First Experimental "High Power Electronic Dog Whistle" Built and Operational

[Steve Gibson]

Hello Everyone!!

So where have I been?

As people who have known me and followed my work for many years know, I have so many various irons in the fire that I shift my work-priorities when the associated need-priorities change.  In the case of the resurrection of this infamous "Portable Dog Killer" project (which was renamed to "Portable Sound Blaster" in recognition of the fact that we're no longer 15) my best friend's troubles with the wayward "dog next door" had abated for last year.  So I after pursuing the design of a second-generation device, I switched my attention to other more pressing work without ever needing to push the theory all the way to practice.  But that changed several weeks ago...

I was hanging out at Mark's house a few weeks ago, helping him set up a new Windows 7, 4-channel CableCard Media Center HD recording system... and the yappy little shit was back at it again! He told me that he had recently been awakened at 2:30am, and that it was becoming a problem.

So I picked up where I had left off late last summer and completed the first working model of what developed into an unbelievably loud super-sonic acoustic blaster:

http://www.grc.com/r&d/HPEDW-Front.jpg

http://www.grc.com/r&d/HPEDW-Back.jpg

Since Mark's application need was for more of an "installation", I chose a configuration that is less a "sound gun" and more of a "deterrent outpost".  And since you really don't want to be anywhere near this thing -- especially not in front of it when it's activated -- I included a 500-foot (nominal) range, single-channel, 315 Mhz RF remote control that can be used to either trigger a fixed 4-second blast, or operate the unit in On/Off Toggle mode.

For audio power generation, I used four synchronously-driven high-power handling, piezoelectric horns which will produce phase-coherent wavefronts. But since I didn't want the system to produce a too-focused and directional beam, the emitters being spaced several wavelengths apart will result in a wider dispersion.

The emitters are driven in parallel though through individual LR networks by an RC-snubbed inductive flyback voltage booster which is switched by a single high voltage n-channel depletion-mode MOSFET which is, in turn, driven by a well-bypassed high-current MOSFET gate driver. And all that is driven by a simple pair of 555 timers, one for the carrier and the second as an FM modulator.

The main power MOSFET switches so fully and rapidly between full on to full off that it runs completely cool.  But the flyback's RC snubber is dissipating more power, in heat, than I would like. But without doing much fancier switching (which I might someday do) there wasn't any obvious cleaner solution.

The system is powered at a nominal 12 volts DC, which will usually be provided by an external "wall wart" power supply, and which can also be sourced by internal batteries. (The box contains two, four-"C"-cell battery holders yielding 12 vdc.)

The inductive flyback generates about 36 to 40 peak-to-peak volts across the piezo transducers, which pushes them right up to their peak RMS operating power limit -- and holy crap are they loud!

Mark wanted something that would seriously deter the canine from barking, while being undetectable by the neighboring humans.  So I pitched the carrier to be adjustable from about 10 kHz to 25 kHz.  The FM modulating sub-carrier is more for fun than anything, but if, by chance, the dog isn't sufficiently annoyed by an extremely high power fixed frequency, operating near the upper end of its hearing, adding some rapid FM modulation flutter ought to do the trick. (It sure does annoy me!)

Speaking of which... The result is SO LOUD that the only way I was able to work with the system during final testing and tuning, was to zip up the "lid" of the box, which contains the four emitters, in a pair of heavily foam padded laptop bags, one inside the other.

My feeling is that this might be overkill. One emitter might have been sufficient. But there was really no way of knowing that ahead of time.  Such is the nature of prototypes. One thing my low-parts-count simple flyback voltage booster doesn't offer... is a volume control. It either generates a 40 volt p-p sinewave, or nothing.

But this DOES mean that a single-emitter "mailman's" or "jogger's" hand-blaster would probably be very useful and practical.

Also... for some reason I started out thinking that using a pair of 555 timer/oscillators would be "easier" than programming a little PIC processor to generate a square wave.  But the required resistors and capacitors all add up.  Before you know it you've got bypass caps and coupling caps, and bias resistors... on and on.  So any next iteration design will definitely use a low pin count PIC chip to generate the square wave which, in turn, drives the MOSFET gate driver.

I just finished this a few hours ago, so I don't yet have any real world report on the device's actual field effectiveness. That begins a few hours from now. Hopefully, before long, we will be one dog down. We'll see.

Leo is away from the podcast through Thanksgiving, which means he won't be back until three podcasts from now.  I plan to leave this with Mark until then, then retrieve it for a show-and-tell at the beginning of the first podcast after Leo's return.

For anyone who might be interested in building a one, two, three, four, or more, emitter system, I need to track down where I found those particular emitters. There are plenty of really crappy ones for a couple of dollars (and they look utterly identical) but they are all Chinese rip-offs of the ones I finally located and barely work at all.  However, I found those a year ago... so it'll take some doing to track down their source.

Keep an eye out for the podcast where I'll be talking about this further, and I will report back here about Mark's experience with the device in practice.

Steve.

[errone...@gmail.com]

Really hope this works out well for you and your friend because i wanna build one.  I have the same problem as your friend. These people who live behind me have a dog that just barks all the damn time during the day and even at night.  The owners don't care either. The type of people who have no business owning a dog.  So I'm really hoping this works and you have a good report.  Been looking online at some manufactured devices and they seem as though they're more for a up close and personal training instead of from a distance, which is basically my only option.  Anyway, hope it works out for you friend and you can give a good report.  

[Steve Gibson]

Thanks for your note.

I'll be giving Mark the system once he's back from a business trip this weekend... and I, too, can't wait to learn how it does.  I configured three remote controls for him, so he can position the unit, power it up and leave it running 24/7.  Then position the remotes around his home so that he can trigger it from wherever he happens to be in his house or backyard.

I learned a LOT during the construction of this first "installation oriented" unit, so I want to see whether I can take what I learned and further evolve the solution.

For one thing, this quad-driver solution may well be overkill.  It's painfully loud for humans, so it is likely unnecessarily loud for controlling barking when there isn't any strong motivation for that barking... just a bored canine.  So I'm going to experiment with a single driver to see how loud I can make it.  I hope to further refine the amplifier concept I tried with that first unit to come up with an absolutely minimal parts-count solution which will also be extremely cost-lean.

But the thought that I can't get off my mind is that what a highly-effective bark-stopper needs, is the ability to LISTEN for the barking of the wayward canine and then automatically, and immediately respond the instant his bark has been positively identified as such.  It seems to me that a tight temporal association between the bark and a loud, annoying, and unexpected sound might go a long way toward rapidly training any dog to stop barking if they want the high-pitched sound to stop.

Anyway... I'll be posting more in the coming days... about Mark's experiences manually stopping the barking and about my progress on the next-iteration solution.

/Steve.

[Robert Osthelder]

There are products that listen to a dogs bark and then emit a hi-frequency sound to deter the dog from barking. I wonder if one if those devices could be easily modified to work with Steve's device. Might be worth a shot. Only problem that device might have is properly picking up the barking from a distance.

I live in a apartment and the dog below me barks constantly whenever it is home alone. Drives me crazy. I wonder if a device like Steve's would penetrate through the floor? I want to build one of these but I'm going to hold off until Steve reports back.

A big thanks to Steve for sharing the build and experiences with us.

Robert

[Steve Gibson]

Hi Robert....

A few points:

All of the reports I've heard -- generically online and also here in this group -- about devices that attempt to detect and respond to a dog's barking, suffer from apparently insurmountable discrimination problems. Either their "sensitivity" is turned down so low (to prevent false positives) that they fail to trigger when they should... or in order to get them to trigger reliably when a dog barks, their "sensitivity" needs to be cranked so far up that they are also responding to all sorts of extraneous environmental noises.

Any low-tech cheap and cheesy solution will rely upon too few parameters for making its determination.

But... it is my feeling that a highly reliable bark detection system, with automatic and instantaneous response to VALID barking, is THE optimal solution for this system. Since this 3rd-generation experimental device will incorporate a microprocessor, I'm already planning to give-a-to to developing a highly accurate dog bark characterization and detection system.

--//--

Next...

As you know, this system uses a high-power and high-frequency transducer to produce a brief sharp "training tone" which will hopefully quickly curtail any dog's boredom-based barking. The bad news is that such high frequency is VERY efficiency blocked by anything standing in the path between the emitter and the receiver (the doggie). LOW frequencies are readily passed by turning the blocking object into a sympathetic re-emitter. We've all experienced the sound of a loud audio system on the other side of a wall. What we hear is predominately the "boom boom boom" of the lower bass frequencies because the intervening wall, which acts as a very effective low-pass filter, blocks the "highs" and transmits the "low" by itself vibrating sympathetically. But the wall has too much mass and inertia to vibrate at a high frequency.

Consequently...

What I would recommend would be that you explain the problem to the dog's owners and offer to provide THEM (on loan) with "The Bark Stops Here" device. That way, the device can be in proximity to the barker.

I also have plans to allow the device to perform logging and reporting. So it COULD be put into a passive bark monitoring mode in order to develop a case about when, how many times, how often a dog is barking. And... if it's running in active mode it will/can record the same data to (hopefully dramatically) demonstrate the training effect of the device's active mode.

/Steve.

[bouwe...@gmail.com]

Hi, Steve long time SN listener, 

just watched sn 11/28/12 and enjoyed the parts about the pdk, and portable one, 

I too have a problem with barking dogs in the neighbors yard, so I am very excited this has been brought to life again. 

cant wait for some updates, 

thanks

brett. 

[Steve Gibson]

Hi Brett!

I'm really glad too.  Even though I have a million unfinished projects, I AM always trying to move various loose ends forward.  I WILL be looking for and soliciting people as "beta testers" to receive a free "hand blaster" in return for reporting their experiences... and short videos would be wonderful too, when possible.

