Constructed Devices Show & Tell

[Steve Gibson]

Everyone...

This top anchored topic serves as a place for builders of various Quiet Canine TrebleShooters and offshoots to show off and describe their work, and for others to react and interact.

/Steve.

[Myron]

Regarding  TrebleShooter V1.1

I am getting ready to make an attempt at making one of these myself. I can solder things and I can read a drawing. But that's about it.  I have a couple questions.
I think I would need some type of board with holes in it to solder to. Any suggestions?

My other question is that I assume the CPU is programmed on the development board and then moved to my circut board. shouldn't I use a socket so I would have the ability to take it off and re-flash it? I might have some old memory cards or video cards that I could de-solder.

Am I on the right track?

Myron

below is my parts list:
1 - board with holes in it (BREAKOUT BOARD)    NEED PART NUMBER?
2 - batteries 9v        GOT
2 - 9v BATTERY HOLDER    http://www.digikey.com/product-detail/en/968/968K-ND/151577
2 - 445-8516-ND        http://www.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go&lang=en&keywords=445-8516-ND&x=16&y=16&cur=USD   
1 - 445-3752-1-ND    http://www.digikey.com/product-detail/en/TSL0808RA-221KR54-PF/445-3752-1-ND/1913616       
1 - CPU                GOT ?? http://www.digikey.com/product-search/en/programmers-development-systems/eval-and-demo-boards-and-kits/2622039?k=MSP430G2553
                    TEXAS INSTRUMENTS
1 - High MOSFET driver     http://www.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go&lang=en&keywords=TC4421AVPA&x=0&y=0&cur=USD
1 - 497-7288-ND        http://www.digikey.com/product-detail/en/L78L33ACZ/497-7288-ND/1038304   
1 - SWITCH            GOT
1 - TWEETER        http://www.ebay.com/itm/ProCraft-Piezo-Tweeter-70-Watts-/261114510098?pt=US_Speaker_Parts_Components&hash=item3ccba33b12   
1 - CASE            Maybe an old 6v large battery flashlight.
1 - ?SOCKET FOR CPU?

[tco...@gmail.com]

You can use perforated board for placing your components, search "perfboard" at radioshack.com to find some. As for the sockets, they're also available at radioshack, just search 14-pin DIP socket and you'll find a couple of options. Of course everything you need is also available at digikey and mouser, as long as you know what to search for.

Also, you may find the "multipurpose PC board" easier to work with, as it comes with holes that have pads and some interconnects already, search "276-150" on radioshack or look for it on the web from any electronics supply site.

[Mike King]

You're on the right track.  (see links below)

Perforated Board: http://www.digikey.com/product-detail/en/8022/V2018-ND

DIP-20 Socket: http://www.digikey.com/product-detail/en/1-390261-6/A100208-ND

[coo...@gmail.com]

[coo...@gmail.com]

On Friday, January 4, 2013 9:35:14 AM UTC-8, Steve Gibson wrote:

[Steve Gibson]

Spec-TACULAR!

The 9v battery poking out of the end of one of the handles give a nice sense of scale, showing that it's not overly thick and would be easy to hold!

Very Nice!!!  Are they loud???  <g>

/Steve.

[coo...@gmail.com]

I have soldered up the components, but have not been able to get sound yet.  Does the U1 side on the schematic indicate the indent on the top of the MOSFET chip?

[Steve Gibson]

I have soldered up the components, but have not been able to get sound yet.  Does the U1 side on the schematic indicate the indent on the top of the MOSFET chip?

I assume that we're talking about one of the v2.x.x designs?  Like v2.2.2?  If so, here's the TC4421A chip diagram:

/Steve.

[Kindanyume]

swwwwweeeeeeet... mail me some.. that is better than I've got so far! LOL

(no such fancy tools handy here)

[coo...@gmail.com]

Re soldered MOSFET chip using TC4421A chip diagram.  Now have high frequency sound.  No low frequency sound.

On Friday, January 4, 2013 9:35:14 AM UTC-8, Steve Gibson wrote:

[Steve Gibson]

Nice!  And about the right frequency!

You need to add the second capacitor, to drop the frequency to around 8 kilohertz so that YOU can get a sense for whether it's a LOUD sound!

/Steve.

[Whistle1940]

Everyone,

Here is my first attempt.
The circuit is v2.2 and the container is a Mainstays Mix & Serve Container 2 cup capacity MS043-430-10-19 from Wal-Mart. The notch in the screw on cover is due to a pour spout on the container.
I cannot hear it working (72 years old, worked in construction, had tinnitus since childhood). But a twenty-something said it was irritating and the dogs do take notice. I have ordered some boards from BatchPCB and will build a v2.2.2 when they arrive. The frequency of a 1mh build showed 17khz on a Fluke meter so I am using a 1.5mh inductor (14.6 khz on the Fluke) .

[Steve Gibson]

Everyone,
 

Here is my first attempt.

Nice!  I love seeing these!  :)   All different.

 

The circuit is v2.2 and the container is a Mainstays Mix & Serve Container 2 cup capacity MS043-430-10-19 from Wal-Mart. The notch in the screw on cover is due to a pour spout on the container.

Clever use.

 

I cannot hear it working (72 years old, worked in construction, had tinnitus since childhood). But a twenty-something said it was irritating and the dogs do take notice. I have ordered some boards from BatchPCB and will build a v2.2.2 when they arrive. The frequency of a 1mh build showed 17khz on a Fluke meter so I am using a 1.5mh inductor (14.6 khz on the Fluke).

I endorse that decision. If the 14.6 kHz isn't annoying to you, I'm of the belief (not based upon empirical research yet) that it would more annoying to a dog than the higher 17 kHz.  And we're unsure where the tweeter's high-end response begins falling off.  So, IMO, the further we are from that unknown point, the better.

/Steve.

[Brian Hall]

1 - board with holes in it (BREAKOUT BOARD)    NEED PART NUMBER?

I bought a 25-pack of Veroboard perf board (individual copper pads per hole) .1" (http://www.amazon.com/Prototype-Perfboard-x1-34-250hole-Epoxy/dp/B008CG62DI/ref=sr_1_7?ie=UTF8&qid=1357618891&sr=8-7&keywords=veroboard) for about $14. I'm sure I will end up using these for other projects as well.

