optimized amplification circuit for ultrasonic sensors

Hello Ajit,

Sounds like a neat project.  I worked on a robotics project in college where we used an array of four hydrophones to detect a sonar beacon and dock with it.  We had to very accurately detect when the sonar signal reached each hydrophone so we could use time-of-arrival math to determine the trajectory of the beacon based on our last reading, so it was very similar to what you're trying to do.   

That said, there are several posts on the forum already devoted to hydrophone, geophone, and microphone applications which are all very similar.  Please consider searching the forum for some other posts like the one below for ideas:

http://e2e.ti.com/support/amplifiers/precision_amplifiers/f/14/p/166388/607712.aspx#607712

http://e2e.ti.com/support/amplifiers/precision_amplifiers/f/14/t/161153.aspx

Regarding the filter-pro results, I'm not sure what you designed but to require a 160MHz amplifier you must have a pretty complicated or advanced design.  The circuit will not properly function if you do not use an amplifier with appropriate GBW, so the 20MHz amplifier will not work for your current circuit.  I would suggest increasing the number of stages, or the order of filter, or gain in the filter to lower the GBW requirements.

It is always desirable to front-load the gain in a sensor system because this will maximize the SNR of the system.  However if your gain requirements are too great and result in excess noise that swamps out the sensor signal then you may have to spread the gain across a few stages.

For a high-speed comparator, check out the TLV3501.

Best Regards,
Collin Wells
Precision Linear Applications

Hello everyone

   Thank you Collin for your prompt reply. For above application I am trying to evaluate some alternatives. Overall objective of analysis is to reduce the variability in the time measured. At present it is about 100 ns. For my application it should 20-40ns or less.  Basic signal processing chain is:

Filter Specification: 40KHz , pass band width 1KHz, Cheby 1, MFB

1)I would like to know whether i should distribute the gain between filter and next amplification stage?

2)What should be the order of the filter? There aren't any major noise source in the environment where the application would be used. 50Hz off course would be there when I am experimenting in lab but smt component based board with short traces i hope will reduce it. So noise added due to different component would be a major consideration. 

2)The GBW that i was talking about is of the op-amp used in the active 40KHz band pass filter. As per Filterpro it should be 160 MHz. So i would like to know whether OPA320 would suffice? Is there any better alternative?

3)How appropriate it would be to use a Instrumentation Amp. Instead of OPA320?

4)A comparator output would be used to stop a digital counter in FPGA which is counting the time elapsed. Will it be better not to use the comparator at all? as There is also a 0 to 1 detector circuit at the FPGA i/o pin.

expecting some insight and use full suggestions.

regards

Ajit Singh

Hi Ajit,

Unfortunately the image of your signal processing chain did not come through.  Would you please try inserting it as a file instead of an image?  I have a few other items I need to close out but will be working this issue after and will reply back as soon as I can.

Regards,
Collin Wells
Precision Linear Applications

hello Colin

       Please find attached updated signal processing chain. GBW of OPAMP is affecting filter response. So Instead of OPA320(GBW - 20 MHz) i have used OPA37(GBW-63 MHz). Instead of TLV2352 Comparator i will be using TLV3501, but it wasn't there in multisim library. I would like to reduce the no of stages.  So for amplification if i use an Instrumentation amplifier will it going to be better? Also if I directly connect amplifier output to FPGA I/O pin, what problem may arise.  

regards

Ajit