RECEIVERS
It is beastly difficult to construct a low-noise,
high-dynamic-range receiver. Under any circumstances. It is
even more difficult when you need (or think you need) to
place the receiver circuitry on the same card as a 125W
transmitter, with oscillators and whatnot running spikes
through your groundplane. Well, I'm not sure that I've come
very close to accomplishing this but I have come up with
some designs that work pretty well.
Historically, the best receiver I ever had was built by
Walt Dillon at Oregon State University. It was a fairly
wide-band (1 kHz - 1 MHz) superhet receiver with linear TVG
(time-varied-gain) and a linear amplitude detector. He
built this as a rack-mounted system we used on shipboard to
process signals from discrete transducer arrays we dangled
over the side of the ship. The last I heard of this system,
it had been sold as scrap to a rare-metals dealer in
Oregon. Sigh ... One of these days, I am going to scan the
schematics for this system, however, and put them here. I
have the parts to make a somewhat broader-band (1 kHz - 6
MHz) version and maybe someday I will. The original
receiver was called the PASS for Plankton Acoustic Survey
System. Acronyms are slippery creatures.
Receivers for use in ultrasonic systems like TAPS, WHAPS,
etc. absolutely must have certain characteristics. They
must have stable gain. They must be low-noise, as the echo
voltages produced at the transducer from reverberation by
plankton are in the range of microvolts to millivolts.
Ideally, they would have a minimum of 75-80 dB of usable
dynamic range (a bit better than, say, a 12-bit converter).
And they absolutely need a fast-response but accurate
envelope detector (unless you push most of the signal
processing into a DSP chip and can deal with ac signals).
Oh, yes, it must also run on a single DC voltage.
The receiver for TAPS-8 came pretty close to meeting these
requirements. I will abstract that design from the manual
data on that system and post the circuit descriptions here.
I have Gerber files for the pcb's; they will find their way
her as well. Maybe you will find them interesting.
Something I haven't seen mentioned anywhere else that is
important for folks building ultrasonic systems using a
single transducer for transmit and receive. At the end of
the transmit pulse the voltage across the transducer does
not suddenly drop to zero. It decays at a seemingly
exponential rate until it enters the noise floor of your
receiver. What is interesting about this -- particularly if
you want to see echoes up close to the transducer -- is the
frequency content of this tail. It begins at the drive
frequency, of course, because this is just mechanical
energy of the not-too-heavily-damped ceramic continuing to
vibrate at that frequency. But rather quickly, it changes
in frequency to match other modes of vibration of the
ceramic element. In the case of the TAPS circular disk
transducers, it appears to transition to a radial or
possibly a circumferential mode because the frequency drops
rapidly away from the thickness resonance frequency.
When I discovered this, it finally dawned on me that the
proper matching circuit for a transducer to a receiver
should be a high-pass filter. Sure enough, doing that
cleaned up the tail of the transmit on the lower-frequency
TAPS transducers enough that I could see echoes from
objects as close as about 10 cm. If you are doing lab or
field work in the 50-500 kHz range and need to see echoes
very close to your transducer, you might try this
trick.