I expect to be demonstrating a working finished prototype unit next week... and to then organizing a beta testing / "beta needers" group.  :)

[Kyle Smith]

Hi Steve,

I have purchased and used your products for many years. You are an extraordinary individual. Because of my admiration for your abilities, I feel compelled to mention that such a device could accelerate the natural high frequency hearing loss associated with physiologic aging in humans. Just because a human cant hear the frequency does not mean that it is benign. You ought to get the professional advise of an ENT physician, before exposing your selves and other bystanders to very high SPL's at any frequency. I don't claim to know what an ENT would say on the matter, however, hearing aides can only amplify high frequencies so much before they begin to negatively effect the patients high frequency hearing loss. And you certainly would want to know if the device could be used safely in the setting of an infant, who may be located close enough to receive a damaging duration of high frequency sound at excessive SPL.

Better to know this stuff, before you deploy it to the uneducated masses. If a competent ENT says its perfectly safe, it won't have cost you much to have checked. However, if they say that it might cause harm, you will certainly be paying much more for some attorney to try and defend you, sooner or later, down the road.

Best wishes and good health to you and your family,

--
Kyle Smith, MS, MD (retired)
air...@me.com

[Noel]

I think there may be an application outside of annoying dogs.  My dog hardly barks and behaves well for the most part except for one annoying thing.  When he wants to come in the house he jumps and scratches at our sliding glass door.  We have been trying for a while to teach him how that doesnt not get him into the house, but he has a hard time figuring this out.  PSB may be the way to show him what not to do.

[Mark Pritchard]

Am I the only one who thinks Steve's prototypes look more polished than some "finished" products I've seen on the shelf? Man, I love good quality engineering. 

I know you have promised parts lists for a more compact and portable unit in a recent episode of SN. Please also post the parts list for the 'Big Boy' pictured in this post. I live in a rural area (out in the woods) and I'll need the full power to gain the same effect on the critters that sit out by the fence line. I've tried some of the current commercial offerings and found them to be too anemic to be effective. 

Thanks for all ya do,

Mark Pritchard

[Dan Stange]

Steve,

             I can not wait for the parts list! My neighbor has a massive Bull Mastiff that will not stop barking and attacking my stockade fence. I really need to teach this dinosaur a lesson! I also know a few of my friends that I will be assembling units for as well.

Thanks,

Dan

[elloc...@gmail.com]

Steve,

This is exciting stuff!  I just wanted to add my voice to the crowd that is pleading for plans for a portable version.

~Gus

[Steve Gibson]

Kyle,

Thanks very much for your well placed thoughts and concerns.  As you might expect, I have researched and considered this VERY important question thoroughly.

And you are 100% correct: Just because WE can't readily hear the sound DOESN'T mean that it's not still loud and potentially dangerous and damaging to us.

And, in fact, acoustic radiation is very much like other forms of radiation: Either an exposure to a large amount in a short time, or a gradual accumulation a lower level over a longer time, set against our body's slowly decelerating rate of repair as we age... can result in long-term damage to our body's delicate systems. (There's a reason the dental tech always leaves the room when our jaws are being x-rayed... and also a reason a lead apron is draped over our gonads too! :)

I absolutely intend to be VERY CLEAR and tutorial -- and to frighten people -- about the very real dangers of chronic exposure to high levels of sound -- both audible and inaudible -- and to also explain the effects of sound level and distance.

The devices which people will be able to manufacture from this work will NOT BE TOYS.  That's why I believe they will be effective... but with that effectiveness, because of that effectiveness, comes responsibility for its usage.  So the resulting devices need to be treated with the same respect as a knife or a loaded gun. The batteries should be removed when it is not being actively used.  It should be off limits to the kids and kept out of the hands of children.  And I'll also be providing graphics for the creation of clear warning labels so that an understanding of the device's nature is transferred to, and appreciated by, anyone else who ever uses it if ownership ever changes.

There was a LOT that I did not have time to discuss during yesterday's podcast (and it wasn't, after all, a "sonic blaster" podcast).  But there are several reasons I chose to use a microprocessor as the system's "oscillator" rather than a much simpler (though, perversely, not much less expensive) analog timer chip:  For one thing, I plan to generate a MUCH more "annoying" sound, by pseudo-randomly varying the positioning of the rising and falling edges of the output square wave.  This will create more of a low-ultrasonic "screech" than a pure sine-wave "tone" by producing a very harmonic rich sound.

And secondly, I plan to incorporate a built-in timeout so that it's NOT POSSIBLE to "tape down" the button of the device in order to run it continuously.  If someone wishes to circumvent those safety measures the responsibility will be theirs.  But I cannot see any typical need for ever generating more than about half a second of sound from the device.  (When my buddy Mark made the neighbor's yappy little dog jump three feet into the side of the house, he had just given it the barest of a "tap".)  The goal is NEVER to "hurt" anything in any way... ONLY to startle and train.  And for that a brief "blip" is probably even more effective than a protracted event.

And lest we get TOO carried away with all this, we should remember that all this is, is a generic audio system piezoelectric tweeter connected to an amplifier and audio oscillator.  It's really not so exotic.  Mostly I've just made such a system convenient, portable, and battery operated.

But that said, I want to and plan to GO OVERBOARD with being as responsible as I can possibly be.  As you originally pointed out, just because we can't readily hear it doesn't mean it's NOT loud and potentially damaging.  In that regard, this device DOES differ from a gun or a knife in that common understanding tells us about the inherent dangers and need to be responsible with those well-known weapons.

I need to similarly educate those who wish to experiment with, and employ, high-power high-frequency sound generation equipment.

(And I feel strongly about clearly labeling the device so that when it inevitably falls into the hands of others that understanding will survive and be communicated.)

/Steve.

[Robert Huttinger]

Sign me up for beta testing. Ill take thorough notes and report usage and findings! We didnt spend summer outside because of thr barkung dogs. Their friendly, just not trained.

If I cant test, please let us know when there is posted plabs and a parts list.

PS just finished reading (yea reading) Daemon on audio.... So good, and scarily plausible.

Cheers!
Bo

[Jim Jensen]

Hi Steve,

After listening to you and Leo earlier this week I ordered a "MSP-EXP430G2" from TI. I'll be watching, listening and/or reading as the portable version progresses. I don't know how you find the time for all the irons in the fire.

Regards,

Jim

[hausm...@gmail.com]

Steve, I am really interested in this.  We have two sheep dogs and a lab next door.    I would love to build and or test one out.  

Dan

[Kindanyume]

Same here.... but one "hunting dog" that the owner is a f'ing tard
about.. leaving it out constantly in a tiny "dog run" (not a proper
one) while everyone ignores him but the neighbors getting po'd/

I feel sorry for that poor dog :(

[Steve Gibson]

I'm GLAD you've got one on the way, Jim.

Since with the free shipping I expect that it takes a while for them to arrive.

Also... since EVERYONE who is going to want to build one of these gizmos, or just generally experiment with the project, is GOING to need one, and since they are $4.30 out the door (or in the mailbox), the sooner they arrive, the better.   :)

I am VERY pleased with this choice of processor.  It's a Win Win Win!

[Steve Gibson]

If you want to poke around at MSP430 code, I have just finished polishing a first little utility for the "Dog Whistle" project.  It's NOT anything that most people will need, since the LaunchPad comes with two nice-sized chips.  But I wanted to explore building the lower-end Dog Whistle -- which, for example, only generates sound upon trigger press and can't listen and train on its own.  (I'm calling THAT possible future project a "Dog Sitter".)  The lower end device is so simple that the smallest and cheapest of the MSP430's, the G2001, can do the job.  But it lacks the factory calibration of its built-in oscillator which the larger chips -- contained in the LaunchPad kit -- contain.

So I wrote a "calibrator" for the low-end chip.

It will run on any of the MSP430's, but there's little point since the target calibration frequency is 12 Mhz and that's one of the standard frequencies for any of the larger MSP430's.  However... IF someone wanted to run an MSP430 at some other frequency than 1, 8, 12, or 16 Mhz (all which are built into the larger chips), then this slick bit of code is the nicest way to do that.  :)   It's here:

http://www.GRC.com/acoustics/DCO_Calibration.asm

http://www.GRC.com/acoustics/Small_Flash_DCO_Calibrator.zip

And it's all pretty well commented and clear.

On Friday, November 30, 2012 8:01:51 AM UTC-8, Jim Jensen wrote:

[Monty Marvian]

Steve, 

  I am a letter carrier and it would be comical to see how many dogs I encounter all day long. At home I have a Black lab to witch it never barks after having experienced all the dogs set out with no guidance. I have been dreaming of building This the day I heard of  it. I too have WASTED a small fortune on dog deterrent devices with Zero success. Thank you for all the effort you have put into this device. I hope to build, and even maybe be a tester of your handheld (PDK). 

Captivated Monty Marvian Sandy, UT 

[Robert Foster]

Does the use of this MSP430 mean that the LPC 1769 (Expresso) previously proposed will not be used?

Thanks in advance.

Bob Foster

Sent from my iPad

[steveho...@gmail.com]

Steve,

   The idea intrigues me as a solution for my neighbors problem. He's no tech guy and being retired is home most of the day dealing with the dog barking across the street all day. I'd like to give the device to him as a gift, but you mentioned earlier that walls would block it. My question is would glass also be enough to dissipate its effect like a wall would? If it would, would it be possible to tune it to a low frequency and still retain its effectiveness? I understand it would unfortunately cause you to lose the covert operation benefit.

[steveho...@gmail.com]

Also, if it would not be too much trouble do you think you could post a wiring diagram. I'm an auto mechanic by trade and these systems are a lot easier for me to visualize as schematics than spoken words.

[caitsi...@gmail.com]

Totally impressed by how cheap the MSP430 development tools are.  Got one on the way, as I too, would like to build one of these.  I have lost count on the number of times I have taken a walk through my local community, and had dogs randomly decide to bark and some of them also give chase.  These dogs need some training.

For a period of time, I had been dealing with atmel AVR microcontrollers.  Of course, the tools to do live debugging have not been all that cheap, and chasing down bugs in the code, is not so fun, without live debugging.

[Martin Jarvis]

What I'd like to understand is how do you test something like this if you can't hear it? My concern is that the test equipment would be more expensive than the build components themselves. Any suggestions, Steve?