Brian Hall 

[Brian Hall]

Here's my first attempt at building v.2.2.2. I used an empty flashlight body and mounted the tweeter about 1/2" below the end where it flares out slightly. I had to remove a lot of the plastic trim around the tweeter with a hacksaw and a file. Also, be careful with the wires connecting to the tweeter--both leads broke off while I was working on them. I had to re-solder the leads back on twice. In hindsight, I probably should have found a larger diameter pipe to mount the tweeter further down, allowing it to resonate better and give it more directionality. The pushbuttons I used are Digikey P/N: EG1900-ND (http://www.digikey.com/product-search/en?x=14&y=20&lang=en&site=us&KeyWords=EG1900-ND), and I really love them. I plan on using this same pushbutton in any further builds, and for other projects. I love how they feel and react.

Here is a photo of all of the components before mounting in flashlight enclosure:

Here is a photo of the finished project:

There is a hole in the bottom of the flashlight case for battery access. The flashlight originally came with a textured, rubber sleeve (not pictured) that fits over the entire flashlight, which I put over the bottom to hold the battery in. I had to cut a small area for the buttons to poke through the rubber casing. I used Gorilla Glue (I love that stuff) to hold the buttons and the tweeter in. Thinking of creative mounting solutions is one of the funnest parts of building this!

[coo...@gmail.com]

v.2.2.1

LOUD!!!...Put your head phones on for the lower frequency as it has a very high pitch sound.

The trick to getting the lower frequency is to hold both buttons down.........I know, some of us are slow on the uptake. The test
button was a great design feature.

Here are some comparisons between the small 1.5" and large 3" speakers at 18" from mic.
 
Frequency's are with both buttons held down showing 8 & 16+ kHz. Both the same.

Oscilloscope is using high frequency button only.  I don't know the significance of the different widths. They get narrower as the
 distance increases.

I tried it on a older dog.  He just looked at me! Older dogs must have hearing issues too.  A raven flew away.

On Friday, January 4, 2013 9:35:14 AM UTC-8, Steve Gibson wrote:

[Sean]

Here is my build of the v2.2.2 TrebleShooter.  I've got no experience soldering up electrical components at all as you will be able to see.  It took about 2 hours to make up the board.  The wife and kids tell me it works really will and loud.  Pressing the test button gives me an instead headache so that works well.  The box is made out of 3mm MDF, wood glue and 6mm dowel rod on the inside for added strength.  The box is 200mm long X 91mm wide X 91mm high.  I used an old cupboard handle I had laying around for a handle.  On the right is the 16khz tone button and the left is the 8khz test button.  I've still got a little more work to do on it but for the most part it is done.  Now to wait and try it out on the next doors dog and see what happens.

TrebleShooter Pics

[Sean]

On Saturday, 12 January 2013 13:37:48 UTC+10, Sean wrote:
> Here is my build of the v2.2.2 TrebleShooter.  I've got no experience soldering up electrical components at all as you will be able to see.  It took about 2 hours to make up the board.  The wife and kids tell me it works really will and loud.  Pressing the test button gives me an instead headache so that works well.  The box is made out of 3mm MDF, wood glue and 6mm dowel rod on the inside for added strength.  The box is 200mm long X 91mm wide X 91mm high.  I used an old cupboard handle I had laying around for a handle.  On the right is the 16khz tone button and the left is the 8khz test button.  I've still got a little more work to do on it but for the most part it is done.  Now to wait and try it out on the next doors dog and see what happens.
>
>
> TrebleShooter Pics

Got a chance to try mine out. My brother in law came around with his dog. A Irish something X, 8 months old. Not really sure what the dog is. It's the size of a small horse. Anyway I was sitting on my front verandah and the dog was sitting, back turned to me, about 2 - 2 1/2m away. I gave the 16khz button a quick press (0.2 seconds). The dog heard it as she turned around to see what the noise was. So I pressed it again for 1 second. She turn her head in a inquisitive look but it never seem to bother her. At this point she started to walk towards me, so I've press the button and held it on. She walked up to the tweeter, sniffed it for a couple of seconds and then laid down at my feet. So I hit her with the 8khz tone. She heard it but she never lifted her head. Looks like my fist test was a failure.

[Steve Plegge]

Sounds like an Irish Wolfhound. I'm not sure that those technically
qualify as dogs. You can't put a saddle on a dog and ride it...

Woof

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

[Steve Gibson]

Sean...

Got a chance to try mine out. My brother in law came around with his dog. A Irish something X, 8 months old. Not really sure what the dog is. It's the size of a small horse. Anyway I was sitting on my front verandah and the dog was sitting, back turned to me, about 2 - 2 1/2m away. I gave the 16khz button a quick press (0.2 seconds). The dog heard it as she turned around to see what the noise was. So I pressed it again for 1 second. She turn her head in a inquisitive look but it never seem to bother her. At this point she started to walk towards me, so I've press the button and held it on. She walked up to the tweeter, sniffed it for a couple of seconds and then laid down at my feet. So I hit her with the 8khz tone. She heard it but she never lifted her head. Looks like my fist test was a failure.

THANK YOU for that perfect testing report.  I'm also SO GLAD that you also tried the 8kHz (instant headache for us) mode, since I would have been dying to know about that otherwise.  And I'm so very glad that the v2.2.2 design allows for that so easily.

Your report is making me think that perhaps I'm wasting my time worrying about a fancy high-end tunable device... and that I would have been better served to build up a hundred of the v2.2.2 designs in order to get them into people's hands for testing.

We should remember that my original PDK was shot at point-blank range (somewhat courageously) directly into the face a charging rabid German Sheppard.  And I did that several more times in the following days until it no longer charged the fence.  So total surprise and ABJECT FEAR was doubtless a HUGE aspect of this.  The dog thought that IT was the "big dog", the pack leader, and that I was prey... and suddenly the tables were turned on it for the first time in its life by experiencing a powerful up-close and totally unexpected counter attack.  Those other factors may have figured more into this than we have accounted for.

As I have recently written elsewhere here (though lord only knows where) I am beginning to think that PROXIMITY to the animal may be KEY to the device's effectiveness -- since sound intensity goes up so quickly with every halving of distance.  If so, it would work (a) as a manually aimed and triggered first-person close-up tactical defensive weapon or (b) as an automatic barking deterrent device when it can be placed (presumably by the dog's owners), within very close proximity to a locally contained animal and respond instantly to a local loud bark.

