Searching \ for 'Serial Sonar Unit (SSU)' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: massmind.org/techref/io/serials.htm?key=serial
Search entire site for: 'Serial Sonar Unit (SSU)'.

Truncated match.
PICList Thread
'Serial Sonar Unit (SSU)'
1997\11\11@162147 by Rick Dickinson

flavicon
face
Fellow PIC-Listers,

Based on the feedback I have received, it seems that a design like the following
would make the most sense:

Break the SSU into two functional units -- the ranging unit and the
communications unit.  Each would be based around a 16F84-10 PIC
microcontroller.  A single 10 MHz oscillator chip will be used to drive both
PICs.  Descriptions of each unit are as follows:

Ranging unit:
=============
10 MHz means a .4nS cycle time.  Using a prescaler value of 250 on Timer0 gives
us a .1mS Timer0 interrupt period to use to keep track of our time-of-flight
count.

Pin assignment will be as follows for this module:

Port A
------
 0 - Sonar Ping Out <-- Drives the out transducer at 40KHz through buffers.
 1 - Data Valid Out <-- Raised by PIC when data out is valid.  Lowered when
                         new ranging in progress.
 2 - Sonar Ping In  <-- Flag raised by analog ckt. when echo detected.
 3 - Range Mode In  <-- High = continuous, low = one-shot
 4 - Range Trigger  <-- Brought high momentarily to trigger one-shot ranging.

Port B
------
0-7 - 8-bit Data Out

Communications Unit:
====================

Pin assignment will be as follows for this module:

Port A
------
 0 - Serial Data Out   <-- Serial Data Out to the world.
 1 - Range Mode Out    <-- Tells the Ranging Unit (RU) what mode to use.
 2 - Range Trigger Out <-- Triggers the RU when in one-shot mode.
 3 - Data Valid In     <-- Input from RU signals that data is valid.
 4 - Serial Data In    <-- Serial Data In from world. Using RA4 allows us
                            to use TMR0 rollover to detect start bit.
Port B
------
0-7 - 8-bit Data In

Overview of Design:
===================

Having the Communications Unit (CU) and the Ranging Unit (RU) live on separate
16F84-10 chips gives us three big advantages:

1) We are free to use TMR0 for both serial communications timing
    and for time-of-flight timing.
2) Different communication algorithms and protocols can be tried
    by simply substituting CU chips.
3) Different range scaling factors or methods can be accomodated
    by simply substituting RU chips.

Additionally, as this is to be a collaborative effort, this allows different
people to work on different pieces of the puzzle without worrying that their
efforts will overlap or conflict.

Does anyone want to volunteer some programming expertise and development time?
Anyone? Anyone?

- Rick "Bueller?" Dickinson

+--------------------------------+------------------------------+
| Rick "The Notes Guy" Dickinson | Views expressed by the author|
| Lotus Notes & Domino Solutions | are not necessarily those of |
| (818)563-1061 spam_OUTrtdTakeThisOuTspamnotesguy.com | Enterprise ArchiTechs.  They |
|   http://www.eArchiTechs.com   | might not even be mine....   |
+--------------------------------+------------------------------+

1997\11\11@191653 by Sean Breheny

face picon face
At 01:19 PM 11/11/97 -0800, you wrote:
>Does anyone want to volunteer some programming expertise and development
time?
> Anyone? Anyone?
>
> - Rick "Bueller?" Dickinson

Rick,

I love your idea and I was actually working on a simpler sonar-ranging
device before this idea got introduced as a thread here. I am mainly
developing the best actual RX and TX circuits as well as what is the best
format for the sonar pings ( # of pulses, null period to avoid transmitted
vibration detection, etc. ). I have only been working on it for a few days,
so far I have a very simple circuit which seems to have very little problem
actually receiving the signal. I am mainly ironing out the timing issues
and digital detection method (I am currently using a PLL, but this is
having trouble locking reliably on only 40 cycles, which I thought might
happen, so I am moving from a PLL to a comparator clipping/direct PIC
detection method). I would like to help, not sure how much time I can
contribute, but I will contribute any major advances that I have. I have
two questions, what type of maximum range do you want this to handle and
what will be the accoustics of the areas where you envision it being used?
This is tough to implement >10meters range when indoors because stray
reflections are detected.

Sean

Sean Breheny,KA3YXM
Electrical Engineering Student

1997\11\11@194637 by Rick Dickinson

flavicon
face
At 07:16 PM 11/11/97 -0500, Sean Breheny wrote:
{Quote hidden}

Sean,

I was actually thinking of simply counting on the transducer's own resonance to
filter things down to just the 40 KHz pulse, and using a couple of op-amps to
boost the 40 KHz echo signal to the point that it clips and is detectable by
the PIC.  It sounds like exactly what you are doing.

Are you using the Murata transducers from All Electronics?  Those seem to be
cheap and readily available, so I was planning on using those.

As far as transmit goes, I was thinking of using a simple FET or BJT amplifier
stage as follows:
                    Vcc
                     |
                     >
                     < Rc
                     >      C1
                     |____||____
    |             | /     ||    |
PIC  |      Rb     |/            |
Port |----/\/\/----|   Q1        [<   Transducer
Pin  |             |\            |
    |             | X           |
                     |          |
                     |          |
                     V Gnd      V Gnd

What sort of drive circuitry had you planned on using?

- Rick
---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at .....rtdKILLspamspam@spam@notesguy.com

1997\11\11@195915 by Rick Dickinson

flavicon
face
At 07:16 PM 11/11/97 -0500, Sean Breheny wrote:
>I have
>two questions, what type of maximum range do you want this to handle and
>what will be the accoustics of the areas where you envision it being used?
>This is tough to implement >10meters range when indoors because stray
>reflections are detected.

Sorry I forgot to actually anser your questions in my first reply... :-)

As I initially envisioned this, it was a tool for model robotics.  As such, I
imagined ranges of a few inches to about 10 feet maximum, and primarily indoor
usage.  Some PICLISTers have suggested other uses, such as measuring stock in
grain silos or water or oil tanks.  In those cases, it would be ranges from
several feet to several dozen yards.

By placing the transducers inside of pipes or cylinders lined with some
sound-absorbing substance (like velvet or some other cloth) we should be able
to get good directivity, and with different resistor values in the op-amp
circuits, we should be able to tweak the gain to accomodate the longer
distances.  I have been planning on 8 bits of granularity in the measurements,
scaled over whatever time period is appropriate.  For robotics work, the scale
would obviously be smaller than for the grain silo application....

Because PICs are so easy to reprogram, it makes sense to have several "standard"
Ranging Unit programs available which scale the 256 possible output values over
various distances.  End users could simply pick whatever range was most
appropriate for their application.

What do you think?

- Rick
---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at rtdspamKILLspamnotesguy.com

1997\11\11@200713 by Ross McKenzie

flavicon
face
Hi Rick and other interested parties,

I don't want to be seen to be raining on your parade, .... but has anyone
thought to look at the sensor science behind this project yet?

Back in the mid '70s I fleshed out the design for an ultrasonic anemometer
based upon an orthogonal array of 40KHz transducers. As I recall it, the
speed of sound is a function of air temperature and moisture content. If you
ignore these, as it appears is your intent from your pin assignments below,
you are likely to reap the consequential accuracy and stability limits.
Perhaps "y'all" need to decide what your project objectives are in terms of
performance and then look at what units, baud rates etc can be applied.

Ross "Often the realist" McKenzie
Melbourne Australia

***********
At 01:19 PM 11/11/97 -0800, you wrote:
>Fellow PIC-Listers,
>
>Based on the feedback I have received, it seems that a design like the
following
{Quote hidden}

1997\11\11@200919 by Sean Breheny

face picon face
At 04:44 PM 11/11/97 -0800, you wrote:
>Sean,
>I was actually thinking of simply counting on the transducer's own
resonance to
> filter things down to just the 40 KHz pulse, and using a couple of
op-amps to
> boost the 40 KHz echo signal to the point that it clips and is detectable by
> the PIC.  It sounds like exactly what you are doing.
>

Yes, I agree that for the most part, you can rely on the transducer's
natural narrow bandwidth to exclude stray, non-ultrasonic signals. I was
simply using the PLL to tell the PIC when there was a signal/ when there
was not a signal. It was more laziness than need! I didn't want to bother
writing code for the PIC to detect pulses when the PLL would deliver a
simple yes/no answer via its "carrier detect" pin. But, as you suggest, it
looks like the direct pic detection scheme will work better.
       My receive amplifier is ONE transistor, operating in maximum gain
configuration. I have yet to design a clipper circuit to add to this. I
tried an op amp for the amplifier, but I have bad luck and little patience
with op amps as linear/semi-linear amps for high frequencies, so I tried
the discreet approach, and it turns out that the single transistor delivers
quite a sizable signal. I plan on adding an op amp in open-loop mode as a
comparator to clip the signal, same as you suggest.