Martin (v. long time listener)

[Dan Stange]

I too ordered a pair of the MSP430s, what are the other components we would need to build the portable version and code?

[Kindanyume]

I did the same.. one for myself (neighbors dog is as I posted before
sadly neglected and untrained. My own dog only "barks" in his sleep..
it's hilarious watching a whippet "run" and attack while out cold.)
The second is for a friend in St.Kits that has problem neighbor dogs
as well. It's too bad there is no test required before someone can
"own" a pet. :(

[Steve Plegge]

Perhaps the PDK can be modified to help "train" humans...

Woof

On 12/01/2012 09:34 AM, Kindanyume wrote:
> I did the same.. one for myself (neighbors dog is as I posted before
> sadly neglected and untrained. My own dog only "barks" in his sleep..
> it's hilarious watching a whippet "run" and attack while out cold.)
> The second is for a friend in St.Kits that has problem neighbor dogs
> as well. It's too bad there is no test required before someone can
> "own" a pet. :(

--
On the Internet, no one knows you're a dog. [Dogs Rule!]

[Steve Gibson]

Hi Dan,

As always, what interests me is not finding "an answer" but rather "the answer."  In this case, it means that there's some iteration required as I move back and forth between the power amplifier, the choice of optimum tweeter, the microcontroller, and the power supply.

I knew I was on the trail of "the answer" when I was able to turn 4vdc into a 60 volt peak-to-peak low-harmonic sine wave across an inexpensive piezo tweeter with only about 200mA of average supply current... and the transistor staying completely cool.

That design uses just a single power MOSFET transistor, two small 1mH inductors, and one each .01uf and 10uf capacitors.  Five components!

The first inductor is in series with the MOSFET, operating as an inductive flyback and taking advantage of the MOSFET's inherent drain-to-source diode to eliminate the need for that separate component. The .01uf capacitor shunts the transistor, allowing it to switch at a time of either zero voltage or zero current across it, thus hugely boosting the system's overall efficiency while allowing the use of a less expensive MOSFET.

That first little sub-system generates a high-voltage synchronous "kick" which pumps the second sub-system, consisting of a tuned LRC oscillator composed of the second identical inductor in series with the piezo tweeter, which appears to the circuit as a 0.13uf capacitor for the LRC oscillator. I avoided the need for a physical load resistor by carefully choosing the resistance of the inductor so that it can do double-duty and eliminate another component.

Finally, since the high-voltage sine wave appearing across the piezo tweeter had a positive DC bias of several volts, and since piezo transducers can have problems over the long term when subjected to DC, I added a 10uf series capacitor to block the DC.  Since the tweeter's effective capacitance is so much lower than 10uf, we split very little power across the larger capacitor.

Since I wanted to work on, and perfect, one isolated portion of the circuit and system at a time, I wanted to initially have a near ideal switch for a transistor so that I wouldn't need to worry about it while working out the details of the downstream components. So I chose an oversize (and over-expensive) power MOSFET for that role. But since oversize power MOSFETs also have oversize gate capacitances, I also used a separate discrete MOSFET driver to briefly source and sink the large spikes of gate current required to rapidly charge and discharge its gate in order to get it rapidly switched on and off.

So where I am now... is just about ready to use the microcontroller's own output signal to directly drive the gate of a properly sized power MOSFET.  Once I have chosen and verified that choice, the electrical design should be complete.

I have designed the entire system for easy assembly (a microcontroller in a DIP package), ultra-low component count, and very low cost (<$10)... while sacrificing NOTHING in final performance or battery life.

I'll have a complete set of schematics and code and Digikey parts list posted once I have everything nailed down. I'll get the hardware posted before the firmware so that people can order the parts and get them on their way.  :)

/Steve.

[Steve Gibson]

Hi Robert,

You asked:

Does the use of this MSP430 mean that the LPC 1769 (Expresso) previously proposed will not be used?

Thanks in advance.

Yes. I feel badly that people grabbed those, but at least the platform for this generation is only $4.30.

That said, it would be TOTALLY POSSIBLE to still use the LPC1769 Expresso for this project -- to drive the output power amplifier / power converter... though calling it "overkill" for square-wave generation would be quite an understatement.  :)

/Steve.

[Steve Gibson]

Steve, you wrote:

Also, if it would not be too much trouble do you think you could post a wiring diagram. I'm an auto mechanic by trade and these systems are a lot easier for me to visualize as schematics than spoken words.

Rest assured that EVERYTHING will be made available just as soon as possible.  This has my full focus and attention now, and I want to get it published ASAP.

I DO expect that I'll be producing a circuit board too, and making them freely available to anyone who wants one.  I'm doing this because I will likely be building a few dozen of the manual Dog Whistles in order to give them to people, who are not electronics builders, to test.  It's easy to manually wire up one simple circuit, but it's much more trouble to do so for 50.  So I'll be creating a PCB for myself... with plenty of extras for anyone who wants one.

/Steve.

[k...@kensmellsgood.com]

I'm pretty excited for this.  My neighbors have two dogs.  One of which is huge and barks constantly.  The other one is smaller and it's bark can be described more as a long howl than anything.  I work, go to school, and have a family so my quiet time is very valuable to me.  I've heard these dogs as late as 1 AM and as early as 5 AM.  I need this product!  Do you have any idea on what the price range will be?  Also if you need anybody to test these things out then I would be happy to.  I was actually going to buy a BB gun, so my neighbor's dogs thank you in advance.

[Kindanyume]

Sign me up! :)

[Kindanyume]

A BB gun? I'm sorry.. yes dogs can be annoying but I seriously hope
you were joking.

Years ago some total moron (for lack of a more polite term.. there are
no lack of impolite terms for this of course) used a BB gun and shot
my doberman in the side. She wouldn't hurt a fly, the 3lb cat "beat
her up" and never barked... even at intruders (she was am 80lb suck to
say the least.)

But if I had not caught the fact she had been shot it could have
caused major problems if not killed her.

NOT cool.

(And I'm very much pro gun btw.. if used in an intelligent manner.
If anyone absolutely MUST use a BB/Pellet gun for such.. use a
spitball not a "proper" projectile!.)

Hence why SG's invention is so important so as to accomplish the goal
w/o causing any real harm.

[tz]

It would help to know which tweeter you are using so I might be able to take advantage of cyber-monday belated deals.

You are sending a 60% approx duty cycle out but getting a sine wave at the tweeter?  I know the harmonics would be cut, but if it is a sine, why not 50%

I use various flavors of Atmel ATmegas and ATtinys, so I will probably do one of those since I have tubes of them (including with serial port and matching bluetooth modules).

I hope the gerbers or even the CAD files would be available at some place like BatchPCB.

[Kindanyume]

LMAO! If only.. then I could train both my brother and brother in law! :)

[Steve Gibson]

TZ:

Replying to you questions in-line...

It would help to know which tweeter you are using so I might be able to take advantage of cyber-monday belated deals.

The topic of the best piezo tweeter will be getting a permanent page of its own.

I still have some experimentation to do to determine whether a $2 Chinese knock-off of the famous Motorola/CTS KSN-1005 Piezo Tweeter performs as well as the $19 original.  But if you Google "KSN-1005", and poke around a bit, you'll quickly learn a lot.  Motorola stopped producing the tweeter a long time ago, after which they sold the injection molds to CTS -- a well-known piezo company -- who continues to produce the tweeter today.  You can tell the "good" tweeters from the Chinese knock-offs because they say "Made in the Philippines" on the back if they are the more expensive ones/

However... the concept and structure of the tweeter is SO SIMPLE -- a floating piezo 'sounder' glued to the back of a rigid cone, which then uses its own inertia as it flexes back and forth to drive the cone forward and backward -- that it's NOT clear that the Chinese $2 knockoffs are any less useful.  I've been buying them from various sources, most recently from Amazon, as the Pyramid TW105:

http://www.amazon.com/Pyramid-TW105-3-75-Inch-Watts-Tweeter/dp/B0013CFOBY

And Pyle also has two closely related offerings:
http://www.amazon.com/Pyle-PSN1167-Dispersion-Piezo-Tweeter/dp/B000EFMP7W
http://www.amazon.com/Pyle-PSN1165-Square-Dispersion-Tweeter/dp/B0002DNZR0

The cool thing about these is that none are expensive and we'll be able to subject them to a great deal of experimentation.  They are virtually indestructible as long as they are not hugely overdriven.

My favorite source of the "real" (non-Chinese) CTS versions of these speakers is US Speaker:
http://www.usspeaker.com/cts%20index-1.htm

You'll see the KSN-1005 there rated at 100 watts RMS... but also check out its big brother, the KSN-1165, shown rated at 400 watts RMS!  The circuit I have designed is fully capable of driving the bigger bro. The only thing holding me back is the much greater cost.

Note, also, that my power amplifier achieves its super low component count by treating the piezo speaker AS the capacitor in a rather tightly tuned LRC resonator. Since Big Bro has a characteristic capacitance of 0.30uf -- as opposed to the 0.13uf capacitance, the series inductor would need to be reduced to 470uH from 1mH in order to get the same 15kHz resonance.  AND... that might REALLY be worth looking in to!  :)

 

You are sending a 60% approx duty cycle out but getting a sine wave at the tweeter?  I know the harmonics would be cut, but if it is a sine, why not 50%.

Ah. That's because the square wave is not driving the LRC resonator directly. The square wave drives the gate of a power MOSFET which, in turn, drives a primary inductor that's setup as a flyback to generate the high voltage which, in turn, drives the LRC resonator containing the second inductor and the piezo.  (I described the amplifier more completely in a post earlier today.)

 

I use various flavors of Atmel ATmegas and ATtinys, so I will probably do one of those since I have tubes of them (including with serial port and matching bluetooth modules).

Yes, absolutely, you should use whatever you are most comfortable with and/or may already have on hand.  I've chosen the MSP430 family because I'll be coding it in assembler, because it has a very nice 16-bit 16-register fully orthogonal Von Neumann instruction set and architecture, and because in the possible future "Dog Sitter" automatic dog bark response aversion trainer, I love the incredibly low-power features of the chip.