I think our notion that the thing would bother a dog from a significant -- on a level of pain to induce it to cease barking -- is likely to be proven wishful.

/Steve.

[Kindanyume]

lmao

Indeed.. my friend had a massive IWH yrs ago.. the dog not even
really chubby and was weighing in at over 240lbs. Almost double my
wight at the time!

I gotta see if I can find some pix of that monster... oh and no one
tried to break into his place LOL

[Kindanyume]

shock and surprise can be a huge advantage to say the least esp w/dogs
of that nature.

[Vinny Valdez]

My apologies that I am not fully caught up on the research, but has the possibility of the dog(s) actually wearing the device been considered?  Our 2 dogs patroll our permiter annoying everyone and everything and I don't always have line of sight from my office window. This makes working from home very difficult at times. A remotely activated TQC collar would be wonderful.

Vinny

> I think our notion that the thing would bother a dog from a significant --
> on a level of pain to induce it to cease barking -- is likely to be proven
> wishful.
>
> /Steve.

--
Vinny

[Steve Gibson]

 

My apologies that I am not fully caught up on the research, but has the possibility of the dog(s) actually wearing the device been considered?  Our 2 dogs patroll our permiter annoying everyone and everything and I don't always have line of sight from my office window. This makes working from home very difficult at times. A remotely activated TQC collar would be wonderful.

When I was writing that I had the same thought.  Though I was thinking of having a bark-response approach, which would be VERY easy to do since the dog's own bark would be SO much louder than other ambient sounds.  So false triggering would not be much trouble.  And with something as small is the little Pyle surface-mount tweeters... it could be VERY small.  In fact... we could use BOTH of the little tweeters, one on each end.

/Steve.

[Vinny Valdez]

On Sunday, January 13, 2013, Steve Gibson wrote:

When I was writing that I had the same thought.  Though I was thinking of having a bark-response approach, which would be VERY easy to do since the dog's own bark would be SO much louder than other ambient sounds.  So false triggering would not be much trouble.  And with something as small is the little Pyle surface-mount tweeters... it could be VERY small.  In fact... we could use BOTH of the little tweeters, one on each end.

/Steve.

 

That would be fantastic Steve, I would love to test anything for you. I will submit my full story via your feedback link. 

Thanks,

Vinny

--
Vinny

[Steve Gibson]

That would be fantastic Steve, I would love to test anything for you. I will submit my full story via your feedback link.

Good Vinny, thanks.  That's my tester database.

/Steve.

[brian...@gmail.com]

No box yet, so, I just taped everything down.

My v2 will use a smaller speaker, assuming this one has an effect.

Cant wait for morning!!!

-Brian

[Steve Cirelli]

I got a chance to breadboard v2.2.2 before I went snowboarding Friday. Just now getting a chance to post some results. 

I have two huskies, very well behaved never make a sound. I'm actually building this because my mom has two very annoying Pomeranian's, not for my dogs. Anyway, they were in the room while I was assembling. When I testing it (not directed at them)  they didn't seem to care. One ever came up and sniffed the tweeter. So I'm not sure if v2.2.2 would work on my dogs. 

To be honest I'm not willing to test it any further on my dogs. I have sensitive hearing and my ears were ringing after the testing. My dogs will walk up to snarling barking German Sheppard's and Rottweilers without blinking an eye and try and sniff them. So it's possible that the noise just doesn't scare them. 

I'm going to mount everything on perf board then I'll try it on my mom's Pom's and post the results. They are much more skittish, so hopefully I'll get a reaction from them.

[Steve Gibson]

Brian...

No box yet, so, I just taped everything down.

My v2 will use a smaller speaker, assuming this one has an effect.

Cant wait for morning!!!

Unfortunately... it appears that you're not using a PIEZO TWEETER.

The current designs are ONLY compatible with Piezo Tweeters.  This is a non-negotiable.  The designs take deliberate advantage of the very different characteristics between piezo and electromagnetic tweeters.  When they are handled properly -- as we have now learned to do quite well -- the piezo tweeters are incredibly efficient.  And since we are able to use their natural capacitance as one of the critical circuit components... we get amazing design simplifications.

Now... that said... the variable-frequency design I am currently wrestling with WILL be able to drive either piezo OR traditional electromagnetic speakers.  It won't care since it's not using the tweeter's capacitance as part of the frequency generating and voltage boosting circuitry.

/Steve.

[Steve Gibson]

Steve,

I got a chance to breadboard v2.2.2 before I went snowboarding Friday. Just now getting a chance to post some results. 

I have two huskies, very well behaved never make a sound. I'm actually building this because my mom has two very annoying Pomeranian's, not for my dogs. Anyway, they were in the room while I was assembling. When I testing it (not directed at them)  they didn't seem to care. One ever came up and sniffed the tweeter. So I'm not sure if v2.2.2 would work on my dogs. 

To be honest I'm not willing to test it any further on my dogs. I have sensitive hearing and my ears were ringing after the testing. My dogs will walk up to snarling barking German Sheppard's and Rottweilers without blinking an eye and try and sniff them. So it's possible that the noise just doesn't scare them. 

I'm going to mount everything on perf board then I'll try it on my mom's Pom's and post the results. They are much more skittish, so hopefully I'll get a reaction from them.

Thanks for the report... and I look forward to more!

What I know is that my buddy Mark's problems have been completely solved with his next door neighbor's two little yappy miniature collies.  They run for cover (back into the house) at the first sound of the blaster I built for him.  But... it's likely FEAR rather than PAIN.

/Steve.

[brian...@gmail.com]

Yep. Ordered the wrong one originally, the giant magnet should have been a giveaway. I ordered 3 different types today to play with form factors once they arrive. Switching out the ‘horn’ will be the easiest part of the build.