>Are you using the Murata transducers from All Electronics?  Those seem to be
> cheap and readily available, so I was planning on using those.
>

They are not from All Electronics, they are from Mouser. I am not sure of
the manufacturer, it looks like it may be Kobitone, it's on the same page
as other Kobitone products. They are abt 1/2 inch diameter cylinders of
about 1/2 inch height also.
They are for 40KHz.

>As far as transmit goes, I was thinking of using a simple FET or BJT
amplifier
{Quote hidden}

I am currently using a single bipolar transistor, base to pic pin, emitter
to GND, collector thru primary of transformer to positive. The transformer
is a 10:1 or so audio matching type. I'm using it to step up the voltage to
the xducer. Xducer is connected across the secondary, of course. I tried
using the transducer directly off two PIC pins, and pulsing them 180 deg
out of phase, to get effectively 10 V across Xducer but that did not work
as well as this method with transformer. This appears to be getting out a
good signal. I hooked up the receive transducer directly to a meter and at
a distance of about two feet from the transmit transducer, got an output
voltage on the receiver of about 400 to 600 mV. With the single transistor
amp, I get about 900 mV when bouncing off an object about 7 feet away.

>
> - Rick
>---
> You can help design a Serial Sonar Unit for model robots!
>  Vist http://www.notesguy.com/notesguy for details, or
>   send mail to Rick Dickinson at EraseMErtdspam_OUTspamTakeThisOuTnotesguy.com
>


Sean Breheny,KA3YXM
Electrical Engineering Student

1997\11\11@202420 by J.Adams

picon face
SNIP! SNIP!

>> - Rick
>>---
>> You can help design a Serial Sonar Unit for model robots!
>>  Vist http://www.notesguy.com/notesguy for details, or
>>   send mail to Rick Dickinson at rtdspamspam_OUTnotesguy.com

Rick,

A few years back Lawrence Livermore Labs was working on a ultra
low power pulsed radar device that would go for about $10US. I am
not however sure what happened to the device except a few companies
were using it to detect rebar and in automobile colision detection
circuits.  I am not sure where the technology lay now but it
may be useful in gleaning design specs from.

I will pull out the article and see what I can pull up in the present
that may help you.  One of these devices hooked to a PIC and serial
chip would be just as cheap as an ultrasonic unit.


John Adams  --  http://pobox.com/~electronics  --  @spam@electronicsKILLspamspampobox.com
Check out H.W.S Internet Guide to Electronics as well. Links on my site.

1997\11\11@202837 by Sean Breheny

face picon face
At 04:58 PM 11/11/97 -0800, you wrote:
>At 07:16 PM 11/11/97 -0500, Sean Breheny wrote:
>>I have
>>two questions, what type of maximum range do you want this to handle and
>>what will be the accoustics of the areas where you envision it being used?
>>This is tough to implement >10meters range when indoors because stray
>>reflections are detected.
>
>Sorry I forgot to actually anser your questions in my first reply... :-)
>
>As I initially envisioned this, it was a tool for model robotics.  As such, I
> imagined ranges of a few inches to about 10 feet maximum, and primarily
indoor
> usage.  Some PICLISTers have suggested other uses, such as measuring
stock in
> grain silos or water or oil tanks.  In those cases, it would be ranges from
> several feet to several dozen yards.
>
>By placing the transducers inside of pipes or cylinders lined with some
> sound-absorbing substance (like velvet or some other cloth) we should be
able
> to get good directivity, and with different resistor values in the op-amp
> circuits, we should be able to tweak the gain to accomodate the longer
> distances.  I have been planning on 8 bits of granularity in the
measurements,
> scaled over whatever time period is appropriate.  For robotics work, the
scale
> would obviously be smaller than for the grain silo application....
>
>Because PICs are so easy to reprogram, it makes sense to have several
"standard"
> Ranging Unit programs available which scale the 256 possible output
values over
> various distances.  End users could simply pick whatever range was most
> appropriate for their application.
>
>What do you think?
>
> - Rick
>---

My design was originally for use in physics demonstrations or velocity and
acceleration, so my goal was resolution of about 3 cm or so over about 7
meters, about the same as your robotics idea. I have worked on the design
of a similar project before, and we then experienced significant difficulty
in working over 15 feet reflected distances (30 feet total round trip) when
indoors, presumably due to reflections. The other guy who was the main
designer of the project (who is also on this list, Louis Marquette), has
since tried placing the transducers in cylinders and I think that he had
good success.

Previously, you said that you wanted to avoid using an AGC. Theoretically,
if we design the comparator/clipper circuit well enough and provide enough
gain, it should respond equally well to weak signals and strong ones. You
may not need to adjust the gain for longer distances. The main thing you
need to do is adjust the timing and guard against the multipath.

You might also have difficulty with getting the circuit to work with very
close ranges, like several inches, for a few reasons:
1.) The transmitting transducer as well as the whole device will vibrate
during the TX period. They will continue to vibrate slightly at 40KHz for a
few microseconds or so after the TX time. If you start listening for an
echo  soon enough to hear one which comes from only a few inches away, you
will most likely hear the vibrations instead of the echo.
2.) The two transducers are probably separated by a few centimeters on the
device. This means that the signal path does not go straight out, but
instead must bounce off at an angle and return, forming an equilateral
triangle, not a straight line. Since the wavefront spreads out over long
distances, and the length of the path becomes long compared to the distance
between transducers, this effect is negligible at several feet. But I
imagine that it may become pronounced at very short distances.

I have also seen in several places recommendations that a specific pattern
of pulses be used rather than a striaght 40KHz burst. This allows you to
throw out noisy readings because the return pulses will not fit the
expected pattern. I imagine that the pattern must be carefully designed so
that it is still recognizable after being passed through the transducers,
which act effectively as a narrow 40KHz band pass filter.

Sean



> You can help design a Serial Sonar Unit for model robots!
>  Vist http://www.notesguy.com/notesguy for details, or
>   send mail to Rick Dickinson at KILLspamrtdKILLspamspamnotesguy.com
>
Sean Breheny,KA3YXM
Electrical Engineering Student

1997\11\11@204943 by Sean Breheny

face picon face
At 12:03 PM 11/12/97 +1100, you wrote:
>Hi Rick and other interested parties,
>
>I don't want to be seen to be raining on your parade, .... but has anyone
>thought to look at the sensor science behind this project yet?
>
>Back in the mid '70s I fleshed out the design for an ultrasonic anemometer
>based upon an orthogonal array of 40KHz transducers. As I recall it, the
>speed of sound is a function of air temperature and moisture content. If you
>ignore these, as it appears is your intent from your pin assignments below,
>you are likely to reap the consequential accuracy and stability limits.
>Perhaps "y'all" need to decide what your project objectives are in terms of
>performance and then look at what units, baud rates etc can be applied.
>
>Ross "Often the realist" McKenzie
>Melbourne Australia

Ross,

I did a little bit of checking and here is how the speed of sound in still
air varies with humidity and temperature:

At 20deg Cel.

0% Rel Humidity for 40KHz : 1127.188 ft/sec
100% RH for 40KHz: 1130.963 ft/sec

Not a bit deal at all for RH, would cause less than 1% error.

For Temp:

30 deg Cel (0% RH): 1145.6 ft/sec
0 deg Cel (0% RH): 1087.4 ft/sec

Here, we have about a 5.3 % variation over the temp range 0 to 30 deg C.
This is not too bad either because it is unlikely that most of the
applications mentioned here (with the possible exception of the grain silo)
would not subject the unit to such a variability in environment. Besides,
does a farmer really care THAT much if he only knows his grain storage
amount to 5% on a real time basis?

In your anemometer design, I would think that you were measuring the change
in the speed of sound in air due to air movement. This is a somewhat more
pronounced effect, since it is almost like velocity addition. However, even
extreme wind speeds are only a fraction of the speed of sound.

Sean


Sean Breheny,KA3YXM
Electrical Engineering Student

1997\11\11@211315 by William Chops Westfield

face picon face
   A few years back Lawrence Livermore Labs was working on a ultra
   low power pulsed radar device that would go for about $10US.

I was thinking of that too.  There's a bunch of info on the web on this
("Micropower Impulse Radar") at http://www.llnl.gov, but they're being pretty
close-mouthed on the details unless you pony up licence fees...

