Interesting how a spectral distribution would prove how annoying dogs
really are.  I think it's getting downright sick how you can't take a walk
without 15 dogs cursing your every step and you can't sleep at night
without 3 or 4 waking you up.  I don't know why people put up with these
scavengers, barking incessantly...

I'm curious to see exactly what this anti-dog device (program) is of yours.
I have 16 poodles next door that take shifts yapping all day.  I've tried
discipline, I've tried politics, I've tried BB guns, but they keep coming
back to their habitual cussing rituals.  I'm looking for a way (chemically
or electronically) to render them a little more silent when I'm trying to
get some peace.

Any ideas you have are appreciated.

Sincerely,
 

I could not agree with you more!!

If I entered the dog owners' home and disrupted their "quality-of-life" equal to what
they allow their dogs' intrusion to do to mine--I would be promptly jailed!

My SDI device(s) have worked for the most part at reducing the neighbors dog(s) from
annoying the crap out of me, however, there have been instances where it was more
trouble than it was worth.

For the actual dog owners, it works much better due to the proximity of the dog to the
device: the quicker the response the more effective. As you play with these things it
becomes very clear just how slow sound travels.

Three approaches:
 1) a method of detecting (reco) the bark, and squeaking back one half second of
ultrasound (~20 kHz @ ~ 20 W).

2) the second method is to record the actual bark in a analog CCD delay line (~100 msec
FIFO) and upon recognition that the sound is a bark, the device plays back--through a
loudspeaker--that last recorded 100 msec of bark. It really confuses them; they start
shortening the number of their barks, until they often give up. It is funny to
watch/hear their initial confusion!

3) The third approach works well on cold winter nights when you can hear every dog in
town, and so can your neighboring dogs!
It is continuous white noise played through a loudspeaker, toward the offending dog(s).
This seems to "mask" the sounds that trigger the neighbor's dog's barking. Be careful
however, because you could run afoul of your local noise ordinance--especially if your
neighboring dog owner is "gunning" for you!

       glen

  Glen, Im interested in your SDI and how it works..  does it repel/aggravate dogs from a long distance your web page doesn't explain it 
thanks.
 

Thanks for writing. 

As you might guess, it depends on the dog, and what is provoking him at the time--it certainly stops random barking. 

My experience has been for distances not exceeding ~75 feet. 

The problem with long distances is excessive time delay between the bark and the "retribution." 

For long distances, I have had success using a 18" parabolic reflector with a Radio Shack 2 1/2" piezo tweeter (75 Watts peak) at the focus.  Of course, with this kind of directivity, you run the risk of missing the dog, if they have a wide roaming area: it is a tradeoff. 

The health concerns can be ameliorated by placing the system up and away from close access by pets or people. 

I hope this helps.

      glen

I saw your circuit when I was web searching for solutions to my problem.
My neighbor allows their dog to bark at unusual hours, outside, in their yard.
I am interested in the circuit (electronic & commo background).
I have a few questions about it... 
What is the circuit driven with? Is it an amplifier or stand-alone noisemaker?
What type of pot core and winding?
What is the freq range of the circuit (and a dogs ultrasonic hearing range, for that matter)
Thanks for helping!
 

* What is the circuit driven with? Is it an amplifier or stand-alone noisemaker?  --It is "stand
alone" Powered Oscillator/Driver.

* What type of pot core and winding?  --Actually, I used several pot cores stacked, such that
the primaries were driven in parallel and the secondaries were connected in series--all phased
correctly. This was done as a quick & dirty way of achieving a "step up" and allow for enough
power--before core saturation. The specific pot core specs are only important relative to
optimum operating frequency: ~10 kHz to ~40 Khz, peaking at ~ 20 kHz. However, this is not
critical due to the broadband nature of ferrite transformers. One could use alternate
transformer technology: a single large ferrite torrid.

* What is the freq range of the circuit (and a dogs ultrasonic hearing range, for that matter)?
--I have typically used a frequency above the human range ~ 20 kHz. Dogs seem to respond to
almost any frequency up to (and maybe above) 30 kHz. My main intent was to not attract the
owner's attention, but use enough power & directivity to get the animal's attention!

I hope this helps, let me know if you have other questions.

      glen

P.S. I would very much enjoy hearing your results...

I have a barking dog problem (not mine) and was wondering about the effectiveness of this design over a longer range.  I'm thinking a the range can be extended using a barrel to focus the sound waves.  Has this been attempted ?  I'm assuming this circuit would be used in combination with a bark activated microphone to blast them. Any health concerns here other than if you blast yourself ?

Any additional info on the design & fabrication would be appreciated.

A parabolic reflector (~18")  has been used to extend the range or to narrow the usonic sound coverage.

The health concerns can be ameliorated by placing the system up and away from close access.

The only possible problem may be--for long distances--excessive time delay between the bark and the "retribution."

My experience has been for distances not exceeding 50 feet.

I really don't have anymore info other than is on the page.

      glen

Do you have direct experience to suggest the device works?

I have ongoing dog problems surrounding my house, and would do just about anything to stop
them.

I've tried an array of 16 piezo tweeters, driven by a wide-band power amp from a function
generator and have tried a wide variety of waveforms, periods, amplitudes, etc.  None have
had much effect.  I've even mounted the array on a 30-foot pole above the house and measured the output with a B & K 1/4-inch mic at 130 dB at a distance of 10 meters (lots of lobing of course) which seems like if it were going to work, it would have.

What can you tell me about your success that I might have been missing?

Thanks.

The most important aspect of this type of system is: whenever the dog barks he should be immediately given a very loud BURST (~20 kHz) lasting ~ 1/2 second. My experience has been the quicker the response the more likely the dog will see it as a "punishment" meant for him.

If the animal is excited by something or someone, e.g., the garbage man, the less likely the system is to work. It is most effective when dogs are "just barking" for no apparent reason--which seems to be most of the time!

There is a second approach which is very effective with MOST dogs: this is where the system records (stores in a CCD delay line-- ~100 msec FIFO) all sounds while analyzing these sounds for a BARK; when a bark is detected the stored/delayed sound is switched to a P.A. amplifier (for the bark interval the recording mic is muted) toward the dog. 
This so confuses the dog, he barks progressively less and less until he gives up. When it works, it is dramatic!

In all cases, the limiting factor is the speed of sound, or the lack thereof; the farther away the dog, the longer the delay...

I hope this helps, let me know your progress.

      glen

Under Construction in Perpetuity