[Kindanyume]

ahh and meanwhile.. the samples from one MFR arrived for me.. and the
dumb asses sent similar models.. claiming they are "piezo like" even
when asked explicitly if they were piezo or not. So much for the
sales dept in that company having a clue lol

Oh well I'll use the samples elsewhere.. or maybe see if I can mod the
horn portion to use on the pyle SM model

[Steve Gibson]

Kindanyume,

ahh and meanwhile.. the samples from one MFR arrived for me.. and the
dumb asses sent similar models.. claiming they are "piezo like" even
when asked explicitly if they were piezo or not.   So much for the
sales dept in that company having a clue   lol

Oh well I'll use the samples elsewhere.. or maybe see if I can mod the
horn portion to use on the pyle SM model

FWIW, the forthcoming variable-frequency design will drive electromagnetic tweeters just as well as piezoelectric ones, since it won't be relying upon the tweeter's natural capacitance to function.

/Steve.

[Kindanyume]

good to know. though these samples are obviously useless (as is) for
the "simple" version

[tr...@smartlivingenergy.com]

Here is my 1st attempt an V2.2.1 .  I can hear both tones, the higher when the red S1 button is press and lower when black S2 pressed.

My dog looks around at me, gives me east sh** eyes, and walks away.  

It is not as loud as I would like, and I would like it to work from longer distances (tuned 1 1/2" pipe rifle), but it was a fun build.

Tripp

[Steve Gibson]

Tripp...

Here is my 1st attempt an V2.2.1 .  I can hear both tones, the higher when the red S1 button is press and lower when black S2 pressed.

My dog looks around at me, gives me east sh** eyes, and walks away.  

It is not as loud as I would like, and I would like it to work from longer distances (tuned 1 1/2" pipe rifle), but it was a fun build.

Thanks for your report.

Unfortunately we don't have any good way to objectively compare loudness.

But... is the lower frequency not what you would describe as "incredibly loud" ??

My own experience (and other reports we've seen here) are that the "both buttons down" volume, when the thing is pointed at you, is REALLY LOUD.

That said... a big YES to the TrebleShooter Pipe Rifle design.  I'll get it prototyped shortly so that we can either get going on it or rule it out as a nice fantasy.

/Steve.

[Jon Katayanagi]

I built the first of ten last night. The 2.2.2 plan worked great!

Jon K

[Steve Gibson]

Jon...

I built the first of ten last night. The 2.2.2 plan worked great!

Is that a coffee mug?  That's Brilliant!!!  <<grin>>

/Steve.

[Jon Katayanagi]

That is an old drink cup I got from a dollar store, I think. It is like a plastic jar with a handle. The lid  is 3 3/8" OD and had a straw hole in it. I used a razor knife to cut the larger hole for the speaker. I rounded the speaker with scissors.

On the bottom it called "Fan Fueler". 

JonK

[sigpoggy]

Here is my V2.2.2 model. The box is a cheap project box from amazon (and I mean cheap as in crappy - but to its credit still functional). My big stroke of genius is using a little bubble wrap bag as the battery holder.

My main project is a remote controlled version (still in development), this one is for lending out to friends.

[Kyle Smith]

Sigpoggy,

It looks like a fine bit of craftsmanship. Do you have any data on its performance, or perhaps SPL levels reached at various distances? Or as Steve is fond of asking, is it loud?

Nice bit of work.

Best wishes,

--

Kyle Smith (AG2F)

air...@me.com

[sigpoggy]

Kyle,

I just measured and got the same results as when it was breadboarded:

8.5KHz is 112db at 1 meter

17KHz is 103db at 1 meter

Measurements made with iPhone app which is not calibrated (ball park calibration of "quiet space"), and at a distance of about 25 feet so it wont peg, then calculated for 1 meter. The drop for the higher frequency appears not to be due to freq response on the phone, the output voltage across the speaker has a corresponding drop.

[Steve Gibson]

Jim...

I just measured and got the same results as when it was breadboarded:

8.5KHz is 112db at 1 meter

17KHz is 103db at 1 meter

Measurements made with iPhone app which is not calibrated (ball park calibration of "quiet space"), and at a distance of about 25 feet so it wont peg, then calculated for 1 meter. The drop for the higher frequency appears not to be due to freq response on the phone, the output voltage across the speaker has a corresponding drop.

If you're curious, you might try dropping the value of the 22k resistor to 15k or 18k which might help with the higher frequency loudness.  The use of 22k was a deliberate compromise for the lower-frequency mode.  I didn't think that it would hurt the high-frequency mode... but I have more work to do there before I'll consider the design finalized.  (I also have all of the new inverters that I haven't yet started playing with.)

ALSO!!!!

What is the iPhone app you've been talking about?  It just occurred to me that IF we could establish some standards -- a device, the app, and distance/environment -- we WOULD have inter-builder numbers.  Also... I think that measurements need to be outdoors to that we don't get "room filling" effects.

So... what app???

/Steve.

[sigpoggy]

Steve,

I'll try the resistor tweak.

The app I'm using is Decibel 10th by SkyPaw Co, Ltd

It also works on my ipod touch.

Note: be sure to aim it properly, the mic it uses is on the bottom of the phone.

[sigpoggy]

Steve,

Also, the here is the website I use for the distance - decibel calculation.

[sigpoggy]

And one more note on the iOS app, there are various claims about the max level it can measure generally ranging from 100 db to 110 db. Mine pegs at 103db on both devices. Be sure to back off so readings are in the high 90's to be safe.

[sigpoggy]

Steve,

I put a 50K pot across the resistor and reduced it resistance while watching the scope. Both the high and low frequencies went down in amplitude as I lowered the resistance. Its not so easy to try increasing it.

[Steve Gibson]

Interesting!

So you're saying that 50K was louder for you than the original 22K ??

I was optimizing for waveform purity and absolute amplitude, not absolute loudness... So perhaps there's a better strategy. :)

/Steve. (Wandering around loose somewhere)

---

From: sigpoggy <sigp...@gmail.com>

Sender: portable-so...@googlegroups.com

Date: Fri, 18 Jan 2013 16:21:48 -0800 (PST)

To: <portable-so...@googlegroups.com>

ReplyTo: portable-so...@googlegroups.com

Subject: Re: [PSB] Re: Constructed Devices Show & Tell

[Brian Hall]

Cool app, thanks!

16KHz @ 1m: 94 dB

8KHz @ 1m: 104 dB

Measurements taken outdoors.