 

I hope the gerbers or even the CAD files would be available at some place like BatchPCB.

Frankly, the circuit is SO SIMPLE, that unless you're wanting to make many (as I expect I will be) you won't need a PCB.  But I WILL be using DipTrace for schematic capture and PCB design... and everything I create will be freely available.

/Steve.

[Steve Gibson]

Gang...

The final pieces were proved today.  Attached is a photo I just took of the breadboard.  That's ALL there is to it!  You give it 6vdc from four series-connected AA cells and this simple and inexpensive little circuit places a lovely 65-volt peak-to-peak sine wave across the piezo tweeter.  And... oh my lord it's loud.  And that's at a frequency that we can barely hear at all.

I'll get a schematic drawn up with a complete bill of materials from www.digikey.com, as well as the source code for the first simple little oscillator.

/Steve.

---

1st_Finished_Dog_Whistle_Electronics.jpg

[Kindanyume]

kickass... I await access to the pic and more info of course..  as will be many others of course! :)

[Kindanyume]

given it's "portable" maybe an even smaller model could be used.

or given the size this one is dirt cheap"

http://www.amazon.com/MCM-Audio-Select-53-820-TWEETER/dp/B008I5R584/ref=sr_1_3?s=electronics&ie=UTF8&qid=1354429134&sr=1-3&keywords=2%22+piezo+tweeter


On Sat, Dec 1, 2012 at 4:50 PM, Steve Gibson <steve....@gmail.com> wrote:
> piezo tweeter

[Stephen Syputa]

Greetings All! 

I placed the TI order on the Nov 30th and it arrived on Saturday Dec 1st, FEDEX Express, with no shipping charge. I suspect that the shipping charge alone outweighed the actual cost for two boards!

I spent my time in the US Navy as a sonar technician. All of our sonars us phased-array piezo-electric transducers to form 2kHz to 18kHz sonic beams in any direction at selectable intensities. Of course, the medium was water where sound behaves quite differently than in air. The transducers are also the receivers..Phasing transducer arrays to form beams is fundamentally a math problem based on physical position, frequency and local speed of sound (which is a variable based on local temperature and density of the medium). I can envision a microprocessor with temperature and humidity sensors regulating the parameters, if necessary. 

Has anyone here explored the practicality of using a phased sonic array in air? 

[Tom Robinson]

So where , when do we see parts list ... software ? I have a bit of experience with Arduino ... I imagine it too would be able to drive the amp. I can build stuff ... Jamco. Allide, etc.  We have terrrible problem with Ravens taking / wrecking all our fruit before its ripe! I don't have the heart to kill and don't desire to own a gun. I'd love to build prototype with single tweeter. If there was a kit available I'd buy just to expidite. 

anxiously awaiting.

On Monday, November 12, 2012 8:58:33 AM UTC-8, Steve Gibson wrote:

Hello Everyone!!

So where have I been?

As people who have known me and followed my work for many years know, I have so many various irons in the fire that I shift my work-priorities when the associated need-priorities change.  In the case of the resurrection of this infamous "Portable Dog Killer" project (which was renamed to "Portable Sound Blaster" in recognition of the fact that we're no longer 15) my best friend's troubles with the wayward "dog next door" had abated for last year.  So I after pursuing the design of a second-generation device, I switched my attention to other more pressing work without ever needing to push the theory all the way to practice.  But that changed several weeks ago...

I was hanging out at Mark's house a few weeks ago, helping him set up a new Windows 7, 4-channel CableCard Media Center HD recording system... and the yappy little shit was back at it again! He told me that he had recently been awakened at 2:30am, and that it was becoming a problem.

So I picked up where I had left off late last summer and completed the first working model of what developed into an unbelievably loud super-sonic acoustic blaster:

http://www.grc.com/r&d/HPEDW-Front.jpg

http://www.grc.com/r&d/HPEDW-Back.jpg

Since Mark's application need was for more of an "installation", I chose a configuration that is less a "sound gun" and more of a "deterrent outpost".  And since you really don't want to be anywhere near this thing -- especially not in front of it when it's activated -- I included a 500-foot (nominal) range, single-channel, 315 Mhz RF remote control that can be used to either trigger a fixed 4-second blast, or operate the unit in On/Off Toggle mode.

For audio power generation, I used four synchronously-driven high-power handling, piezoelectric horns which will produce phase-coherent wavefronts. But since I didn't want the system to produce a too-focused and directional beam, the emitters being spaced several wavelengths apart will result in a wider dispersion.

The emitters are driven in parallel though through individual LR networks by an RC-snubbed inductive flyback voltage booster which is switched by a single high voltage n-channel depletion-mode MOSFET which is, in turn, driven by a well-bypassed high-current MOSFET gate driver. And all that is driven by a simple pair of 555 timers, one for the carrier and the second as an FM modulator.

The main power MOSFET switches so fully and rapidly between full on to full off that it runs completely cool.  But the flyback's RC snubber is dissipating more power, in heat, than I would like. But without doing much fancier switching (which I might someday do) there wasn't any obvious cleaner solution.

The system is powered at a nominal 12 volts DC, which will usually be provided by an external "wall wart" power supply, and which can also be sourced by internal batteries. (The box contains two, four-"C"-cell battery holders yielding 12 vdc.)

The inductive flyback generates about 36 to 40 peak-to-peak volts across the piezo transducers, which pushes them right up to their peak RMS operating power limit -- and holy crap are they loud!

Mark wanted something that would seriously deter the canine from barking, while being undetectable by the neighboring humans.  So I pitched the carrier to be adjustable from about 10 kHz to 25 kHz.  The FM modulating sub-carrier is more for fun than anything, but if, by chance, the dog isn't sufficiently annoyed by an extremely high power fixed frequency, operating near the upper end of its hearing, adding some rapid FM modulation flutter ought to do the trick. (It sure does annoy me!)

Speaking of which... The result is SO LOUD that the only way I was able to work with the system during final testing and tuning, was to zip up the "lid" of the box, which contains the four emitters, in a pair of heavily foam padded laptop bags, one inside the other.

My feeling is that this might be overkill. One emitter might have been sufficient. But there was really no way of knowing that ahead of time.  Such is the nature of prototypes. One thing my low-parts-count simple flyback voltage booster doesn't offer... is a volume control. It either generates a 40 volt p-p sinewave, or nothing.

But this DOES mean that a single-emitter "mailman's" or "jogger's" hand-blaster would probably be very useful and practical.

Also... for some reason I started out thinking that using a pair of 555 timer/oscillators would be "easier" than programming a little PIC processor to generate a square wave.  But the required resistors and capacitors all add up.  Before you know it you've got bypass caps and coupling caps, and bias resistors... on and on.  So any next iteration design will definitely use a low pin count PIC chip to generate the square wave which, in turn, drives the MOSFET gate driver.

I just finished this a few hours ago, so I don't yet have any real world report on the device's actual field effectiveness. That begins a few hours from now. Hopefully, before long, we will be one dog down. We'll see.

Leo is away from the podcast through Thanksgiving, which means he won't be back until three podcasts from now.  I plan to leave this with Mark until then, then retrieve it for a show-and-tell at the beginning of the first podcast after Leo's return.

For anyone who might be interested in building a one, two, three, four, or more, emitter system, I need to track down where I found those particular emitters. There are plenty of really crappy ones for a couple of dollars (and they look utterly identical) but they are all Chinese rip-offs of the ones I finally located and barely work at all.  However, I found those a year ago... so it'll take some doing to track down their source.

Keep an eye out for the podcast where I'll be talking about this further, and I will report back here about Mark's experience with the device in practice.

Steve.

[Steve Gibson]

Tom...

You wrote:

So where , when do we see parts list ... software ? I have a bit of experience with Arduino ... I imagine it too would be able to drive the amp. I can build stuff ... Jamco. Allide, etc.  We have terrrible problem with Ravens taking / wrecking all our fruit before its ripe! I don't have the heart to kill and don't desire to own a gun. I'd love to build prototype with single tweeter. If there was a kit available I'd buy just to expidite. 

anxiously awaiting.

I'm working on the project full time.  Last night I was playing with battery size and power amp input voltage, and may be switching everything from 6vdc to 9vdc.  A "transistor radio" battery has much less total capacity than a set of four AA's, but it appears that I'm able to get enough additional power out that it would be worthwhile, and the operating efficiency of the system is close to 100%, so we're not wasting anything.  Also, a single 9vdc battery is enough smaller to allow the use of a smaller case, which is a big convenience.

But since the MSP430's can only run on 1.8v to 3.6v, I was planning to center-tap four AA's to get a nice "free" 3vdc for the processor.  If I switch over to a single, closed, 9vdc battery, I'll need a little fixed-output three-pin regulator.  That would annoy me, but this morning I definitely eliminated two of the existing components and possibly a third...

My point is... the design is NEARLY, but not yet, fully complete.  It'll happen when it happens.  The whole reason I do these things is because it's unknown and unexplored territory, which is what makes it interesting and useful for me.  But a side effect of that is that I have no idea when anything will be "finished".

So, without wanting to be a smart-ass... it'll be ready when it is... but that won't be long from now... and it'll be SERIOUSLY tight and wonderful once it's done.  :)

Note, also, that birds DO NOT HAVE the same acute high-frequency hearing that canines possess.  You can try it as is (and we'd all love to know how it works).  But it might be necessary for you to drop the frequency by half, to 7.5 KHz for it to be readily audible by Ravens.  That would be a trivial change to the processor's code, and you could just double the value of the series-connected inductor from 1mH to 2mH.

More soon!

/Steve.

[hausm...@gmail.com]

Brushed up on my project skills, just put together a minty boost with my son.    Really psyched about this PDK.   I would probably be willing to donate some time a build a few for people that are not into making stuff.

Dan

[Steve Gibson]

Stephen...

You wrote:

Greetings All! 