BillW

1997\11\11@214602 by Joe Little

flavicon
face
    I'll create the GERBER files and put them on the WEB when the schematic and
    board outline appears.
    Joe

1997\11\11@221939 by Rick Dickinson

flavicon
face
At 08:27 PM 11/11/97 -0500, Sean Breheny wrote:
>My design was originally for use in physics demonstrations or velocity and
>acceleration, so my goal was resolution of about 3 cm or so over about 7
>meters, about the same as your robotics idea. I have worked on the design
>of a similar project before, and we then experienced significant difficulty
>in working over 15 feet reflected distances (30 feet total round trip) when
>indoors, presumably due to reflections. The other guy who was the main
>designer of the project (who is also on this list, Louis Marquette), has
>since tried placing the transducers in cylinders and I think that he had
>good success.

I can't claim complete credit for this idea, myself, either.  I had heard that
people had successully done this with the Polaroid transducers.

>Previously, you said that you wanted to avoid using an AGC. Theoretically,
>if we design the comparator/clipper circuit well enough and provide enough
>gain, it should respond equally well to weak signals and strong ones. You
>may not need to adjust the gain for longer distances. The main thing you
>need to do is adjust the timing and guard against the multipath.

Agreed.  The main design goals are low cost and reliable operation.  If we can
make this device work with a simpler circuit, then the costs stay low.  Also, I
don't want the end user to need to "fine tune" anything. I want it to work when
assembled.

>You might also have difficulty with getting the circuit to work with very
>close ranges, like several inches, for a few reasons:
> 1.) The transmitting transducer as well as the whole device will vibrate
>during the TX period. They will continue to vibrate slightly at 40KHz for a
>few microseconds or so after the TX time. If you start listening for an
>echo  soon enough to hear one which comes from only a few inches away, you
>will most likely hear the vibrations instead of the echo.

True.  This is a big problem with the Polaroid unit, from what I hear.  Since
the same transducer is doing both transmit and receive functions, it is even
more susceptible to false triggering if its "blanking interval" (the time
period it doesn't look for an echo) is set too low.

I was thinking that the transducers should be fairly snugly mounted in some sort
of damping medium.  Maybe something along the lines of cork or rubber, or maybe
some sort of styrofoam would work.  Also, placing them in tubes to increase
their directivity should help.

>2.) The two transducers are probably separated by a few centimeters on the
>device. This means that the signal path does not go straight out, but
>instead must bounce off at an angle and return, forming an equilateral

Isosceles, actually, but I understand your point.

>triangle, not a straight line. Since the wavefront spreads out over long
>distances, and the length of the path becomes long compared to the distance
>between transducers, this effect is negligible at several feet. But I
>imagine that it may become pronounced at very short distances.

Although the absolute accuracy may drop off for very short measurements if you
assume that the return time is a linear function of distance to the target, the
measurements would still be useful, in any case, as a way of distinguishing
"closer" or "farther".  Also, if accuracy is needed at close range, consider
the following diagram:

            D
____________________________ Wall or other object
            ^
           /|\            A & B are the transducers.
          / | \           CD is the distance from the sensor
         /  |  \               array to the wall
        /   |   \         AD + DB is the distance measured (AD=DB)
       /    |    \        By the Pythagorean theorem:
      /-----+-----\         2     2     2
      A     C      B      AD  = AC  + CD

Since we know AD and AC, we can solve for CD as follows:
                 __________
                /   2     2
          CD = V  AD  - AC

>I have also seen in several places recommendations that a specific pattern
>of pulses be used rather than a striaght 40KHz burst. This allows you to
>throw out noisy readings because the return pulses will not fit the
>expected pattern. I imagine that the pattern must be carefully designed so
>that it is still recognizable after being passed through the transducers,
>which act effectively as a narrow 40KHz band pass filter.

Probably a good idea, but well beyond the scope of a "first pass" at the
project.  This might make a good area to experiment with in the future,
however.  Hopefully, there won't be much 40KHz noise in the environments this
will be operating in.

- Rick "Once we finish the ping, next we do tracert" Dickinson


---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at RemoveMErtdTakeThisOuTspamnotesguy.com

1997\11\11@224652 by Rick Dickinson

flavicon
face
At 08:48 PM 11/11/97 -0500, Sean Breheny wrote:

{Quote hidden}

Sean,

Thanks for the research into the actual numbers.  It doesn't sound like either
effect will cause much of a problem for us.  I can't imagine that our pulse
detection circuitry will have an accuracy much better than a few percent over
short distances, anyway.  As long as we can come "pretty close", I think that
we will be fine.

- Rick

---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at spamBeGonertdspamBeGonespamnotesguy.com

1997\11\11@224858 by Rick Dickinson

flavicon
face
At 04:44 PM 11/11/97 -0500, Joe Little wrote:
>     I'll create the GERBER files and put them on the WEB when the schematic
and
>     board outline appears.
>     Joe

Thanks, Joe.  Consider yourself "drafted" onto the project.

- Rick

---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at TakeThisOuTrtdEraseMEspamspam_OUTnotesguy.com

1997\11\11@232419 by Clyde Smith-Stubbs

flavicon
face
On Tue, Nov 11, 1997 at 04:58:23PM -0800, Rick Dickinson wrote:

> By placing the transducers inside of pipes or cylinders lined with some
>  sound-absorbing substance (like velvet or some other cloth) we should be able

I've found the ultrasonic transducers are highly directional anyway. For
liquid-level measurements it's probably necessary to use a pipe, to avoid
reflections from the tank sides etc, but for free-air ranging it's not
particular necessary.

--
Clyde Smith-Stubbs               |            HI-TECH Software
Email: RemoveMEclydespamTakeThisOuThtsoft.com          |          Phone            Fax
WWW:   http://www.htsoft.com/    | USA: (408) 490 2885  (408) 490 2885
PGP:   finger clydeEraseMEspam.....htsoft.com   | AUS: +61 7 3354 2411 +61 7 3354 2422
---------------------------------------------------------------------------
ANSI C for the PIC! Now shipping! See http://www.htsoft.com for more info.

1997\11\11@232428 by J.Adams

picon face
>> 1.) The transmitting transducer as well as the whole device will vibrate
>>during the TX period. They will continue to vibrate slightly at 40KHz for a
>>few microseconds or so after the TX time. If you start listening for an
>>echo  soon enough to hear one which comes from only a few inches away, you
>>will most likely hear the vibrations instead of the echo.
>
>True.  This is a big problem with the Polaroid unit, from what I hear.  Since
> the same transducer is doing both transmit and receive functions, it is even
> more susceptible to false triggering if its "blanking interval" (the time
> period it doesn't look for an echo) is set too low.
>
>I was thinking that the transducers should be fairly snugly mounted in
some sort
> of damping medium.  Maybe something along the lines of cork or rubber, or
maybe
> some sort of styrofoam would work.  Also, placing them in tubes to increase
> their directivity should help.


You may want to send out a distinct pulse that varies.  Example: A 4 ping
pulse
then a 20 pin pulse....etc.  This might negate the receiver reading the
wrong pulse or outside influences or bounces.


John Adams ------- http://pobox.com/~electronics ---------
EraseMEelectronicsspampobox.com
Check out the H.W.S Internet Guide to Electronics book as well. Links on my
site

1997\11\12@002630 by Robert Nansel

flavicon
face
Some thoughts on the analog side:

Use a transformer to match impedances, for sure. Also, what I've seen of
ultrasonic transducers, especially the piezo variety, is that they are such
high Q resonators that you don't even really need to drive them with an
accurate frequency. A friend of mine in the Seattle Robotics Society built
a system where just hitting the transducer with a pulse was enough to
elicit an ultrasonic pulse (much like ringing a bell with a hammer tap).
He used three op amps (actually, a TL082 dual op amp and an LM311
comparator) for the receive function, if I recall correctly. It didn't use
AGC, but had a very simple timed gain control, so that the longer you
waited for the pulse to return (and therefor the weaker the pulse you are
looking for), the more gain. I can dig up the schematics if anyone is
interested.

One thing that always intrigued me about Keith's design was how simple it
was: he used just a transistor switch (TIP122, darlington configuration), a
few resistors & caps, and a 1:5 impedance matching transformer (a TK2002, I
think) to drive the transducer. There was a 1500 pF cap in parallel with
the transmit transducer to make it behave like a resonant tank circuit.

I remember Keith told me that he could watch the transducer ring decay on
his scope; the decay period was long enough that it was this that limited
the close-in range. It occurred to me that, since this design was entirely
software driven (with some artful analog electronics where needed), that if
you paid very careful attention to the frequency and phase of the forcing
function that it should be possible to get both a good loud chirp _and_ a
quick decay time. It would go something like this:

       1) Bang on the transducer for N pulses, where N < 10, with a pulse
          frequency selected to match the 40 KHz center frequency.

