Originally Posted by JD-EE
What are the noise power characteristics vs frequency for 100Hz through 100kHz. (I understand 100kHz is basically down to thermal noise levels.) And I'd LOVE to know some transducer sensitivity characteristics. Alas, I don't know where to look for this data.
I assume you are asking about the ambient noise of the ocean in the 100Hz-100kHz range. I don't have a copy handy of "Principles of Underwater Sound", Robert J. Urick. Over that wide a range the answer is complicated, because different effects and sources dominate the noise in different bands. Generally the noise decreases from low frequencies to high, and then above 100-200kHz, it starts to rise again as the thermal noise of the water becomes dominant. I have a chart at work, but I have not memorized it.
Of course, transducer sensitivity is important in system design, but in the range that I have worked 10Hz-30kHz, it is possible to design transducer/pre-amp combinations that have less noise than the ocean ambient. Thus the ocean noise is measurable and the transducer parameters are not critical for determining detectability. I could find some numbers when I'm at work, or you could look at data sheets from vendors like Benthos, Teledyne, EDO (if they're still in that business), ITC, Massa (I think I'm showing my age, so I'll stop...).
I'm sure the various contributors are aware, but I want to clarify that we are mixing two separate discussions: how to best detect existing pingers, and how to design a replacement pinger that is more detectable.
With regard to the latter, and without addressing every point that has been raised in the past couple of days, I offer that all REMUS vehicles operate with digitally coded transponders (transponders ping in response to interrogation). This greatly improves the ability to operate in a multipath environment, because the chip time is much shorter than the transmission length, which is still on the order of the traditional 10ms. I'm not sure that we get much improvement in detection range.
The receivers operate continuously, running 4 simultaneous auto-correlations (to detect any of 4 different codes) for an expense of about 50mW, including the preamp. That's about 3 alkaline D-cells for 30 days. If that is changed to lithium primary, the size probably becomes similar to what's in the existing pingers, though of course it would need still more energy so that it can transmit as well as receive.