I placed the TI order on the Nov 30th and it arrived on Saturday Dec 1st, FEDEX Express, with no shipping charge. I suspect that the shipping charge alone outweighed the actual cost for two boards!

I spent my time in the US Navy as a sonar technician. All of our sonars us phased-array piezo-electric transducers to form 2kHz to 18kHz sonic beams in any direction at selectable intensities. Of course, the medium was water where sound behaves quite differently than in air. The transducers are also the receivers..Phasing transducer arrays to form beams is fundamentally a math problem based on physical position, frequency and local speed of sound (which is a variable based on local temperature and density of the medium). I can envision a microprocessor with temperature and humidity sensors regulating the parameters, if necessary. 

Has anyone here explored the practicality of using a phased sonic array in air?

Yes.  It's DEFINITELY possible.  There are commercial devices (Google "Sound Spotlight") and hobby projects (kickstarter has such a project.)  My goal for this round is to explore a "just get the job done" solution that delivers the very most bang for the buck.  My goal is to create a baseline design and get it, and the resulting device, into the hands of people in need... so that we can evaluate its short- and long-term utility.  Since every component wired increases both the cost and the construction difficulty, and decreases the reliability... I am shooting for a highly economical design.  And... it's looking good!

And thanks for the news of how TI handles orders.  VERY nice!

/Steve.

[Tom Robinson]

Wow:

You response is fast. ( I don't tweet much ) ... found Google Project site not very active.

 What's the best feed to keep informed ?  I'm concerned about freeking out my pets and farm animals at large. Tweeter focus ? 

I understand the short burst notion.

Can you point to article about sound projection via harmonic reflection off wall ? 

I have listened to all security now podcasts ... thank you so much, have learned soooo much..

[Kindanyume]

Hey Steve,

have you considered the use of a "AA" sized LiOn cell. (Trustfire /
Ultrafire etc) They are typically 3.7v and would be smaller still.
(Or a diff case size for that matter). Of course recharging such is
"harder" but given the crown that will actually build this I don't see
that as being a huge problem. (I thought of a built in charging
circuit as well but that could get far more complicated.. unless
someone has a suitable design handy that uses minimal
components/space.

As for costs.. from one of my suppliers a separate charger and 2 cells
can be had for $12 including shipping.

On Sun, Dec 2, 2012 at 11:34 AM, Steve Gibson <steve....@gmail.com> wrote:

> I'm working on the project full time. Last night I was playing with battery
> size and power amp input voltage, and may be switching everything from 6vdc
> to 9vdc. A "transistor radio" battery has much less total capacity than a
> set of four AA's, but it appears that I'm able to get enough additional
> power out that it would be worthwhile, and the operating efficiency of the
> system is close to 100%, so we're not wasting anything. Also, a single 9vdc
> battery is enough smaller to allow the use of a smaller case, which is a big
> convenience.
>
> But since the MSP430's can only run on 1.8v to 3.6v, I was planning to
> center-tap four AA's to get a nice "free" 3vdc for the processor. If I
> switch over to a single, closed, 9vdc battery, I'll need a little
> fixed-output three-pin regulator. That would annoy me, but this morning I
> definitely eliminated two of the existing components and possibly a third...

> /Steve.

[Kindanyume]

One more thought.. I have not investigated "yet" but in theory it
could be possible to have a very small circuit for charging an
internal AA class LiOn cell via a Micro USB port and using the now
uber common 'cell phone' chargers and/or a straight cable from almost
any computer.

That could add another aspect of portability w/o undue cost I would
think. (However again I have NOT done real investigation into it
yet).

[Steve Gibson]

Kindanyume...

FWIW, I'm back to 6vdc from four standard AA cells.  The 9vdc "transistor radio" battery wasn't really able to source the current that the system needs -- about 220 mA.  And using a single 9vdc battery then requires the use of a regulator for the processor.  By using four AA's, we can center tap the chain to easily obtain 3vdc for the processor.  And a battery's impedance is so low that there's no need for a fat tantalum cap near the processor.

And, really... the damn thing is REALLY quite loud and penetrating with JUST two AA's for a total of 3vdc power which puts 34vac p-p across the tweeter.

/Steve.

[Kindanyume]

Hmm ok.. but just one more thought.. if you considered size for a
truly portable version... maybe a CR123A class cell would help?
Either standard or LiOn type etc.

I'm also checking w/my suppliers to see what the smallest piezo I can
source and that might work for this. (or maybe 2-3 array wide
depending on size and/or case used.. you know what I mean I'm sure)

Just thinking along the lines of something that could be very handy
for say a Mail carrier.. where they could easily keep such a device in
a pocket.. etc

[Steve Gibson]

Kindanyume,

ALL of the various piezo tweeters I have referenced are two pieces held together with three or sometimes two screws.  It would be entirely feasible to use ONLY the "back portion" of the speaker, which houses the piezo element and its attached paper cone, to create an entirely "pocketable" device.  Lacking the front horn section, it would have less effective range, and less directivity.  And the horn also serves an impedance matching roll, of the piezo/cone to the air.  But it would definitely work that way too.

/Steve.

[Kindanyume]

Indeed that is true.. but till worth a lil work checking on a
"mini-horn" type just in case :)

Oh and no.. I'm not a mail carrier etc at all. Just thinking of
others whom would appreciate your invention.

[Stephen Syputa]

Dear Steve, Thanks for the details on your major aims with respect to this project. You are first and foremost a practical man. I have many more incomplete projects than completed and think that I will try to duplicate your efforts first, yielding a much needed functional device, before attempting anything more involved. My son was killed by a rabid dog attack years ago and my wife is in absolute fear of walking alone in our rather safe San Antonio neighborhood. She, a non-techie, has become a follower and admirer of your podcasts, believes that you are in essence doing what you are doing, and trusts your judgement and recommendations. I am hoping that this device will give her the courage to enjoy our beautiful environs without being in utter fear and will help her ongoing reconciliation of the death of our son. May you never thirst!

[Steve Gibson]

Stephen,

I'm so sorry to hear about your loss of a son to a rabid dog.  Wow.

Users of this "Dog Whistle" device will likely fall into two general categories: (a) Those who want to and are capable of building the device from its detailed specification -- which I will provide on my web site -- and (b) those who very much want to have one to experiment with but who lack the background to build one themselves.

For the first group I'll provide everything they need -- a parts list, schematics, theory of operation, etc. and I'll be very interested to learn of their experiences, as well as of any changes and/or improvements they may choose to make.

For the second group, my plan is to build a sufficiently large batch of devices that I can simply send one to anyone who will be willing to provide feedback about its effectiveness for whatever purpose they may have.  I suppose that a great many people would LIKE to have something that I had designed and built by hand, so I'll be asking people to really have a need for it.

And depending upon how all that goes, I may well arrange to have many more manufactured and available for sale.  We'll see.  :)

/Steve.

[Kindanyume]

Stephen,

Indeed.. such attacks and losses are very sad. But do happen
particularly via dogs w/very "bad" owners sadly. (not just rabid dogs
can kill and/or harm)

My sympathies.. though I have not been there personally I'm sure
everyone else will agree we sympathize with your situation.


Steve,

LMK about that once you are fin. Myself and one friend here will be
building of course.. but I have another non techie friend in another
city that I'm sure would want one of yours. Maybe I can setup a test
of the "3" at once and try to note any differences in build etc.
Hopefully refining and/or confirming the finished builds and any mods
done etc etc.

[tz]

Thanks for the description, and I think I will enjoy the tweeter page.  I wondered which tweeters would work the best.

(Ignore these if they will be on the page eventually)

Could you put two of the 400W models in series and use 1uH, though the frequency would be off a bit - but I thought most inductors were only within 20%, and caps can be worse.  How close does the RLC and the stimulation frequency have to match?  I'm trying to think of a way to use feedback so the micro can tune itself to the best frequency.  Now if there were only variable components - ferrite cores...  Well, it is for my motorcycle (deer, dog, kamikaze bird scatterer), and it is in for the winter, but that is when I build my toys (harleyhacking.blogspot.com).

The ATtinyX5 can go to 5v and is only 8 pins (DIP too) and has some fairly beefy output stages - and the arduino build has the tools though you need a programmer (which I have - the AVR Dragon).

[Steve Gibson]

TZ...

It would DEFINITELY be possible to have the processor self-adjust to find the optimal frequency.  All you would need is a simple peak-detector: A diode riding the positive leg of the tweeter, which pulls up a resistor divider (in order to bring the voltage down to something that won't instantly fry the processor.  And a capacitor across the lower resister so that it's charged up by the high-going sine wave through the upper resistor and diode.  But then the diode blocks quick discharge when the sine wave goes negative.  Then hook the processor's A/D converter to the capacitor... and we've given the processor the ability to read the amplitude at the current frequency and duty-cycle.  Then simply write some code to "hunt" for the optimal parameters!  :)

And, yes, you COULD put two larger tweeters in series, but then they would be dividing the voltage (and the power) between themselves.  It would be better to give each tweeter its own L and C.  (I'll have a first rough schematic posted very shortly.

/Steve.

[aaron...@gmail.com]

Steve we have a dog who's owner leaved her alone most of the day and barks for up to 8 hours at a time till he returns at night. We've asked, begged, pleaded with him to do something about it and he dismisses us every time. I would love to be on the beta testing group for this. I hope that it could end our suffering. The poor dog is just bored but this is probably our only hope at some sanity again. Thanks for all that you do, if you can put me on the best test group I promise to take detailed notes and documentation for further development of the project. Thanks!

-Aaron

On Thursday, November 29, 2012 9:49:00 AM UTC-8, Steve Gibson wrote:

Kyle,

Thanks very much for your well placed thoughts and concerns.  As you might expect, I have researched and considered this VERY important question thoroughly.

And you are 100% correct: Just because WE can't readily hear the sound DOESN'T mean that it's not still loud and potentially dangerous and damaging to us.