Brian Hall

[sigpoggy]

No, I put the 50K  pot in parallel with the 22K, thus reducing the resistance. I then continued to reduce resistance by turning the pot so its resistance diminished. There was no gain increase, only gain decrease as I effectively reduced R1. I am unwilling at the moment to try to increase R1 as it requires some work to take the thing apart.

[Steve Gibson]

Got it!! Thanks!!

/Steve. (Wandering around loose somewhere)

---

From: sigpoggy <sigp...@gmail.com>

Date: Fri, 18 Jan 2013 17:00:54 -0800 (PST)

To: <portable-so...@googlegroups.com>

Cc: <steve....@gmail.com>

[Álvaro Prieto]

I just got my parts today,so this is just a breadboard prototype. I'm using the pyramid TW-105. I can hear both the 16kHz and the 8kHz. To me, the 8kHz is much more noticeable, but I guess that's expected.

I don't have a dog to try it on, but birds on the power line above the sidewalk don't seem to react.

Not sure if this was brought up already, but keep in mind the frequency response of the phone's microphone might not always do well at 16kHz. http://blog.faberacoustical.com/2009/ios/iphone/iphone-microphone-frequency-response-comparison/

Never mind, looks like it's just the first models that had the not-so-good mics: http://blog.faberacoustical.com/2010/ios/iphone/iphone-4-audio-and-frequency-response-limitations/

Cheers,

Alvaro

[Jon Katayanagi]

[sigpoggy]

Álvaro,

Interesting, your signals are about the same amplitude. Here are my readings:

[stede....@gmail.com]

OK, so I've been lurking and getting up to speed on the design while doing some other projects.
My MOSFET driver came today so I built up a test circuit of v2.2.2 (minus the power bypass cap)...

Initial volume measurements with an Extech 600 @ .5m are 107.1 dB @~8kHz. This meter rolls off at 8kHz so the lower ~99 dB @~16kHz I measured is expected. I have a calibration microphone that is nearly flat to 20kHz and can measure upwards of 130dB. A little fiddling with stuff and I should be able to get real readings for this set of components.

I noticed there is an audible click noise as the power is applied and removed that's only apparent with the high frequency setting. Also, the folded-over kitchen towel + mouse pad sound trap is patent-pending :P

I plan to test the tube idea - I have some 1.5" SCH40 PVC (which is 40.0mm inside diameter for reasons I do not understand) that I was going to try. Anyone think this is a good size? I have a range of cardboard tube cores from large paper rolls to play with as well (from Ø 2" to 6" or so).

Also what about removing the horn part of the clone speakers? It seems like this would make it comparable to the surface mount version, but if anyone already has data I would like to hear it.

+s

[sigpoggy]

Hi stede,

Have you taken voltage measurements across the tweeter?

I have taken the horn off mine and one problem is the horn holds the speaker in place with regards to its back housing and terminals. Either that has to be glued or clamped (or eliminated?). The diameter of the speaker is about 1.5".

I wonder if the click is due to the missing cap.

[Steve Gibson]

 

I wonder if the click is due to the missing cap.

I was thinking thinking that perhaps the click is always there, but since the 16+ khz tone is so much less audible, the click is more audible.  It would be virtually impossible to hear any sort of click at the start of the 8khz blaster tone!  :)

/Steve.

[Steve Gibson]

Jim...

I have taken the horn off mine and one problem is the horn holds the speaker in place with regards to its back housing and terminals. Either that has to be glued or clamped (or eliminated?). The diameter of the speaker is about 1.5".

At the back of the tweeter there's a black paper cone permanently GLUED to the center of a floating piezo disc.  Are you saying that your black paper cone is loose and is being held in place by the front and back plastic pieces being clamped together by the three screws?

It so... your speaker has died.  This is one of the ways they die.  The piezo disc breaks free from the black paper cone due to the excessive mechanical stresses.  

Perhaps that not what you mean... but if it is, you need some good tweeters!  :)

/Steve.

[Stephen]

Hi Jon

Great work and great photos..thanks for posting.
Slightly puzzled by the inductor, seems different to the one suggest by Steve in v2.2.2 any reason for this ?

Regards Stephen

[Jon Katayanagi]

I ordered the parts a while ago, maybe the 2.2.1? I used the part numbers from the drawing. 

JonK

[Steve Gibson]

Stephen & Jon...

Great work and great photos..thanks for posting.
Slightly puzzled by the inductor, seems different to the one suggest by Steve in v2.2.2 any reason for this ?

At one point I was playing with various families and lines of inductors, and I briefly felt that the black "axial" inductor (which was just a bit less expensive) might be a superior choice. But it performed identically to the blue "radial" inductor.  Since the "blue tank" was overall physically smaller, though identical in performance, I decided to switch back to it just for reasons of physical packaging.  So it make zero difference to the performance of the system. the "blue tank" is a tad more expensive, but I prefer its packaging.  :)

/Steve.

[hausm...@gmail.com]

Posted yesterday in the Q&A thread while I was trying to get it to work on the breadboard.    I got it working and put it on perf board this morning.  It went well and I am happy.  I would do the project box different next time, but this was a good first time at making anything.     Have not seen any barking dogs yet today, but maybe later.   Attached pictures of my progress.

[duf...@gmail.com]

Hello all. I wanted to share my build. It is version 2.2.0 residing in a 3D printable package.

I made a few mistakes along the way. I ordered an absolutely massive push button on accident but it looks nice and inviting on the top. The cavity in the handle was just barely big enough to fit everything. I can't really get the battery out of the bottom without taking the handle apart. I didn't create any holes for the button or the tweeter hook-ups so I had to drill. If you have ever drilled a 3D printed object you know it just kind of splinters with a normal bit. I printed on a Thing-O-Matic which are an older technology and don't produce quality prints most of the time. I had to do some Dremel-ing to get things to fit.

All that said, I think it turned out pretty well. I've attached the 3D files if anyone wants to print them or improve them.



-Griffin

[Thomas Couey]

Steve,

I know that you were looking to profile as many speakers as you could to find out if the less expensive ones worked as well as the premium ones. Forgive me if you already have this data, but I haven't had time to completely read through all the posts. Anyway, I just got my speakers today as I had to have my parents on the mainland re-mail them to me because the seller will not ship to Hawaii (very annoying).

