From: Ender Kuntsal [EKuntsal@itc-transducers.com] Sent: Friday, September 25, 2009 8:56 AM To: Massion, Gene Cc: Bill Bunker Subject: Re: ITC8201 Questions Gene, Pls. see my answers under your questions. Regards, Ender ======================== Ender Kuntsal, Engr. Mngr. ITC, 869 Ward Drive, Santa Barbara CA 93111 USA Tel: (805) 683 2575 Ext. 514, Fax: (805) 967 8199 EKuntsal-itc@channeltech.com, www.itc-transducers.com >>> "Massion, Gene" 9/24/2009 6:41 PM >>> Ender Thanks for trying to reach me. Let’s see if I can ask my questions in an intelligible way over e-mail. A little background on our project may help and may actually be of interest to you. We have several ITC8201 hydrophones S/N 207, 208 and 209 that we are using for a somewhat novel application. We are trying to listen for sediment transport events in the Monterey Canyon. We can’t put instrumentation (current meters, turbidity sensors, etc) in the canyon as it gets wiped out by the sediment transport events we are listening for. So our current approach is to put a hydrophone and data logger on a shelf on the side of the canyon that seem protected from these events. We a very simple, cheap device in the canyon that releases a float and phones the time of the event in to us when an event occurs. Now we can look in our acoustic record around the time of the event to see if we see a unique acoustic signature that we can say is a sediment transport event. We aren’t looking for a signature in the sense of a submarine acoustic signature. We are just looking for a build up in noise (probably low frequency) and a long, loud event; or a sequence of short loud events or ???. As a first guess we are looking from 10 Hz to 20 kHz. In any event, we don’t need a really low noise system, we think these events are loud and we are only about 150 meters away from the centerline of the canyon. It will even be OK if a sediment transport events saturates our digitizer. COMMENT: Pls. note that this unit has a maximum of 2 Vrms output w/o clipping; if the SPL on the unit causes this output exceed this value then the preamp will saturate (maybe before digitizer) and start clipping (but will come back to normal when the sound is gone)... Q1) The phones were calibrated in your facility on May 18, 1999. Is it correct to assume that the Receive Sensitivity vs Frequency curves that came with the phones include the preamplifier gain? For example, S/N 209 has a receive sensitivity from the curve of -158 dB/V/uPa. So if I hit the phone with a 158 dB re 1uPa sound pressure level, will I see 0 dBV out of the preamplifier? ANSWER: The measurements include the preamp gain which is about 23 dB (pls. note that this is NOMINAL and it is correct when the load is infinite; i.e., unterminated). The NOM -3 dB roll-off frequency is near 10 Hz so, the receive response (OCV) is going to be about 3 dB less than otherwise "flat" LF response. Using -158 dB OCV, a 158 dB SPL will generate a 1 Vrms output, or as you stated 0 dBV. Q2) Bill Bunker sent me a copy of a memo (attached) you wrote to Bert Neales May 19, 1999 describing the 8201 output signal DC offset. We see the same .6-.7 VDC offset which creates a problem given our digitizer. We are following the 8201 preamplifier with a simple instrumentation amplifier (an Analog Devices AD8220 JFET input device). This amp can’t be completely AC coupled as it needs a DC return path on the inputs to provide a bias current. Analog Devices recommends a RC high pass filter on the front end with the R value chosen to provide the return path and minimize offset, noise and so on. I’d like to set the low end of my bandwidth about 10 Hz which means I only need a .1 uF cap and a 150 kohm resistor. Your memo talks about much larger caps: 15 uF to 220 uF. I don’t know as much as I should about this so I’d like to ask if you see a problem with this approach for our application. ANSWER: I reviewed that e.mail again (10 years old!) and I see the suggested cap values are only suggestions and one can use different values. As far as I know the suggested option worked for the 1999 application but an RC network should also work. Q3) the ITC document 168-0047.htm date 09/10/1999 says the “Cal Pad” is 20 dB. I’m not sure how to interpret this and I’d like to make sure I don’t damage anything before I put a signal in the cal input. If I put a 100mVp-p (or some other suitable level) into the Cal input, what level signal should I expect at the preamplifier outputs? ANSWER: You can input even 1 Vrms into CAL circuit and it will be OK. The intent of the CAL pad is to make sure the preamp is functional and provides the intended gain. We need to consider the front end (between the ceramic and preamp) cable loss as well in these calculations. Assuming the ceramic is about 60 pF and cable is 15 pF, then the cable loss is: 20log(60/75)=-1.9 dB.At 1 kHz if the preamp gain is 23 dB, and if the CAL PAD is 20 dB, then with a certain input voltage, the preamplifier output will be: OUTPUT(dB)=INPUT(dB)-1.9-20+23=INPUT(dB)-1.1 dB. Please keep in mind that the ceramic and cable capacitance, along with actual preamp gain and cal-pad resistor values vary from unit to unit and if you are getting a input-output difference within a couple of dB it shows that everything is OK.