And, in fact, acoustic radiation is very much like other forms of radiation: Either an exposure to a large amount in a short time, or a gradual accumulation a lower level over a longer time, set against our body's slowly decelerating rate of repair as we age... can result in long-term damage to our body's delicate systems. (There's a reason the dental tech always leaves the room when our jaws are being x-rayed... and also a reason a lead apron is draped over our gonads too! :)

I absolutely intend to be VERY CLEAR and tutorial -- and to frighten people -- about the very real dangers of chronic exposure to high levels of sound -- both audible and inaudible -- and to also explain the effects of sound level and distance.

The devices which people will be able to manufacture from this work will NOT BE TOYS.  That's why I believe they will be effective... but with that effectiveness, because of that effectiveness, comes responsibility for its usage.  So the resulting devices need to be treated with the same respect as a knife or a loaded gun. The batteries should be removed when it is not being actively used.  It should be off limits to the kids and kept out of the hands of children.  And I'll also be providing graphics for the creation of clear warning labels so that an understanding of the device's nature is transferred to, and appreciated by, anyone else who ever uses it if ownership ever changes.

There was a LOT that I did not have time to discuss during yesterday's podcast (and it wasn't, after all, a "sonic blaster" podcast).  But there are several reasons I chose to use a microprocessor as the system's "oscillator" rather than a much simpler (though, perversely, not much less expensive) analog timer chip:  For one thing, I plan to generate a MUCH more "annoying" sound, by pseudo-randomly varying the positioning of the rising and falling edges of the output square wave.  This will create more of a low-ultrasonic "screech" than a pure sine-wave "tone" by producing a very harmonic rich sound.

And secondly, I plan to incorporate a built-in timeout so that it's NOT POSSIBLE to "tape down" the button of the device in order to run it continuously.  If someone wishes to circumvent those safety measures the responsibility will be theirs.  But I cannot see any typical need for ever generating more than about half a second of sound from the device.  (When my buddy Mark made the neighbor's yappy little dog jump three feet into the side of the house, he had just given it the barest of a "tap".)  The goal is NEVER to "hurt" anything in any way... ONLY to startle and train.  And for that a brief "blip" is probably even more effective than a protracted event.

And lest we get TOO carried away with all this, we should remember that all this is, is a generic audio system piezoelectric tweeter connected to an amplifier and audio oscillator.  It's really not so exotic.  Mostly I've just made such a system convenient, portable, and battery operated.

But that said, I want to and plan to GO OVERBOARD with being as responsible as I can possibly be.  As you originally pointed out, just because we can't readily hear it doesn't mean it's NOT loud and potentially damaging.  In that regard, this device DOES differ from a gun or a knife in that common understanding tells us about the inherent dangers and need to be responsible with those well-known weapons.

I need to similarly educate those who wish to experiment with, and employ, high-power high-frequency sound generation equipment.

(And I feel strongly about clearly labeling the device so that when it inevitably falls into the hands of others that understanding will survive and be communicated.)

/Steve.

[Ben Sims]

Hi Steve, 

    I have two dogs, both rescues, A German Shorthaired Pointer, and a Vizsla/Britney mix that are super well behaved. That is until I let them alone in the back yard of our new home at which point, they work together to dig holes and chase the root systems of a nearby tree. 

    I have tried scolding them, squirts of water from a spray bottle, Stomping on the deck, and distracting them with rawhide bones, but the wife and I draw the line somewhere before a shock collar. I mean these dogs suffered various levels of abuse before we came into their lives, and we don't want to risk injuring them.  Anyway nothing has had a lasting effect. 

    I would LOVE to get into your test program to use the PSB, as a training tool to teach them to stop digging. And like Aaron said, I would be more than happy to take detailed notes, perhaps even submit video to help further the project.

-Ben Sims

C|EH

[Kindanyume]

Sounds exactly the same as the situation here. I feel sorry for the
dog actually.. but legally here there is nothing I can do to make the
"owner" give the dog up to a better (and caring) home.

[Boilerbots]

Watch out for inductor saturation current.

I ran some numbers through a simple calculator and see that your inductor through the FET to ground might reach saturation at a frequency below 18KHz at 9V. The calculator says that at 6V you might be able to get down the 12KHz but this is an ideal calculator, your mileage may very. You can use your current shunt resistor to measure the current passing through L1 and the FET.

You mentioned in a different thread that someone might be able to turn the frequency down for Crows but not without an inductor change.

Generally your output voltage is directly related to your duty cycle and you should keep the inductor out of saturation. You can very the output voltage and thus the volume by changing the duty cycle, well at least that works for a boost converter which is it minus the output diode.

Curt

[Steve Gibson]

Curt,

Thank you for your analysis.

Watch out for inductor saturation current.

I ran some numbers through a simple calculator and see that your inductor through the FET to ground might reach saturation at a frequency below 18KHz at 9V. The calculator says that at 6V you might be able to get down the 12KHz but this is an ideal calculator, your mileage may very. You can use your current shunt resistor to measure the current passing through L1 and the FET.

THAT is a terrific idea that hadn't occurred to me.  I'll definitely do that.  That's a great thought.  I had played with using a 1.5mH inductor, but for some reason I've had it stuck in my head that it would be "better" to use the same value for both roles... which is clearly not the case since their roles are entirely different.

You mentioned in a different thread that someone might be able to turn the frequency down for Crows but not without an inductor change.

Right... somewhere here I did talk about switching to 1.5mH inductors if the frequency were dropped to 10kHz.
 

Generally your output voltage is directly related to your duty cycle...

Yes it is.

 

...and you should keep the inductor out of saturation. You can vary the output voltage and thus the volume by changing the duty cycle, well at least that works for a boost converter which is it minus the output diode.

Reducing the 'on' time does indeed lower the volume nicely while bringing the total current consumption down.  And when it's being driven less hard, the tuned output LC tank (the 'C' being the piezo) generates a nicer (purer) sine wave too.

Thanks again Curt.  It's great to have more pairs of eyes on all this!  :)  I'm going to play with the boost inductor and see about perhaps bumping it up to 1.5mH to keep it further away from saturation.

/Steve.

[Kianoosh Salami]

Steve, I know you were against using an external regulator, however, just in case you decide you want to use one. I found these awesome switching regulators awhile back. The reason I think they are awesome is because they are switching regulators and all the components are sealed with a 3 pin output. Vin - Gnd - Vout. Can't get simpler than that. 

For example, the R-78E3.3-0.5 can take 6V input and output 3.3v with efficiency ~92%. They go for about $3.65 a piece on digikey.

http://www.recom-power.com/?id=271&no_cache=1&user_recom_pi2[xseries]=R-78xx-0.5+R-78HBxx-0.5(L)+R-78HB24-0.3(L)+R-78Exx-0.5+ROF-78xx-0.5

[Boilerbots]

All inductors are not equal but generally a ferrite core will be the most effecient for this application. Coilcraft has good calculators to pick an inductor and they will ship you free samples pretty fast.

Also choose low resistance caps for the output since you are putting a sine wave across it, otherwise it is going to heat up during long bouts of testing. State of the art ceramic caps are hard to beat for this reason and they are fairly forgiving if you exceed their voltage specs. I lost track of where your schematic is located in this groups thing but the photo looks like you have ceramic caps.

Curt

[Steve Gibson]

Thanks for the pointer kiantech.

When I was considering a regulator, I was just going with a super-cheap (30 cents) 3-terminal 100mA, very-low-dropout regulator, since the only thing it needed to supply was the processor chip itself... and that chip only needs a mA at most.  So anything else would have been overkill.

However... I will DEFINITELY remember that for the future.  It's sweet indeed!

/Steve.

[Steve Gibson]

Curt...

Very good points all.  Thanks.  And you're right, I'm using exclusively ceramic caps throughout.

Here's the schematic: http://www.GRC.com/acoustics/Dog_Whistle_Tuned_Power_Amp.jpg

Any suggestions and feedback you have will be more than welcome!  The more eyes on this, the better!  :)

After our discussions yesterday, the first thought that popped into my head as I awoke this morning, was that there must be some wicked transients on the +3vdc VCC supply line to the processor.  Since it's power needs are so ridiculously low, I'm going to add a series resistor in the line powering it and probably increase its bypass cap from 0.01uf to 0.1uf.

/Steve.

[kiantech]

If you center tap the batteries, you will be discharging the batteries at a different rate, correct? This would shorten the battery life. However, you might argue the power consumed by the MSP430 relative to the tweeter is so small that this cost/benefit does not make a noticeable difference in battery performance.

[Steve Gibson]

kiantech, you wrote:

If you center tap the batteries, you will be discharging the batteries at a different rate, correct? This would shorten the battery life. However, you might argue the power consumed by the MSP430 relative to the tweeter is so small that this cost/benefit does not make a noticeable difference in battery performance.

Precisely!  That is exactly my thinking.  And even better, the center-tap will be drawing nearly zero current during the blast, since the instant I finish setting up the free-running timer in the chip... I shut down the processor!  <grin>

/Steve.

[greg...@gmail.com]

Hi Steve,

I just wanted to chime in and (if it's worth the shipping cost) offer my time, (somewhat-decent; I can build these pretty reliably) skill, and soldering station if you could use more pairs of hands to assemble the CAAT test units. It does look beautifully simple, but like you said, making 1 custom circuit is a different ballgame than 50+ :).

--Greg

[Steve Gibson]

Hi Greg...

What a beautiful little antenna!  Nice.

I'll definitely keep your offer in mind.  It's all a function of how many need to be built.  I might switch to surface mount components if a lot are needed.  We'll see.  Time will reveal all.  :)

/Steve.

[dakin...@gmail.com]

On Sunday, December 2, 2012 12:46:18 PM UTC-7, Steve Gibson wrote:

Stephen,

Steve, 

 

Put me in group 2. I want to purchase your Portable Sound Blaster ready made and assembled by you. I want to reward you for your efforts.

 

I believe that I could assemble the parts to make one but to own somethhing built by you would simply be the coolest thing ever!

 

Additionally I really do have a dog that comes to visit every weekend. Camby needs training as she barks altogether too much and after buying 2 commercially made devices that didn't work or we decided we couldn't try (electro shock training) I'm almost desperate.