I used one of these: 4 Pieces of Piezo Tweeters Replacement for KSN1005A Motorola T1005-4

And attached is a scope trace of the output. I don't have a decibel meter, so I can't speak to the actual volume, except that it's very loud.

In case you can't see it in the re-sized picture, it's 44.8V peak-to-peak, at 16.81kHz. It looks like it has a bit of a DC component, but I suspect that's to be expected. I didn't measure the capacitance or resistance on them, but if you want that data, I can.

I literally just received these and threw the parts on the breadboard, so I don't have a finished unit to show off yet, but I'll post a pic when I do.

/Tom.

[Steve Gibson]

VERY nice!

That's a perfectly sized little protoboard too.

/Steve.

[Steve Gibson]

What a neat looking contraption!!  Ha!  I love it!

/Steve.

[Thomas Couey]

Funny you mention it, I just got it to add to my collection. This is my first use of it, super cheap and Amazon Prime eligible.

BB400 Solderless Plug-in BreadBoard, 400 tie-points, 4 power rails, 3.3 x 2.2 x 0.3in (84 x 55 x 9mm)

/Tom.

[stede....@gmail.com]

On Friday, January 18, 2013 11:38:46 PM UTC-5, sigpoggy wrote:

Hi stede,

Have you taken voltage measurements across the tweeter?

On my trusty analog o-scope I'm getting:

45.0V p-p at 15.4kHz 

53.5V p-p at 8.33kHz

[stede....@gmail.com]

Just recorded the output on my C01U at my desk. I attached a sample of the beginning and end of each waveform - you can hear the click sound and see the un-loaded resonant frequency. I'm getting closer to 8.2kHz and 16.4kHz looking at these.

This is the startup of the oscillator (48kHz sample rate)

+s

[Thomas Couey]

Here it is on a protoboard:

And the back (I haven't cleaned off the the flux yet...):

I guess I need a case now.

I tested it on the neighbor's dog with somewhat disappointing results. There is foliage and a vinyl fence between the sidewalk and the backyard where he is. It did sound like he backed up a little, but he was still barking. I pulsed it for about a second after each bark, but he kept on. I ended up with a headache and ringing ears. I think I'll use earplugs next time I use it (even outside, pointed away from me).

I noticed that the freq. shifted up a few hundred Hz on the soldered version (just over 17kHz). I wonder if I should add a little capacitance to the output to drop it down a few hundred cycles. Maybe I'll see if that makes it any more effective. Worst case, maybe I'll just drive the tweeter with a square wave and see if that is more annoying.

-Tom

[Steve Gibson]

Just recorded the output on my C01U at my desk. I attached a sample of the beginning and end of each waveform - you can hear the click sound and see the un-loaded resonant frequency. I'm getting closer to 8.2kHz and 16.4kHz looking at these.

This is the startup of the oscillator (48kHz sample rate)

The problem is... unfortunately... that 48khz is not a fast enough sample rate to capture the true waveform of a 16khz sine wave.  That sample rate is only three times the period of the waveform you are trying to capture.  So you'll only be able to see three points along one entire cycle of the sine wave.  You CAN get a feeling for it -- absolutely -- but not a picture of the waveform.  It's just moving too fast.  (The lower-frequency mode ought to be better.)

/Steve.

[Steve Gibson]

Tom...

Here it is on a protoboard:

That's a TERRIFIC little protoboard.  Where's you find it?  I may recommend it for this project.

Thanks!!

/Steve.

[Steve Gibson]

And... PLEASE keep experimenting!

(What AC voltage (RMS or P-P) are you seeing across the tweeter in your soldered version??)

/Steve.

[Steve Cirelli]

Here's pics of my first attempt at v2.2.2. Been forever since I soldered anything so it's ugly but it works. Now I need to buy a battery holder, and make a case for the entire thing.

[Steve Gibson]

Steve...

Here's pics of my first attempt at v2.2.2. Been forever since I soldered anything so it's ugly but it works. Now I need to buy a battery holder, and make a case for the entire thing.

Yeah... I've made a note to add a 9v battery clip to the next version's Bill of Materials.  :)

/Steve.

[stede....@gmail.com]

On Sunday, January 20, 2013 12:26:07 PM UTC-5, Steve Gibson wrote:

The problem is... unfortunately... that 48khz is not a fast enough sample rate to capture the true waveform of a 16khz sine wave.  That sample rate is only three times the period of the waveform you are trying to capture.  So you'll only be able to see three points along one entire cycle of the sine wave.  You CAN get a feeling for it -- absolutely -- but not a picture of the waveform.  It's just moving too fast.  (The lower-frequency mode ought to be better.)

/Steve.

Right, and I could see on the waveform I wasn't getting a sine-shape at that samplerate. Another try at 192/24 using my EMM-6 and Alesis interface:

The signal seems to stabilze after ~1ms.

Biggest thing to take note of is that the 16kHz signal is -6dB relative to the 8kHz (that's 1/4 the power). This is as predicted by your most recent post in the other thread regarding use of the 22k resistor. I'll play with a few values and see what I can get.

+s

[Steve Gibson]

Ah!  VERY nice!  Now we're talking!

I'll be very interested to see what you find with various values for R1.

The ONLY thing to watch for is that the inverter's input is not over-driven by a too-low value.  I would think that anything larger than 4.7K would be okay, but larger than 10K might be better.

/Steve.

[Thomas Couey]

It's from a Chinese company that I've bought quite a lot of parts from on eBay. They now have a their own web store.

1 Panel Of Double -Sided Prototype PCB, 576 Holes (CS-BP002)

The price is $1.99 with free shipping, if you use the Hong Kong Post Office ;-)  I thought you'd like that. Everything I've bought from them came through Hong Kong Post with no problems, it's just a little slow (customs and such).

One thing to note is that they're break apart boards, so you actually get four for that $1.99. I've left them connected for larger projects, but for things like this, the quarter panels are great. So, it works out to $.50 a board, if you can deal with the shipping time (anywhere from about 1-3 weeks, depending).

When I bought them on eBay (a while back), they came in 5 packs:

-Tom

[Thomas Couey]

44.6 VAC pk-pk, 15.0 VAC RMS, 17.07 kHz, at present.

I've confirmed that it doesn't seem to bother my neighbor's dog, so I might try a 555 driving a transistor with a 27V power source (something closer to your original PDK). I may need to look at a different transducer though.

