- Thanks, but I think i will stay with the digital pot. variant since it gives more gain values. And I can lower the final stage gain later if I encounter problems. AD812 might be expensive but it has two opamps in one case. There aren't a lot of opamps like these I can choose from without waiting for them for two weeks to be delivered. |
| David Hess:
I found some datasheets and the capacitance is 1000s to 10,000s of picofarads like I expected. So at 240kHz, that is 664 ohms and 66.4 ohms of reactance so a low noise bipolar input operational amplifier *is* suitable because the current noise of such a part into such a low reactance will produce an equal or lower voltage noise than the input voltage noise of the amplifier itself. (1) Of course the details will depend on exactly the capacitance of the piezoelectric hydrophone. At lower frequencies and with smaller transducers which have lower capacitance, the current noise needs to be lower because the transducer's reactance is higher giving JFET or MOSFET input amplifiers the advantage and the same thing happens if an input transformer is used to raise the transducer's impedance. So with the current situation assuming a 1000pF transducer, I think you would have about a 6 to 12 dB noise advantage with suitable bipolar input operational amplifier at 240kHz because the input capacitance is so high but with a 4:1 or higher ratio transformer, a JFET or MOSFET input amplifier will be necessary because bipolar parts with a suitably low current noise will not be fast enough. With a 10,000pF transducer, a transformer could even be used with a bipolar input part. (1) You still have to deal with returning the input bias current to ground, whether a bipolar, JFET, or MOSFET input is used, but that just means shunting it with an inductance or resistance. A shunt resistor forms the first pole of your high pass response attenuating low frequencies. |
| Sparker:
Could you share the datasheets, please? I'd like to see the real-imaginary impedance plots of real assemblies. Thanks a lot for such in-depth analysis of my problem. :-+ I have learned a lot from your messages! I might be wrong again, but I expected that the antenna will work at its resonant frequency, and its impedance should be active at this frequency. |
| David Hess:
This is the datasheet I found: http://www.lthe.fr/LTHE/IMG/pdf/DocBruelKjaerHydro.pdf I only used it to make a rough estimate and I suspect your device being an array will have a capacitance at the higher end of what I estimated. The short summary of what I posted is that you cannot go wrong with a low noise JFET or MOSFET operational amplifier but I think under ideal conditions at least another 6 to 12 dB is possible with a bipolar part and even more if a transformer is used because the source impedance is so low. |
| Navigation |
| Message Index |
| Next page |
| Previous page |