 

Thanks for Security Now and of of your great free apps, not to mention SpinRite.

 

Dakineathon

\For the first group I'll provide everything they need -- a parts list, schematics, theory of operation, etc. and I'll be very interested to learn of their experiences, as well as of any changes and/or improvements they may choose to make.

[Mike Rogers]

Hi Steve:

I printed off the “Hush Puppy” schematic v0.1 as soon as it became available & ordered the parts the next day. For some reason, I can no longer find it online.  I have gone over the schematic many times to ensure I assembled it correctly however, when I apply power, I hear a small, barely audible “tick” then nothing every time power is applied. I measure 3.21 V between pins 1 & 14 of U1 & 6.43 V between L2 & C3 & ground. I’m using a Pyramid tweeter that says: “Typical Impedance appears as .13 uF”, 200 watts, 94 dB @ 1w/1m.  The speakers came as a pair & I tried both. Should you have a spare moment, I could use some guidance to why I cannot get this to work. FYI, I have a basic knowledge of electricity.

Also, I work at an International airport in Canada & part of our duties is wildlife control, which 80% - 90% would involve displacing birds from the vicinity of the runway.  I’m confident you can appreciate how hazardous the larger ones can be to aircraft. We generally use pyrotechnics to scare them away however, when they don’t leave, we are left with no choice but to shoot them for safety reasons. This is something I am not fond of doing. Since the PDK episode, I’ve been waiting for over 2 years for something to come to life regarding the PDK to try on birds as you tried it on a seagull when you were younger & said the bird definitely knew it. I would be more than willing to document it for you complete with video. It would most likely have to be a larger unit like the pics you posted of the one built for your friend. You can advise me of any costs associated with the device as well as shipping. Hopefully we could save some stubborn wildlife. Thanks again for all that you & Leo do! First time posting to Google groups, hopefully it's done correctly.

[Nate Hamilton]

>... but to own something built by you would simply be the coolest thing ever!

That is exactly what I was thinking! I don't have a dog problem but I think it would be cool to own one!

--Nate

[Steve Gibson]

Mike!

Congratulations on being the first, so far as anyone knows, to assemble a "Gibson-Derived" latest-generation "PDK" !!  :)

Your circuit looks 100% correct to my eyes.  But the question I have is... I see a 14-pin DIP without any visible markings on its back.  So I assume that's the MSP430G2001 that I referred to on the schematic?

If so... are you saying that you also have one of the $4.30 Texas Instruments LaunchPads.  That you downloaded and installed the 1.4 gigabyte Code Composer Studio v5.3.  Then grabbed the early posting I made of the firmware I wrote to calibrate the G2001 chip's clock from its default of 1Mhz to 12Mhz?  Then assembled and downloaded that into the chip and ran it to update the chip's flash calibration?  Then you used the later source code I posted, of the Dog Whistle operating firmware for the G2001, assembled and loaded that into the chip... and then plugged the chip into your breadboard??  And that's what you're using?

After having just run though the steps needed to get a "blank" MSP430G2001 chip to operate... and since I haven't heard from you before now... I think that the answer to my questions must be... "whoops!!"  :)

Here's what I propose:  Let me just SEND YOU RIGHT NOW a ready-to-go, pre-programmed, G2001 microcontroller chip.  I'd LOVE to see your circuit run.  :)

I have 100 of them, and they cost me 53 cents each (Digikey gave me a 100-piece discount, that's why I bought 100).  So the postage to you is going to outweigh the parts cost!  <g>  And I'd be GLAD to.  I think it's SO COOL that you built one so quickly and I'd love to support your efforts.

It must be that you didn't appreciate how much of the technology is inside the chip... and that it needs to be custom programmed as a consequence.  But having just said that... it's also 100% true that I COULD have designed the solution around an off-the-shelf analog 555 timer, with one pot for frequency and another pot for duty cycle.  But I did that the first time when I built "Big Bertha" for Mark.  So this time I was aiming more for an ultra-low component count design.  Which I did get, but at the cost of moving about half of the "moving parts" into firmware.  That's WHY your circuit is SO clean and tiny.  If we were using a 555, we'd need two pots and two capacitors.  The famous "555 Timer" is so ubiquitous that Wikipedia even has en entry for it:  http://en.wikipedia.org/wiki/555_timer_IC

Check out the "Astable" section... that's what you want.  Notice, also, that the "high time" is R1+R2, whereas the "low time" is only R2.  This is PRECISELY what we want, since we want the output to be high for 2/3rds of the time and low for 1/3rd of the time.  So using a 555 timer to generate the 66% duty-cycle waveform to drive the MOSFET should be VERY straightforward.

That said... I would be GLAD to shoot a ready-to-go pre-programmed and tested (on my prototype) G2001 chip to you immediately.

Now, or in the future, there IS also an easy solution that doesn't require me to physically send you anything:

You could get one of the $4.30 LaunchPads, either from T.I. or DigiKey or Mouser.  Then you would have the ability to download the firmware into them yourself.  And I recently went looking for, and found, a solution that will allow people to download firmware into their chips WITHOUT needing to mess around with assembling any of my code.  But if you just want to get something going now... I'd be glad to send you a ready-to-go chip.  :)

I have just created an eMail account for this project: hpedw2013 (at) grc.com (Hush Puppy Electronic Dog Whistle 2013).  Send me your mailing address and I'll send you a G2001 chip. (And if you could use more than one, and have parts to use more than one, tell me how many.)

/Steve.

[Kindanyume]

Nice discount on the chip Steve ;)

PDK comin alone very quickly and looking very impressive IMO.

[roger...@gmail.com]

Steve:
           You are correct, it is the MSP430G2001 that I purchased. It's definitely a big "whoops!!" on my part. I did not realize the chip had to be programmed hence why I stated: I have a basic knowledge of electrical. If I had read the previous posts more closely, it would have been apparent. I will purchase the LaunchPad once the design is final just in case some more parts are required as I have over $30.00 worth of shipping charges on 2 orders of $10 parts so far (shipping to Canada). The steps listed in your previous post for programming may be above my skill level but I'm willing to learn. I will take you up on your offer of the pre-programmed controller.  Also, in version v0.4 you changed C1 from a .01 uF to a 10 uF. What is the purpose/effect of doing this? Looking forward when you have a chance to post an explanation of how this circuit actually works. Thanks again!

[Kindanyume]

To those in Canada... I will be working on importing parts etc once
the design is finalized. It will probably be cheaper to do it in
batches.

Depending on where you are located etc etc.. or Maybe a cpl of us can
do it in a distributed fashion across the country.

(GTA area here)

[Steve Gibson]

Mike...

You are correct, it is the MSP430G2001 that I purchased. It's definitely a big "whoops!!" on my part.

Once you have the $4.30 LaunchPad (free shipping even to Australia and presumably to Canada) you'll be able to EASILY download my finished code into that 'blank" chip... so you'll be able to use it anyway.  :)
 

I will purchase the LaunchPad once the design is final just in case some more parts are required as I have over $30.00 worth of shipping charges on 2 orders of $10 parts so far (shipping to Canada).

FWIW... DigiKey DOES also offer the LaunchPad (MSP-EXP430G2), so combining it with an existing order would make sense.  But I believe you'll find that Texas Instruments will SHIP it to you, in Canada, for the same $4.30 *total*... so no reason to wait on that one.  :)

 

The steps listed in your previous post for programming may be above my skill level but I'm willing to learn.

Learning is good.  But we'll also soon have a short-cut means of downloading code into the chips -- through the $4.30 LaunchPad -- but WITHOUT the need to learn anything at all.  :)

 

I will take you up on your offer of the pre-programmed controller.

They are headed your way!  :)

 

Also, in version v0.4 you changed C1 from a .01 uF to a 10 uF. What is the purpose/effect of doing this?

Just me being me -- endlessly fine tuning the design.  They are non-critical changes.  I wanted to use two of the same component for convenience, and I also determined that their voltage could be reduced to just 6.3v... down from 50v and thus MUCH less expensive.
 

Looking forward when you have a chance to post an explanation of how this circuit actually works. Thanks again!

Me too!!!   :)

/Steve.

[Steve Gibson]

Yes... and DigiKey offers reachable price breaks, often at just qty. 10.   :)

/Steve.

[robtu...@gmail.com]

Hi Steve

I too would like to be in group 2 and test, demo or purchase a unit from you when they become available

Thanks

Rob

[Steve Gibson]

Sounds great Rob!  I'll be pulling all that together soon!

Thanks for your note.

/Steve.

[dscur...@gmail.com]

Hi Steve, I have been listening to Security Now and many other Twit shows for a few years now and yours is my favorite, although most of is over my head I have learned a lot. I don't think I could build one of these but would like to be in test group because of the neighbors dog my wife and I have considered moving. Other sound devices we have tried have not worked and want to give this a try. 

Thanks 

Darren

[book...@gmail.com]

Hi Steve, just got my LaunchPad Kit yesterday, I am in Canada and Texas Instruments sent it to me free shipping and it took only a day for the shipping part, so I can confirm that it is also possible to buy the kits from Canada no problem.

 

I will make a couple of PDK for myself, I'm thinking one portable to train one dog that scares the kids (and myself when I'm with them) when they go to school. Another for the house, so that I can train the neighbor dog. And probably one more portable to lent to friend and family when needed.

 

And since it's winter here (I'm in Montreal), I have time to build them since all windows are closed and the dogs don't go out much now, so they will be used only around May next year.

 

If there are some people near Montreal in need, I can help and build some; I have the skills for it, I went to school in electronics in my college years and made many project myself, the last one I can think of was a gift for my girlfriend "Etched Copper Board Valentine".

 

Thank you for security now, I was listening from the beginning, I was even watching The Screen Savers when it was on TechTV. The PDK episode of Secutiry Now is one of my favorite; it was so funny to me.