-Tom

[Thomas Couey]

I forgot to mention that my handheld meter reads 6.265 VAC RMS, I'm not sure why there's a discrepancy (maybe the frequency is beyond the meter's rating?). I'll have to look into that, so I'd trust the scope's reading.

-Tom

[sigpoggy]

I just did some new measurements while testing various ios devices.

First difference is I put a brand new battery in.

Second, I took the reading in a different context - the ios devices where about 10 feet outside my front door, the PSB was across my patio out the back door and the living/dining room and a hall way to the front door were in the middle - total 60 feet.

I tested 8KHz which is a pretty clean sine wave and ended up with the three newest devices agreeing on 92 db which calculates out to about 117 db at 1 meter.

I gave up on 16KHz, the readings were at least 20 db lower and were all over the map. Subsequent testing showed differences due to device type.  I hear a spritzy sound, too deaf to hear the 16KHz, but a friend says he hears it in his spine.

Here is a break down of the audio. I got this using a free windows program called Soundcard Scope.

[Kindanyume]

Don't forget.. the Mic in a phone is way off overall compared to a proper mic

sad but very true :(

[g...@watkins-home.com]

built V 2.2.2 on a breadboard.  The cat ignores it.  The dog cocks its head.  My wife complains.  22 year old daughter, upstairs in her room, says it hurts.  I can hear the thing, but it does not seem loud to me.  In 8K mode it is really loud to me.  Now using ear protection!

Can't wait to test it with the neighbors dog.

[stede....@gmail.com]

So I found a 25k pot that was amenable to fitting in my breadboard and put it in series with a 10k giving me ~10-35k to play with.

I checked the input side and I am getting 8v p-p at 12.25k, which goes down as resistance goes up (as it should).

I was all ready to make a spiffy graph of resistance to voltage or dB, but it would look like this --------- because:

For any value of R1 from 12.25k to 25k my tweeter voltage is ~45Vp-p. A second run through with the scope disconnected and using my dB meter and a sound trap gave me 73dB±0.5 (which is equivalent to 99dB in open space).

I also did some testing at 8kHz and although I could hear a pitch shift in the signal (which is visible on the scope as well) there is no significant change in amplitude.

So, capacitance. The only thing I had handy was another 4700pF so I connected it in series which gives me 2350pF and I wound my pot up to 24.5k to retain my 8vp-p input and I got... 74.5dB. Dropping down to 22...@8.5Vp-p on the input I topped 75dB in the trap @50Vp-p on the tweeter.

Guess we need to test more capacitor values (or do some maths), I have some 1μF tantalum around here somewhere...

Also, any lower resistance and I could see the clipping on the scope as it reached 8.7 which jives with the datasheet for the 4421 (Vdd-.3V max).

+s

[g...@watkins-home.com]

I used xoscope (software) and it shows about 13,000Hz and 8,000Hz.  I tested a 15,000Hz sound source and xoscope reported 15,151Hz, so I think xoscope is giving reasonable results.  So, if 13,000 and 8,000Hz do not "train" the dog, what should I change to increase the frequency?

Wait, I tried another speaker (bought 4) and I get about 15,800Hz and 8130Hz.  I did not expect that.
#3 gives me 15,600Hz and 8,130Hz
#4 gives me 15,800Hz and 8,130Hz

I guess 1 of 4 is out of spec?
I would like to increase the frequency so that wife and kids can't hear it.

Question, if the frequency is too high to hear, can it still cause hearing damage?  Silent but deadly (damaging)?  :)

[Thomas Couey]

Yes, high frequency can damage your normal hearing. It has more to do with the volume (decibels) than the frequency.

Noise-Induced Hearing Loss (NIHL)

[Kyle Smith]

Guy,

Your concern with respect to hearing damage is well founded.  One thing is certain, if your daughter was reporting pain associated with the higher frequency, you may have already caused inadvertent high frequency hearing loss in her.  It is never safe to be in an environment in which the sound pressure level is causing frank ear pain.

As to your question regarding hearing damage at frequencies an individual cannot hear, but are still within the range of normal hearing for a human, the answer is that it most certainly can cause increased high frequency hearing loss, which is normally associated with aging.  In particular, it could cause damage to your daughter's very acute high frequency hearing, which is one of the great benefits of being young, causing her to prematurely loose that acute high frequency hearing, which adds a great deal of enjoyment when listening to music and increases the intelligibility of speech. Consonants are relatively speaking, higher in frequency than are vowels, and the loss of high frequency hearing therefore, effects the detection of those consonants.  

However, I am not your doctor, and I do not consider myself to be an expert in this area, so I advise you, and any others who are concerned about hearing damage to discuss it with your physician, and to research the topic independently, if you feel comfortable reading academic medical journals.  OSHA is a government agency which has set standards for acceptable exposure to sound, primarily to regulate sound exposure within the work place. An excellent resource would be to talk with an ENT, although this is a pricey option to undertake for an individual.  Audiologists, who may work in accord with an ENT, are another excellent source of information, and are likely much cheaper to talk with.

Stay healthy and happy.

Best wishes,

 

--kyle AG2F

Kyle Smith, MS, MD (retired)

air...@me.com

[Steve Gibson]

So, capacitance. The only thing I had handy was another 4700pF so I connected it in series which gives me 2350pF and I wound my pot up to 24.5k to retain my 8vp-p input and I got... 74.5dB. Dropping down to 22...@8.5Vp-p on the input I topped 75dB in the trap @50Vp-p on the tweeter.

Ah... that's better.  But getting it up to around 70Vp-p would be perfect.

Have you measured the voltage on your 9v battery while the sound it being generated?  I realized that my testing had been done with my lab bench supply and not enough testing with an actual 9v transistor radio battery.  I'll be switching to pure battery power from now on.

Also, any lower resistance and I could see the clipping on the scope as it reached 8.7 which jives with the datasheet for the 4421 (Vdd-.3V max).

Exactly right.  Though it's a function of BOTH the R and the C... since a larger C with a lower R would also limit the filter's output amplitude.

/Steve.

[Steve Gibson]

I used xoscope (software) and it shows about 13,000Hz and 8,000Hz.  I tested a 15,000Hz sound source and xoscope reported 15,151Hz, so I think xoscope is giving reasonable results.  So, if 13,000 and 8,000Hz do not "train" the dog, what should I change to increase the frequency?