       2) Wait an integral number of cycles (however long you want the
          chirp to be) plus 1/2 cycle.

       3) Bang on the transducer again for M pulses (M<N), but now you
          are 180 degrees out of phase with the transducer oscillation,
          so this has the effect of _damping_ the oscillation, thus
          giving a cleaner and quicker pulse.


Now, I don't know of anybody who's tried this, but it sure sounds worth a
try. Thoughts?

--BN

----------------------------------------------------------------------------
  Country Robot            "Modular robot components
  69 S. Fremont Ave. # 2     for education and industry"
  Pittsburgh, PA 15202
----------------------------------------------------------------------------

1997\11\12@024304 by J.Adams

picon face
Rick,

I have been doing some mental calculations to help you:

Sound travels 1130 feet/sec at 75 degrees far. This is a great
place to start.  (The deviation is 1.1 Ft / degree far by the way)

That is 13,560 inches per second.

Using the formula  d = t * 13,560/2 you can figure out the distance
(in inches) from the unit.
d= distance in inches
t= time to ping an object and back

Therefore:
distance to object = time for ping to get there and back times 6780

d=t*6780

A 40Khz sound wave has a wavelength of .33 inches so I am not sure
if you can get much better resolution then .33 inches. Maybe I am
wrong on this - Does anyone else dispute this?

Now remember this is all in the AIR medium.  Water is about 4 times
these numbers as sound travels through liquids at 4,757 ft/sec.
And Steel is around 16,500 Ft/sec.

Interesting subject, Rick. Thanks for instigating it.


John Adams ------- http://pobox.com/~electronics ---------
RemoveMEelectronicsEraseMEspamEraseMEpobox.com
Check out the H.W.S Internet Guide to Electronics book as well. Links on my
site

1997\11\12@064017 by Octavio Nogueira

flavicon
face
>At 01:19 PM 11/11/97 -0800, you wrote:
>Does anyone want to volunteer some programming expertise and development
time?
> Anyone? Anyone?
>
> - Rick "Bueller?" Dickinson

Consider myself a volunteer, Rick.

Octavio
======================================================
Octavio Nogueira  - e-mail:   RemoveMEnogueiraspam_OUTspamKILLspammandic.com.br
http://www.geocities.com/~oct_nogueira
"ProPic" Production PIC Programmer Windows under US$20
======================================================

1997\11\12@072856 by Pasi T Mustalahti

picon face
On Wed, 12 Nov 1997, J.Adams wrote:

> Rick,
>
> A 40Khz sound wave has a wavelength of .33 inches so I am not sure
> if you can get much better resolution then .33 inches. Maybe I am
> wrong on this - Does anyone else dispute this?
PTM: Use phase lock to get more decimals :)
--------------------------------------------------------------------------
PTM, RemoveMEpasi.mustalahtiTakeThisOuTspamspamutu.fi, EraseMEptmustaspamspamspamBeGoneutu.fi, http://www.utu.fi/~ptmusta
Lab.ins. (mikrotuki) ATK-keskus/Mat.Luon.Tdk                    OH1HEK
Lab.engineer (PC support) Computer Center                       OI7234
Mail: Turun Yliopisto / Fysla, Vesilinnantie 5, 20014
Pt 02-3336669, FAX 02-3335632 (Pk 02-2387010, NMT 049-555577)
--------------------------------------------------------------------------

1997\11\12@095034 by Tom Rogers

flavicon
face
Check out section 2.2 of the article "The production of acoustic impulses in
air" by Ian Marshall, published in the J. Phys. E (European Journal of
Physics E, Measurement Science and Technology 1 1990 413-418). This
describes a "step" method of driving a ringing transducer, which is what you
are trying to describe in your post.

--Tom Rogers  VP-R&D  Time Tech Inc.


{Original Message removed}

1997\11\12@105052 by Rick Dickinson

flavicon
face
At 09:26 AM 11/12/97 -0200, Octavio Nogueira wrote:
>>At 01:19 PM 11/11/97 -0800, you wrote:
>>Does anyone want to volunteer some programming expertise and development
>time?
>> Anyone? Anyone?
>>
>> - Rick "Bueller?" Dickinson
>
>Consider myself a volunteer, Rick.

Any particular area you'd like to work on?

Currently, I see three sections:
 - Analog circuitry (transducer drive and echo detection)
 - Ranging Unit (PIC program to pulse transducer & time return of pulse)
 - Communications Unit (PIC program to handle serial comm/protocol)

- Rick

Analog circuit
---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at RemoveMErtdKILLspamspamnotesguy.com

1997\11\12@123044 by Paul BRITTON

flavicon
face
Hi there,
Ive been listening in to this...sounds interesting....personally I have no
need for such a unit at present...
but still interesting.
Q. Will ambient air pressure variation affect sound velocity
significantly?

Just asking....

TTYL Paul

>At 08:48 PM 11/11/97 -0500, Sean Breheny wrote:
>
>>I did a little bit of checking and here is how the speed of sound in
still
{Quote hidden}

silo)
>>would not subject the unit to such a variability in environment.
Besides,
>>does a farmer really care THAT much if he only knows his grain storage
>>amount to 5% on a real time basis?
>>
>>In your anemometer design, I would think that you were measuring the
change
>>in the speed of sound in air due to air movement. This is a somewhat
more
>>pronounced effect, since it is almost like velocity addition. However,
even
>>extreme wind speeds are only a fraction of the speed of sound.
>
>Sean,
>
>Thanks for the research into the actual numbers.  It doesn't sound like
either
> effect will cause much of a problem for us.  I can't imagine that our
pulse
> detection circuitry will have an accuracy much better than a few percent
over
> short distances, anyway.  As long as we can come "pretty close", I think
that
> we will be fine.
>
> - Rick
>
>---
> You can help design a Serial Sonar Unit for model robots!
>  Vist http://www.notesguy.com/notesguy for details, or
>   send mail to Rick Dickinson at rtdSTOPspamspamspam_OUTnotesguy.com

1997\11\12@132000 by Alessandro Zummo

flavicon
face
Il 11-Nov-97, Rick Dickinson scrisse:


>Does anyone want to volunteer some programming expertise and development
time?
> Anyone? Anyone?

i'm a newbie, so i'm not able to write such a code, but i think
that cross-platform compatibility must be guaranteed.
i.e. only standard PIC assembly instructions should be used, without using
MPASm auto macros or something similar.. they aren't supportd
by all pic assemblers...
--

  - *Alex* -

     (spamBeGoneazummoSTOPspamspamEraseMEita.flashnet.it)

1997\11\12@160958 by Alan G. Smith

flavicon
face
Rick you can sign me up for working on a communication chip for I2C and
one
for Rx/Tx.

--Alan

+---------------
| Alan G. Smith
| KILLspamagsspamBeGonespampoboxes.com
| http://www.innovatus.com/ags

1997\11\12@174558 by Rick Dickinson

flavicon
face
At 03:57 PM 11/12/97 -0500, Alan G. Smith wrote:
>Rick you can sign me up for working on a communication chip for I2C and
>one
>for Rx/Tx.
>
>--Alan

Thanks, Alan.  You're on the list.

- Rick

---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at EraseMErtdspamEraseMEnotesguy.com

1997\11\12@183127 by Robert Nansel

flavicon
face
>Additionally, as this is to be a collaborative effort, this allows different
> people to work on different pieces of the puzzle without worrying that their
> efforts will overlap or conflict.
>
>Does anyone want to volunteer some programming expertise and development time?
> Anyone? Anyone?
>
> - Rick "Bueller?" Dickinson

I'm interested in working on the analog design of the Range unit & the
comm. protocol for the Communication processor.

-BN

----------------------------------------------------------------------------
  Country Robot            "Modular robot components
  69 S. Fremont Ave. # 2     for education and industry"
  Pittsburgh, PA 15202
----------------------------------------------------------------------------

1997\11\12@200321 by Rick Dickinson

flavicon
face
At 06:27 PM 11/12/97 -0500, Robert Nansel wrote:
>>Additionally, as this is to be a collaborative effort, this allows different
>> people to work on different pieces of the puzzle without worrying that their
>> efforts will overlap or conflict.
>>
>>Does anyone want to volunteer some programming expertise and development time?
>> Anyone? Anyone?
>>
>> - Rick "Bueller?" Dickinson
>
>I'm interested in working on the analog design of the Range unit & the
>comm. protocol for the Communication processor.
>
>-BN
>
>----------------------------------------------------------------------------
>   Country Robot            "Modular robot components
>   69 S. Fremont Ave. # 2     for education and industry"
>   Pittsburgh, PA 15202
>----------------------------------------------------------------------------

Consider yourself on the team.  Thanks!