[Raymond Ilano]

Hello Steve,
I would like to be added into Group2. I have the perfect candidate for testing the device. A German Shepard next door who barks incessantly day and night. The hours I don't hear him are fewer in 24 than I do. When I went next door to discuss this with my new neighbor (we just moved here) he  told me, "Oh sorry about the noise. He just does that I don't know why. I even had a bark collar on him and it didn't do anything. You'll get used to it after a while." I'll get used to it?! I'll get used to it?! Yeah right whatever. This thing barks at the wee hours of the morning and all day long for no reason. I have sat in the backyard and watched him just run back and forth and bark. It's like a machine gun. I contacted the local authorities and they said I have to have support from other neighbors or they won't do anything. The other immediate neighbors don't seem to care. Please let me know if I can either beta test the device or just purchase it outright. Thanks.

[gilb...@gmail.com]

Add me to your Group Two list!  We have two neighbor dogs that can use some training.

Thanks for all your efforts and the postcast!

Gilbert

[Thomas Trostel]

We used these parts on a project a while back.  They are about 80% efficient converting 9v to 3.3v and fairly inexpensive if you want to waste less power

Unfortunately at their price they are more expensive than anything else in the design!!  Thats really quite a testament to the design though.

http://www.digikey.com/product-detail/en/V7803W-500/102-2446-ND/2757916

On Dec 2, 2012, at 11:34 AM, Steve Gibson <steve....@gmail.com> wrote:

Tom...

You wrote:

So where , when do we see parts list ... software ? I have a bit of experience with Arduino ... I imagine it too would be able to drive the amp. I can build stuff ... Jamco. Allide, etc.  We have terrrible problem with Ravens taking / wrecking all our fruit before its ripe! I don't have the heart to kill and don't desire to own a gun. I'd love to build prototype with single tweeter. If there was a kit available I'd buy just to expidite. 

anxiously awaiting.

I'm working on the project full time.  Last night I was playing with battery size and power amp input voltage, and may be switching everything from 6vdc to 9vdc.  A "transistor radio" battery has much less total capacity than a set of four AA's, but it appears that I'm able to get enough additional power out that it would be worthwhile, and the operating efficiency of the system is close to 100%, so we're not wasting anything.  Also, a single 9vdc battery is enough smaller to allow the use of a smaller case, which is a big convenience.

But since the MSP430's can only run on 1.8v to 3.6v, I was planning to center-tap four AA's to get a nice "free" 3vdc for the processor.  If I switch over to a single, closed, 9vdc battery, I'll need a little fixed-output three-pin regulator.  That would annoy me, but this morning I definitely eliminated two of the existing components and possibly a third...

My point is... the design is NEARLY, but not yet, fully complete.  It'll happen when it happens.  The whole reason I do these things is because it's unknown and unexplored territory, which is what makes it interesting and useful for me.  But a side effect of that is that I have no idea when anything will be "finished".

So, without wanting to be a smart-ass... it'll be ready when it is... but that won't be long from now... and it'll be SERIOUSLY tight and wonderful once it's done.  :)

Note, also, that birds DO NOT HAVE the same acute high-frequency hearing that canines possess.  You can try it as is (and we'd all love to know how it works).  But it might be necessary for you to drop the frequency by half, to 7.5 KHz for it to be readily audible by Ravens.  That would be a trivial change to the processor's code, and you could just double the value of the series-connected inductor from 1mH to 2mH.

More soon!

/Steve.

[Steve Gibson]

Thanks Thomas.

That's a nifty little module!

[Braden Kear]

There will be a GRC "T.Q.C." device operational in BC next week. I got the TI kits (to Canada) in about 3 days, and the rest of the components from my parts pile and local shop. I was able to load both of your hex files successfully to the chip with that nifty command line tool. I was able to figure it out without much command line experience, reading the included PDF and running the example batch file helped a lot to understand the way to use this tool. I attached a .txt of my experience programming a G2001 chip on the EXP430G2 board, it looks OK to me.

The only delay is the tweeter! Tues. next week looks like the ETA for them. I verified a pulse output from the G2001 pin 4 tonight, I'll grab some fresh AA's and bring home my pulse counting meter from work tomorrow to check it's function, frequency, and duty cycle before the tweeters arrive next week. I have a variety of AA holders here, and once the driver is in my hands, I'll see what I can do about building a compact handheld unit, probably with some sort of safety lockout and warning decal, too much fun!

I have a variety of stubborn and sometimes aggressive animals to test this with, including a cow and a pig, I'll report on it's effectiveness.

-Braden

[Steve Gibson]

Braden!

Very cool all!  And you've gone farther than I had with using the "Flasher" Flash memory programmer.  Nice.

One thing... while you had the device programmed for calibration of the G2001, did you RUN the code that you had loaded into the chip?  I'm unsure whether the Flasher RUNS the code that it loads.  But you do need to do that in order for the G2001 to be able to calibrate itself.  And... for that you do need the optional 32,768 Hz crystal added to the LaunchPad.

When you run the Calibrator code with the 32,768 Hz Xtal, the LEDs first light Red, then Green... and finally flashing Green in indicate successful calibration.  If the crystal is not in place or running, I turn on the RED LED steady.

So... if at some point you DID see a flashing GREEN LED... the first step completed and you're then ready to load the code for the permanent oscillator.

(I do feel somewhat awkward that I published the approach for using a G2001 chip (and the schematic too) since going through all added hassle of needing to solder the tiny surface-mount watch crystal to the LaunchPad and run a calibration program once... well... NONE of that is required for use of the higher-end chips that accompany the LaunchPad.)

[Braden Kear]

Steve, yes! I have a flashing green LED after calibration  I manually ran the program after, but adding "[VCC]" to the end of the programming command would do that too (just tried it, it works). Without explicit instructions, I had soldered the micro crystal to the Q2 pad on the launchpad board (pin 12+13 on the G2001), since you had referenced it.

Thanks again!
Braden

[Steve Gibson]

Nice going Braden,

And, in that case... I'll bet that the signal you are seeing on Pin 4 of the chip when it's running the 15khz oscillator code WILL indeed turn out to be a 2/3rds high 15khz square wave!  :)

/Steve.

[sag...@gmail.com]

Can someone build me one I will pay for it?  There are these two Chihuahua's that circle me while barking aggresively.  They never attack but I am sure if I lost focus they would try to bite me. 

[Steve Gibson]

Hi!

My plan is to produce a bunch of these fully functional units and, to some degree, use quantity to bring the cost down while making a hand-built commercial grade unit.  And, because FEEDBACK of users' experiences with them will mean everything to me, my offer will be to provide to a set of "beta testers" in return for their agreement to provide as much feedback of any kind -- candid cellphone videos would be welcome -- so that I/we can learn about the effectiveness of this approach and these devices for discouraging unprovoked dog barking.

Can someone build me one I will pay for it?

There are these two Chihuahua's that circle me while barking aggresively.

They never attack but I am sure if I lost focus they would try to bite me.

I am a dog lover.  I grew up with dogs.  I "understand" dogs and cats.  And I would NEVER hurt any of either.  But I can certainly understand the pleasure I would feel blasting the literal crap out of two annoying, and actually bullying -- because that's what it is -- Chihuahua's.  Again... NEVER to hurt them (and this device doesn't, can't and won't) but to turn the tables on them a bit and say, in a "voice" than they would understand... "Get the hell away from me you little yappy shits!"

Please go over to the "Feedback" page at "The Quiet Canine" area of GRC and sent me a note about your problem and need for one of these:

http://www.GRC.com/tqc/feedback.htm

/Steve.

[rfsc...@gmail.com]

Good evening Steve (evening EST):

First, thanks for everything!  Great Podcast (you and Leo on Security Now), great products (SpinRite), services (all the freebies).  Loyal listener (although I didn't start day 1 - maybe a year in) I have been through all of the SN Podcasts.

If you need a beta tester I will assist in any testing, documenting, etc.  If the dance card is full I will proceed in building one as I think my skills are just at the level to follow a schematic, easily in soldering and assembly.  I would be proud to own a RC built one too so any methos will work.

I am in a rural setting and I own dogs (and more - we have a hobby farm). On our long walks there are a few "unfriendly" beasts and this would be great for myself or my wife to keep those critters at bay.  It may also be a good trainer for new pups to stay away from the other animals.

Keep up the good work and don't ever stop!

Regards,

Ray

[Braden Kear]

This meter isn't a high precision unit, but the readings are pretty close! I'd say the flash was successful. Now just the wait for my tweeters!

Braden

[Joe H]

Some feedback on the cheap pyramid tweeters from Amazon: http://www.amazon.com/Pyramid-TW105-3-75-Inch-Watts-Tweeter/dp/B0013CFOBY

I did some experimenting when they arrived yesterday. I was not impressed with the sound output of these at all, compared to my previous experience with piezo buzzers. Some of which can be ear splitting driven with just a 12 V square wave.

I've also destroyed one after driving it with your amp. The mosfet drive came from a function gen set to 60% duty cycle. As the piezo heated up, I chased the resonant frequency (decreasing) while viewing the voltage across the tweeter on an O'scope. It died in less than a minute (somewhere around 10Khz). No smoke or fire :-(, just a very decreased output, some audible ”crackling" and the resonant frequency jumped up to around 30 Khz (even after cooling off).

The failure was caused from chasing resonance and operating continously, which obviously wouldn't happen with the unit your designing, but it might speak of the quality of these cheap tweeters.

Failure aside, my main concern was that they just weren't very loud. I haven't driven the second surviving one hard yet but the sound output at low levels is comparable to the first (tested in the very audible range of around 4 to 10 Khz driven directly from the function gen at around 20Vp-p sine wave).

[Steve Gibson]

Yes, Braden, I completely agree.

The tuning of the chip's oscillator is also not infinitely sensitive.  I scan up from low-freq until I hit the target.  What I might have done, but it seemed unnecessary, was also check the next higher frequency setting to see whether it was CLOSER to the goal than the first one we found.  But, again, given the large 10% to 20% tolerances on all of the rest of the analog components... it was clear then that there was really little point.

/Steve.