Wait, I tried another speaker (bought 4) and I get about 15,800Hz and 8130Hz.  I did not expect that.
#3 gives me 15,600Hz and 8,130Hz
#4 gives me 15,800Hz and 8,130Hz

I guess 1 of 4 is out of spec?

Yes.  I purchased 250 of the "El Cheapo" tweeters expecting to select from them and find around 100 good ones.

 

I would like to increase the frequency so that wife and kids can't hear it.

First of all...

I deliberately pitched the device at around 15 to 16 khz so that it WOULD be marginally audible to people.  I worry about people not appreciating that it is generating powerful acoustic energy if they cannot hear it at all.  The lower-frequency button was designed SPECIFICALLY to help people understand just how loud the thing really is.

But also... we don't really know how the tweeters fall off at the high end.  So pushing the frequency much higher might result in less actual energy output.

All that said, the way to move the frequencies (both high and lower) upward would be to use a slightly lower value of inductance which would increase the oscillator's frequency.

THIS: (Digikey: AIUR-02H-821K-ND) is the inductor I would choose. It's 820uH which ought to move the 15,800 hz up to arond 19,270 hz... which ought to still be (barely) within the tweeter's frequency response, yet also out of your family's.

/Steve.

[sigpoggy]

I have been playing with the circuit and have come to the conclusion that the output frequency is the main determiner of the output voltage across the tweeter. Playing with R1 an d C2 have a small affect on amplitude and frequency. Picking a value for C2 then tweaking R1 with a pot you can find a sweet spot but you only gain a few volts.

Adding capacitance across the tweeter to change the tank frequency causes the amplitude to increase dramatically as you go down in frequency. Again, R1 C2 adjustments at lower frequencies cause minimal change.

Pulling the frequency down to 15.7 kHz and tweaking the feedback gave me a high of about 45vpp - (vs 39vpp at 16.9kHz). Don't have other inductors to play with, but seems that would not really change the dynamics.

[stede....@gmail.com]

On Tuesday, January 22, 2013 2:14:31 PM UTC-5, Steve Gibson wrote:

 

Ah... that's better.  But getting it up to around 70Vp-p would be perfect.

Have you measured the voltage on your 9v battery while the sound it being generated?  I realized that my testing had been done with my lab bench supply and not enough testing with an actual 9v transistor radio battery.  I'll be switching to pure battery power from now on.

/Steve.

I tried a couple values for C2 down to 1000pF (where the circuit stops resonating, BTW) and I tried a slightly larger and smaller inductor (I used speaker crossover parts with very low DC resitance, <0.5Ω). I don't see much more than 50vp-p at any point.

The battery drops as much as 0.4V under load. Reducing the input voltage with larger values of R1 reduces the voltage drop (less current). If a nominal 9V battery has a Ri of 1.5Ω that should indicate 600mA load, but the actual power draw is an order-of-magnitude less at ~62mA. (Then there is also the possibility I suck at math).

So there's that...

+s

[Álvaro Prieto]

Wow, just received the Goldwood GT-1005. What a HUGE difference it makes compared to the Pyramid TW105! It's about 20Vpp more, which is definitely noticeable.

Alvaro

[sigpoggy]

I notice a big difference in frequency. I have found the lower the frequency the higher the amplitude.

[Steve Gibson]

Wow, just received the Goldwood GT-1005. What a HUGE difference it makes compared to the Pyramid TW105! It's about 20Vpp more, which is definitely noticeable.

That's REALLY interesting.  Thanks for posting and sharing that Alvaro!

And... yes... 20Vpp would DEFINITELY make a HUGE difference.

/Steve.

[Steve Gibson]

I notice a big difference in frequency. I have found the lower the frequency the higher the amplitude.

That MAY be supporting my contention that we're seeing high-end bandwidth falloff at the higher end.

In general, though, I really need to figure out why no one else has been seeing the 70Vpp voltages that I have been.

/Steve.

[Steve Gibson]

I tried a couple values for C2 down to 1000pF (where the circuit stops resonating, BTW) and I tried a slightly larger and smaller inductor (I used speaker crossover parts with very low DC resitance, <0.5Ω). I don't see much more than 50vp-p at any point.

The battery drops as much as 0.4V under load. Reducing the input voltage with larger values of R1 reduces the voltage drop (less current). If a nominal 9V battery has a Ri of 1.5Ω that should indicate 600mA load, but the actual power draw is an order-of-magnitude less at ~62mA. (Then there is also the possibility I suck at math).

So there's that...

The battery drop could easily explain the differences we're seeing in output amplitude since I have always been using a regulated 9v supply.  I'll be switching over to 9v batteries from here on out.

/Steve.

[Steve Gibson]

Jim (at al)...

I have been playing with the circuit and have come to the conclusion that the output frequency is the main determiner of the output voltage across the tweeter. Playing with R1 an d C2 have a small affect on amplitude and frequency. Picking a value for C2 then tweaking R1 with a pot you can find a sweet spot but you only gain a few volts.

Yes. Thanks for confirming my findings.  That matches my theory and design.  Those SHOULD be non-critical, thus their ability to operate usefully both at 16khz and 8khz... a comparatively HUGE difference.

Adding capacitance across the tweeter to change the tank frequency causes the amplitude to increase dramatically as you go down in frequency. Again, R1 C2 adjustments at lower frequencies cause minimal change.

Yep.

Pulling the frequency down to 15.7 kHz and tweaking the feedback gave me a high of about 45vpp - (vs 39vpp at 16.9kHz). Don't have other inductors to play with, but seems that would not really change the dynamics.

Jim... others have (recently) reported seeing larger tweeter-to-tweeter variations than I encountered in my testing.  Have you tried more than one tweeter?

/Steve.

[sigpoggy]

It has occurred to me that the speaker is the final arbiter of output voltage level.

I only have two, they are the same model: Pyramid TW105. No difference between them.

I have others on order and they should be arriving soon.

Speaking of speakers I seem to remember you saying that the version you built for your friend had 4 speakers. Is that correct? 

If so, that would mean it would be producing a hell of lot more sound that these devices assuming the output of each speaker is comparable.