- Rick
---
You can help design a Serial Sonar Unit for model robots!
 Vist http://www.notesguy.com/notesguy for details, or
  send mail to Rick Dickinson at @spam@rtd@spam@spamspam_OUTnotesguy.com

1997\11\12@220904 by Ron Kreymborg

flavicon
face
Rick

My background is in both hardware and software design. I will help
wherever I can.

Ron

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ron Kreymborg                   Computer Systems Manager
Monash University               CRC for Southern Hemisphere Meteorology
Wellington Road
Clayton, VIC 3168               Phone     : 061-3-9905-9671
Australia                       Fax       : 061-3-9905-9689
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

>Additionally, as this is to be a collaborative effort, this allows different
> people to work on different pieces of the puzzle without worrying that their
> efforts will overlap or conflict.
>
>Does anyone want to volunteer some programming expertise and development time?

1997\11\12@234447 by pburdick

flavicon
face
Robert Nansel wrote:
{Quote hidden}

Very interested in schematics!  Please send.

               Perry

spamBeGonepburdickspamKILLspambcn.net

1997\11\13@015437 by William Chops Westfield

face picon face
I have my doubts that this device could be sold for much less than the $60
price of the polaroid unit once "real" costs are taken into account (ie no
more getting your transducers from surplus outlets.)  Still, I wouldn't look
a circuit, PCB, and SW working design in the mouth - sounds like lots of fun,
and sure to be at least publishable in one of the hobby magazines (or maybe
it would win Microchips 12C50x design contest.)

BillW

1997\11\13@063738 by Robert Nansel

flavicon
face
>I have my doubts that this device could be sold for much less than the $60
>price of the polaroid unit once "real" costs are taken into account (ie no
>more getting your transducers from surplus outlets.)  Still, I wouldn't look
>a circuit, PCB, and SW working design in the mouth - sounds like lots of fun,
>and sure to be at least publishable in one of the hobby magazines (or maybe
>it would win Microchips 12C50x design contest.)
>
>BillW


Keith's 2-chip ultrasonic system--the analog parts, anyway--cost around $35
in '93, and most of his parts came from Digi-Key, not surplus. I do notice
that Digi-Key no longer carries a few of parts, but there look to be
reasonable substitutes. Which just means we'll have to improve the old
design. I do like the idea of using 12C50X parts for this, though, to
reduce costs.

Actually, I think it would be best to produce several preliminary analog
designs and race them. We could figure out what version has the highest
performance/$, the best noise immunity, is the easiest to build & tune up,
etc. Maybe someone could volunteer to be the project tester/evaluator to
characterize the performance of the system(s) we come up with?


--BN

----------------------------------------------------------------------------
  Country Robot            "Modular robot components
  69 S. Fremont Ave. # 2     for education and industry"
  Pittsburgh, PA 15202
----------------------------------------------------------------------------

1997\11\13@063944 by Robert Nansel

flavicon
face
>Robert Nansel wrote:
>>
>> Some thoughts on the analog side:

   >Snip!<

>
>Very interested in schematics!  Please send.
>
>                Perry
>
>.....pburdickspam_OUTspambcn.net


Since these schematics are in the form of TIFF scans, how would you like me
to send them? I can e-mail the TIFFs directly, or I can convert them to
JPEGs. Or I 'spose I could do some ASCII graphics, though that could take
me a while (I'm going out of town for a few days).

The schematics originally appeared in the March '93 issue of the Encoder,
the newsletter of the Seattle Robotics Society, which I was editing at the
time, so there is an article that goes along with them. Maybe the best
route would be to just put photocopies in the mail?

--BN

----------------------------------------------------------------------------
  Country Robot            "Modular robot components
  69 S. Fremont Ave. # 2     for education and industry"
  Pittsburgh, PA 15202
----------------------------------------------------------------------------

1997\11\13@134511 by Eric van Es

flavicon
face
Robert Nansel wrote:

{Quote hidden}

Hi Robert

I think that the JPG is probably the best standard to use. Most ppl can view
them straight from win95 etc.

Anyone agree?

Can't you put it up on a homepage?

--
Eric van Es               | Cape Town, South Africa
TakeThisOuTvanesKILLspamspamspamilink.nis.za | http://www.nis.za/~vanes
LOOKING FOR TEMPORARY / HOLIDAY ACCOMODATION?
http://www.nis.za/~vanes/accom.htm

1997\11\13@155027 by Matt Bonner

flavicon
face
Eric van Es wrote:
>
> Robert Nansel wrote:
>
> > Since these schematics are in the form of TIFF scans, how would you like me
> > to send them? I can e-mail the TIFFs directly, or I can convert them to
> > JPEGs. Or I 'spose I could do some ASCII graphics, though that could take
> > me a while (I'm going out of town for a few days).
> >
> > The schematics originally appeared in the March '93 issue of the Encoder,
> > the newsletter of the Seattle Robotics Society, which I was editing at the
> > time, so there is an article that goes along with them. Maybe the best
> > route would be to just put photocopies in the mail?

Not the mail! In Canada, our snail mail is going to get even slower -
the biennial Christmas postal strike is looming...
>
> I think that the JPG is probably the best standard to use. Most ppl can view
> them straight from win95 etc.
>
> Anyone agree?
Agree
>
> Can't you put it up on a homepage?
Any volunteers, Eric? <g>

--Matt

1997\11\14@170243 by Eric van Es

flavicon
face
Matt Bonner wrote:

> Eric van Es wrote:
>



> Not the mail! In Canada, our snail mail is going to get even slower -
> the biennial Christmas postal strike is looming...
>

Ha! Around here mail from 200km from me takes 2 (two) days. Mail from Cape Town
itself (30km) takes 7 to 10 days!

Then the idiot of a minister of ptt increases stamp revenue because "services
have
been upgraded." Bull$%#@

> > I think that the JPG is probably the best standard to use. Most ppl can view
> > them straight from win95 etc.

I've been told that that is not the case. However .BMP file are too large. So if
you want it it get a .JPG-viewer. I recommend Irfanview32. Its freeware and
excellent.

> > Can't you put it up on a homepage?
> Any volunteers, Eric? <g>

OK Matt - you got me. If Robert mails them to me I'll zip and zip2exe 'em and
put
them up on my homepage.

Cheers!
--
Eric van Es               | Cape Town, South Africa
.....vanesspamRemoveMEilink.nis.za | http://www.nis.za/~vanes
LOOKING FOR TEMPORARY / HOLIDAY ACCOMODATION?
http://www.nis.za/~vanes/accom.htm

1997\11\14@172924 by Roger Books

flavicon
face
{Quote hidden}

You can view jpeg's with Netscape and Internet Explorerer.  Just do an
open/local file.  As for .BMP's being too large, if you are going to 'zip
em up you will be amazed at how much they shrink, I know I was.

Roger

1997\11\16@062604 by paulb

flavicon
face
Robert Nansel wrote:

> Since these schematics are in the form of TIFF scans, how would you
> like me to send them? I can e-mail the TIFFs directly, or I can
> convert them to JPEGs.

 PLEEASE!  For the sanity of we "technical" types, never EVER convert
any image of importance to JPEG!  And particularly never any image
containing text!  JPEG may be OK for the odd naughty postcard, but if
you have *information* to send, please use GIF.

 Cheers,
       Paul B.

1997\11\16@070133 by Tim Kerby

picon face
Hi
I was looking at a method for measuring the speed of sound a few years ago
at school where you measured a distance to a wall, clapped your hands from
that distance and clapped on each echo.  This gave surprisingly good
results.  A method like this for sonar could work quite simply as follows:

40kHz ping is output

Analogue electronics receives this and starts a feedback loop where another
ping is instantly sent

The rangefinding PIC then counts the number of pings in a given time.

An inverse function is applied to this to give a large number for larger
distance using the speed of sound, time taken for measurement and number of
pings received.

This leaves the pic dealing with only counting and maths.



In terms of comms, keeping it simple is wise.  Suggestions of multi node
networks would complicate things and could be implemented in another pic
for that particular design.

I would stick to a serial output perhaps the first byte indicating with a
binary sequence the range (mm, cm, m etc) and the second the actual data.
A spare pin could be used to fix the range for applications where a wide
range is not required.  This would save problems with 256 bits and allow a
high accuracty.

An alternatice method would be to use three bytes.  The first two of the 8
bits indicate range (as data would only need to be up to 9mm, 99cm etc) and
the last 6 indicate data.  This would allow high precision on even the
metre ranfe where all three bytes could be used.


Some sort of calibration function would be useful where a distance of 1
metre is measured and the unit calibrated into its eeprom (on the f84) with
the speed of sound in those conditions.


If you need help on checking that components are available in the UK then I
have a pile of catalogues in my bedroom and I am more than happy to help.
I also have unlimited web space with my ISP so if you need it, it is there.



Tim Kerby


------------------------------------------------------------------
Personal Web Pages: http://web.ukonline.co.uk/members/tim.kerby/
Email: RemoveMEtim.kerbyspamspamBeGoneukonline.co.uk
------------------------------------------------------------------

1997\11\17@061232 by Scott Stephens

picon face
Here's my 2 cents on driving the transucer:

Have a look at Nation Semi's LM1812 Sonar Chip and app note, this is a chip
used for 'Humingbird' fishing sonar's. They use 1 transducer for transmit &
recieve.

I believe the spec's on the cheap 40KHz transducer is 300mw.
Measure the impedance of your transducer.
If it's over 75 ohms, at fr. you will need a driver because your PIC only
does 5volts.

So wrap a ferrite toriod coil to resonate with the transducer.
wrap a driver winding; each PIC port is good for +/- 20ma @ 5v, (100mw) so
you may want to parrallel 3 port lines.

{Quote hidden}

You only need to do 1), because after you chirp the step-up transformer's
primary, you can 'short it' and allow the PIC's port pins to dissipate the
stored energy, damping the reverberations on the transducer.

After a SHORT damping interval, put the PIC port pins in HI-Z state, so they
don't dampen the transducer when you want to use it as a microphone.


For a reciever, you realy want TDC- time dependant gain control. I like the
idea of ripping off an IR remote module. Maybe a FET in the feedback path of
an op-amp, used as a voltage-variable resistor, and opened after the
transmit pulse through a time constant. Try to use a hi-z input amp directly
off the transducer. Maybe not too hard at only 40KHz, if Z-in is just 1-10
Kohms.

Have fun.

>

1997\11\17@071621 by paulb

flavicon
face
Rick Dickinson wrote:

> Currently, I see three sections:
> - Analog circuitry (transducer drive and echo detection)
> - Ranging Unit (PIC program to pulse transducer & time return of
> pulse)
>   - Communications Unit (PIC program to handle serial comm/protocol)

 If, as I gather, the timing resolution is to be in the order of
milliseconds, or hundreds of microseconds, might it not be appropriate
to make the split between tasks at the point of timing; that is, one PIC
generates the pulses, controls receive gain ramping, correlates the
received signals and issues an "echo" pulse, while the other times the
echo pulse and formats serial data.

 The latter chip in this scheme would be very negotiable; MANY users
would choose to do their own timing.  It would simplify inter-chip
interfacing extremely; to two wires.  Protocol would go like:

1) Co-ordinator waits till "receive" line is idle, asserts "send" line
to start a cycle.

2) Ranging unit starts a send cycle and asserts RX line.

3) On receiving echo, it clears RX line so that the pulse length on
this line is exactly the round trip time (with any corrections to make
this the case being the responsibility or the Ranging unit).  There are
possibilities here for multiple echoes to be measured.

4) Clearing the TX line resets the Ranging unit.  Note that if a
further "ping" is requested before the Ranging unit is ready, it simply
declines to assert the RX line until it actually IS ready and has
started its "ping".

 AGC:  You may be aware that gain ramping (the exact term eludes me at
present) is an essential part of ultrasonic imaging, used perhaps most
extensively in medicine.  I suspect this is one of the design tricks of
the Polaroid system.  If you run at full gain to start, with or without
clipping, you are really likely to be troubled by ringing of the
transducer, mounting, etc.  And if you DO use gain ramping, you can
certainly consider a single transducer design.

 Most of the popular Doppler alarms seem to use rubber grommets to
mount the transducers anyway, presumably isolation is one reason.
Again, the main reason thy use two transducers is that the transmitter
is continuous.

 There must be a wealth of design information in the marine depth
sounder field.  They mostly use single transducers too, and the 4:1
scaling factor makes them very comparable to the presently considered
dimensions in air.  There are analog ICs designed for this purpose.

 The idea of damping or active damping of the transducer is
fascinating.  I suspect the marine ones probably do this too.

 Cosine error:  Probably not going to trouble anybody in practice.  For
close ranging you would have the transducers (all assuming you use two)
close anyway.  If you had to make corrections from a table and/ or
interpolation for these distances, it shouldn't be difficult.  I thought
the "ringing"/ gain consideration was more important.

 Cheers,
       Paul B.

1997\11\17@080739 by John Hallam

flavicon
picon face
On Tue, 11 Nov 1997, Rick Dickinson wrote:

> Fellow PIC-Listers,
>
> Based on the feedback I have received, it seems that a design like the
following
{Quote hidden}

Using these pin assignments is going to make life difficult for anyone
who wants to implement a full speed multimaster I^2 C bus interface...
But maybe few people want that?

John Hallam, AI Dept, Edinburgh University.

1997\11\17@111457 by Robert Nansel

flavicon
face
{Quote hidden}

Why not reverse port pins 1 & 4 so both serial lines are next to each
other? This would allow those folks who want to give the I2C option a try
to use the same PCB layout, yet still be able to do an AN541-style
bit-banger.

Also, do we really want to lock this into a PIC that must have a bit 4 on
Port A? Especially considering you are giving up the T0CKI line to do so.
Maybe another port would be more appropriate? We could go with nybble-mode
transfers instead of using all of Port B for this purpose (this would be my
preference). Or maybe we should look for a way to get rid of the Range Mode
pin altogether.

Also, what's the feeling about taking the SSU stuff to a private
distribution list & posting periodic updates back to the PIC list?

--BN

----------------------------------------------------------------------------
  Country Robot            "Modular robot components
  69 S. Fremont Ave. # 2     for education and industry"
  Pittsburgh, PA 15202
----------------------------------------------------------------------------

1997\11\20@001315 by Louis Marquette

flavicon
face
> >By placing the transducers inside of pipes or cylinders lined with
> some
> > sound-absorbing substance (like velvet or some other cloth) we
> should be
> able
> > to get good directivity, and with different resistor values in the
> op-amp
> > circuits, we should be able to tweak the gain to accomodate the
> longer
> > distances.  I have been planning on 8 bits of granularity in the
> measurements,
> indoors, presumably due to reflections. The other guy who was the main
>
> designer of the project (who is also on this list, Louis Marquette),
> has
> since tried placing the transducers in cylinders and I think that he
> had
> good success.
>

When I tried to resolve the return signal detected,  I seemed to always
have some signal being detected. Looking back, it seems that i used a 3
transister amplifier, and i tried all sorts of gain configurations. I
think that just using 3 transistors rather than on  with alot of gain,
inserted some noise at each stage of the amp. you might want to try a
comparator to compare the 2 terminals of the transducer, that may work
well, but back to the tubes, I wasnt sure if it was my amplifier
introducing the noise or of it was amplifing small signals made my the
transducer. for all of you out there who say that the transducer has a
narrow bacdwith and so does not need filtering, it is a filter, well,
true enough, but if you tap one, even lightly, it will generate a small
signal, i dont know if this is because tapping it causes it to resonate
at 40khz or if the mechanical stress is transfered to the piezo element
and converted to a voltage, either way it is more or less the same
effect. It will also create a small voltage if ait rushes past it or it
'hears' a loud noise, even if the noise is MUCH less than 40k. I placed
a little conductive foam used for static sensitive ic's(hey, its handy!)
around the transducers outer edges, used that as an insulator of sorts
from a 5cm long tude i put it in, transducer diameter was abt 1 cm, the
tubes was abt 2cm. then i insulated the tube from shock from the surcace
it was mounted on with mre foam. it seemed to work better, but lots of
false detection problems still.oh well thats life! if anyone has any
really good breakthroughd, please let me know, as i would like to
continue but grew frustrated. well, my osciliscope died too, and i'm not
done with my homemade pc based on yet.
Louis Marquette


{Quote hidden}

1997\11\20@111804 by Wayne Foletta

flavicon
face
Louis and other Picsters:

Generally all piezoelectric transducers have many response peaks above
and below the designed "fundamental" operating frequency. Depending on
how it is connected in the circuit non-fundamental responses may be the
highest. That is, coupled with the amplifier bandpass characteristics,
the transducer may in fact produce a very high lower frequency response
as you observed.

From graduate research work I and my colleagues did at Stanford on
medical ultrasound (both CW and pulsed Doppler) in the early 70's we
found:
1. Impedance match transducer to circuit and transmission medium (be it
air or water) for highest sensitivity.
2. Dampen high Q transducers with mechanical or electronic loading to
improve spacial resolution in pulsed mode applications.
3. Limit out of band responses (measure the bandpass of loaded
transducer, impedance matching network and preamp) by adding low pass or
high elements. (It takes real skill in the front end to get just one
response peak with high sensitivity.)
4. Use lowest noise preamp possible with gain of 10x to set noise floor
of system. (You use a lower first stage gain but that puts lower noise
requirements on the following stages.)
5. Isolate TX and RX transducers electrically and acoustically if a two
transducer design or use active damping (and release) and switched
preamp gain for pulsed single transducer designs.
6. Isolate the transducers from the ambient environment except for the
desired acoustic viewing 'window' to the target.

Hope this helps. Have fun.

- Wayne Foletta
BMI - Santa Clara

{Quote hidden}

1997\11\20@145407 by Robert Nansel

flavicon
face
Thur 11/20/97 Wayne Foletta wrote:

>From graduate research work I and my colleagues did at Stanford on
>medical ultrasound (both CW and pulsed Doppler) in the early 70's we
>found:
>1. Impedance match transducer to circuit and transmission medium (be it
>air or water) for highest sensitivity.

This is the big trouble with using piezo transducers in air. Piezos are
"stiff" transducers, that is, they generate large forces but small
displacements; they couple very nicely into like-stiffness media, such as
water or solids, but poorly into a low-density medium like air. Another way
of saying this is that there is a large impedance mismatch between air and
piezo. I've heard some folks have done stuff with impedance matching
layers. One group I read about used a one-quarter wavelength silcone-rubber
matching layer to achieve better coupling into air. They were working at
1-2 MHz, though the principle should be applicable to lower frequencies
with suitable adjustments in dimension and material.

For lower frequencies (below a couple hundred KHz) electrostatic
transducers like Polaroid uses don't produce much force but have much
larger displacements, so they couple into air quite well. They use a foil
membrane as one plate of a capacitor, so it is mechanically much less stiff
than a piezo. They also tend to have a broader band response with no real
sharp resonances. With electrostatic transducers is that they need to be
driven with about 400 V p-p, so a step up transformer is essential.

Even so, it might be worth it use electrostatic transducers and put the
desired bandpass characteristics into the electronics. On the transmit side
we can control the frequency by judicious selection of the PIC's loop
timing and crystal frequency.


>2. Dampen high Q transducers with mechanical or electronic loading to
>improve spacial resolution in pulsed mode applications.

I'd be interested in seeing some specific circuit topologies for doing
this, and I'm sure others would be, too.


>3. Limit out of band responses (measure the bandpass of loaded
>transducer, impedance matching network and preamp) by adding low pass or
>high elements. (It takes real skill in the front end to get just one
>response peak with high sensitivity.)

What sort of numbers are we talking here? What passband would you suggest?


>4. Use lowest noise preamp possible with gain of 10x to set noise floor
>of system. (You use a lower first stage gain but that puts lower noise
>requirements on the following stages.)

Should the preamp be broadband, or would it be better to start filtering
from the start?

Thanks for your ideas!

--BN

----------------------------------------------------------------------------
  Country Robot            "Modular robot components
  69 S. Fremont Ave. # 2     for education and industry"
  Pittsburgh, PA 15202
----------------------------------------------------------------------------

1997\11\20@185544 by Sean Breheny

face picon face
>>4. Use lowest noise preamp possible with gain of 10x to set noise floor
>>of system. (You use a lower first stage gain but that puts lower noise
>>requirements on the following stages.)

Don't you want as much gain in this first amp stage as possible? Each amp
in the chain creates noise, so you want to amplify it as little as
possible, so concentrating your gain in the first stage would be best,
wouldn't it?

Sean


+--------------------------------+
| Sean Breheny                   |
| Amateur Radio Callsign: KA3YXM |
| Electrical Engineering Student |
+--------------------------------+
http://www.people.cornell.edu/pages/shb7
Web Page Under Construction!
shb7EraseMEspamcornell.edu

1997\11\21@105703 by Wayne Foletta

flavicon
face
Sean:

Yes, if cost is important factor alone you would want to set the noise
figure or performance of the system by the first stage. To do that the
first stage power gain must be 3dB or more. So if you've put you money
into a good front-end amp, get as much gain-bandwidth as you can. Get it
while the signal swing is small so the non-linear effects of the
"linear" amps are minimal.

However, since as you point out, other stages contribute noise and
spurious signals if you are forced to use distributed gain (limited by
stage gain-bandwidth, for example). The amplifier second order
characteristics (PSRR, CMRR, THD, etc.), physical layout and supply
bypassing become more important for following stages. So total costs
will go up.

- Wayne Foletta
BMI - Santa Clara

{Quote hidden}

1997\11\21@115524 by Sean Breheny

face picon face
At 10:45 AM 11/21/97 -0500, you wrote:
>Sean:
>
>Yes, if cost is important factor alone you would want to set the noise
>figure or performance of the system by the first stage. To do that the
>first stage power gain must be 3dB or more. So if you've put you money
>into a good front-end amp, get as much gain-bandwidth as you can. Get it
>while the signal swing is small so the non-linear effects of the
>"linear" amps are minimal.
>
>However, since as you point out, other stages contribute noise and
>spurious signals if you are forced to use distributed gain (limited by
>stage gain-bandwidth, for example). The amplifier second order
>characteristics (PSRR, CMRR, THD, etc.), physical layout and supply
>bypassing become more important for following stages. So total costs
>will go up.
>
>- Wayne Foletta
>BMI - Santa Clara
>

Hi Wayne,

Thanks for the answer, but I am still not sure I understand what you are
saying. Lets assume that one has to use distributed gain. Lets say two
identical op amps in a row. They both generate noise of power N and each
can have gain of either 10 or 20 (only two choices to make it easier).
Wouldn't it practically always be better to make the first one 20 and the
second one 10 (assuming you need 200 times gain, not just 100)??!

I assume that in your message you were saying that it might be bad to
concentrate the gain in the first op amp because then the signal going into
the second one would be greater and therefore would have a greater voltage
swing and it would cause the second amp to exhibit non-linearity and
generate higher-order spurious products. The only thing I don't see here is
that we are STILL dealing with a very small signal, I would think that we
could ignore the 2nd order IMD here, especially since we are not designing
a high quality audio system or an RF power amp, but only a pulse
amplification/detection system. If our signal is strong compared to the
noise, THEN it might be strong enough that the 2nd order IMD would be a
bigger problem than the noise, so we would worry more about that. At least,
with my limited amplifier knowledge, this would be my guess.

Sean

+--------------------------------+
| Sean Breheny                   |
| Amateur Radio Callsign: KA3YXM |
| Electrical Engineering Student |
+--------------------------------+
http://www.people.cornell.edu/pages/shb7
Web Page Under Construction!
@spam@shb7spam_OUTspam.....cornell.edu

1997\11\21@125403 by Wayne Foletta

flavicon
face
Sean:

Boy, now we are starting to real get deep into this amplifier stuff. I
hope not too deep for a 'digital' type list.
Answers below:

- Wayne Foletta
BMI - Santa Clara

{Quote hidden}

OK - but what is the gain-bandwith product of the amplifier. In a pulsed
signal application the bandwidth is large because you want to resolve
the timing of the RX pulse envelope with respect to the TX pulse. If
have limited (positive) gain at the highest frequency of interest (based
on the spacial resolving power required) you will have use both amps
with the same gain. If you have more than enough gain-bandwidth in one
amp - then take all with the first one. The large signal artifacts (like
crossover related THD) come mostly from the output stage of amp
(assuming op amps). Noise and other small signal artifacts come mostly
from the input stage or elements. It also depends on if we are talking
about an open-loop amp or closed-loop to make this discussion even more
deep (because some factor get improved with closed-loop - others do
not).

> I assume that in your message you were saying that it might be bad to
> concentrate the gain in the first op amp because then the signal going
> into
> the second one would be greater and therefore would have a greater
> voltage
> swing and it would cause the second amp to exhibit non-linearity and
> generate higher-order spurious products.
>
The most important spurious components originate in the input stage or
elements - period. Assuming identical amps and no clipping or such what
happens in the following stages, the other stages are not as important.
You use them just to get the signal up to a level needed for the
detector - be it just a comparator threshold or A/D with DSP
post-processing.

> The only thing I don't see here is
> that we are STILL dealing with a very small signal, I would think that
> we
> could ignore the 2nd order IMD here, especially since we are not
> designing
> a high quality audio system or an RF power amp, but only a pulse
> amplification/detection system.
>
       A pulse system is more demanding than a high quality audio
system because most people can't detect a 1 dB amplitude change or THD
in the 0.5% range (8 bits).

{Quote hidden}

1997\11\21@171701 by Montaigne, Mike

flavicon
face
This probably won't work but I'm trying to understand:
Why can't we use a horn coupling to the transducer?
My understanding is that is what a horn does, it is a air
transformer which couples compacted air in the throat to the
ordinary air at the mouth of the horn.  Are the frequencies too high
(~40K?) and the wavelengths too short to be useful?

{Quote hidden}

1997\11\22@124348 by John Griessen

flavicon
face
At 05:14 PM 11/21/97 -0500, you wrote:
>This probably won't work but I'm trying to understand:
>Why can't we use a horn coupling to the transducer?
>My understanding is that is what a horn does, it is a air
>transformer which couples compacted air in the throat to the
>ordinary air at the mouth of the horn.  Are the frequencies too high
>(~40K?) and the wavelengths too short to be useful?
>
>>------
Yes and Yes.  A speaker horn can couple a wavetrain, but not a pulse to the
tightness needed for a ultrsound transducer.  Another way of looking at
what a horn does is to think of it as a reflector, then taking into account
some phase reversals and delays, you see that it won't necessarily focus a
single wavefront to all land at the transducer face in a planar wave.  A
plane wave slamming into the trandsucer face is the desired outcome here.
Likewise, a transducer bang is desired to turn into a plane wave at the
horn output.

It's a toughie.  The horn would have to be made with phase changing
coatings, and not let any path go directly in a line of sight to the
transducer face to get the desired outcome.  It might work well if set up
like a reflecting telescope that lets light land at the objective lens
after two bounces.

John Griessen
Austin TX

1997\11\23@001650 by paulh

flavicon
face
My sonar is working on my bench.  It's good up to about 2 feet, which is
all I'll need for my robot.  I suspect I'll be able to improve the range
after I experiment some.

I'm using a PIC16F84 (10 Mhz), an L293D "push pull 4 channel driver", an
LM339 "quad comparator", 2 ultrasonic transducers "251-1603" and a bunch
of resistors.  I suspect I'll need to tweak things a bit before it'll work
reliably on my robot.

The transducers were $6.20 each from Ciruit Specialists.  They came with
copy of the first page of the Mouser Electronics data sheet.  The L293D
came from Ben Wirz, http://www.wirz.com.  The PIC came from DigiKey.  The LM339
came from Radio Shack.  Except for the transducers, the parts were ones
that I had on hand.

The PIC generates a pulse of 16 cycles at 40.3 Khz.  More cycles should
improve the range.  The PIC didn't have enough oomph for the transducers,
so I use half of an L293D motor driver chip (err, transducer driver chip).
It uses a crude 20 volt power supply (2 somewhat used 9 volt batteries and
2 D cells (they were handy)).  There are 3 wires from the PIC to the
L293D, enable, +pulse and -pulse.  The -pulse is just the +pulse inverted.
All the logic circuits are running at 5 volts.

The input transducer is connected to one of the comparators in the 339.  I
use the resistors to bias the circuit so that the 339 is off with no
signal.  One side of the transducer/comparator has a voltage divider made
out of two 2.5Kohm resistors.  So that pin is at 2.5 volts.  There is a
470Kohm resistor across the transducer and a 10Mohm resistor to ground. So
with no signal, the other transducer is about 2.4 volts.  All these
resistor values were based on what resistors were handy.  They probably
aren't close to optimal, they're just the ones that worked first.  I'm
open to better ways of doing the input side.  I suspect a little tuning
will dramatically improve the range.

The PIC only enables the driver chip while sending the pulse.  When not
enabled, the driver chip is "high" impedance.  If it is high enough, I'll
be able to use the same chip to recieve the echo.  The transducer rings
for a bit after sending the pulse, so the minimum range that way would be
6 inches, unless I can dampen the ringing.

On the bench, I've been cheating.  My PIC code doesn't do anything with
the comparator output yet.  I've been looking at the output with an
oscilliscope.

You'll note that a complete circuit only needs 6 PIC I/O pins.  So this
could run on an 8 pin PIC.  The I/O pins are:
    +pulse to driver
    -pulse to driver
    enable to driver
    from comparator
    serial in
    serial out

That is assuming asynch serial I/O.  Since -pulse is an inverted +pulse,
it could be generated with an inverter instead.  This would save a PIC I/O
pin.  There are unused comparators in the 339 that could be used as
inverters.  So a 12C508 could use one of the 3 wire serial protocols, if
needed.

Life's been hectic recently, so I may not spend more time on this project
for a few more weeks.  So don't be surprised if I'm slow responding to
email.  There is more to do to make it work, but the basic concept
certainly works.

--
.....paulhSTOPspamspam@spam@hamjudo.com  http://www.hamjudo.com
The April 97 WebSight magazine describes me as "(presumably) normal".

1997\11\23@160311 by Felix Centeno

flavicon
face
why you dont'n put the circuit squematics in a
WWW page, or sent me to my personal E-mail
please...


Felix Centeno
centellaEraseMEspam@spam@telcel.net.ve
http://www.geocities.com/CapeCanaveral/Hangar/2372/


----------
{Quote hidden}

1997\11\24@160913 by paulb

flavicon
face
Sean Breheny wrote:

> Don't you want as much gain in this first amp stage as possible? Each
> amp in the chain creates noise, so you want to amplify it as little as
> possible, so concentrating your gain in the first stage would be best,
> wouldn't it?

 Oh dear, Sean, you embarrass me!  KA3YXM, what did you study?  The
noise figure is set by the first stage, assuming it has REASONABLE
(20dB or so) gain and the following stages aren't dreadful.  If for
example the NF of the first stage was 0.25µV with a voltage gain of 10
(20dB) only, then its internal noise appears as 2.5µV to the next stage.

 Even if the next stage has 1µV of internal noise, this is about 8dB
down on the amplified version of the first stage noise and will be
barely discernible.  2.5µV noise + 1µV noise = approximately 2.7µV or
so, NOT 3.5µV.

 (Figures typical for VHF radios rather than ultrasonics, by way of
reference.)

 Cheers,
       Paul B.

1997\11\24@164424 by Sean Breheny

face picon face
At 11:56 PM 11/22/97 +1000, you wrote:
{Quote hidden}

Paul,

I guess I was getting confused between which was more important in the
front end, noise figure or gain. I am only learning bit by bit about RF
design,(remember, I am only a 1st year student right now) and I thought
that I have remembered that gain was a big consideration in the front end
>from my Extra class theory. I did the calculations for a simple system like
you suggest, and you are right, it makes only a little difference. However,
if we have two amps, one with say .25 uV internal noise and one with 1 uV
internal noise, WE BETTER put the .25 uV one first. At least, this is what
I saw from my calculations. This makes a big difference in SN ratio. The
same would go for distortion, we want the lowest distortion earliest in the
chain.

Thanks,

Sean


+--------------------------------+
| Sean Breheny                   |
| Amateur Radio Callsign: KA3YXM |
| Electrical Engineering Student |
+--------------------------------+
http://www.people.cornell.edu/pages/shb7
Web Page Under Construction!
spamBeGoneshb7@spam@spamcornell.edu

1997\11\24@235430 by Peter van Hoof

flavicon
face
An additional tip

Noise cancels out(partial) in parallel amps, this works especially well in preamps where you can hang several amps to an input signal and (with series resistors) ty all outputs together

Peter van Hoof


'Serial Sonar Unit (SSU)'
1997\12\03@042339 by Michael J. Ghormley
flavicon
face
Mike Keitz wrote:

>Regardless of all this bickering about amplifiers, examine the big
>picture to see if the application's (sonar ranger attached to robot)
>performance will indeed be limited by amplifier noise.

>Acoustic and electrical interference from the machine and environment
>will likely be predominant sources of noise, even using a preamp of only
>moderate performance.

<SNIP>

Hear! Hear!  (...or is it, Here! Here! I can never remember...)

Mike has hit the nail on the head!  Following this thread has reminded me
of the old saying:

       "A camel is a racehorse designed by a committee."

My admittedly limited experience with the subject, as I stated in a
previous post, was that the Murata transducers worked fine in detecting a
binder to about 20 feet by hitting the Tx with some half-cycle pulses and
detecting them with a simple AGC circuit.  Of course, your mileage may
vary...  d;^)

Michael

1997\12\03@145013 by Steve Smith

picon face
In a message dated 03/12/97  09:24:47, you write:

<<   "A camel is a racehorse designed by a committee." >>

Therefore is a Goat a racehorce designed by Parlement.

Well it might be Steve.....

More... (looser matching)
- Last day of these posts
- In 1997 , 1998 only
- Today
- New search...