amplify ultrasound signal

[ned707]

I'm working on a project that will produce ultrasound waves through a piezo tweeter in short bursts (target is 105dB).

I'm planning to use a "DIY function generator" kit that will produce the input signal I want. Its adjustable and I intend to tune it to a square wave of 8VPP (max) and 22KHz. The generator handles between 9 and 18 V of DC supply input.

The problem I am facing is how to amplify the generator signal (8VPP and probably low current) to drive a 200W piezo tweeter. I wasn't able to find out the max V/I that the tweeter can handle. The tweeter operates between 5KHz-50KHz, 200W max, 105dB, 4 or 8 Ohm.

I'm looking for a low cost solution, so don't want to use a home/car multi-channel amp.

I plan to drive as many as 4 or 8 tweeters.

Can this tweeter be driven to the max using some discrete transistors / inductors on a breadboard? I'm not interested in sound quality (won't even hear it) other than the tweeter to spit out an ugly 20+ KHz signal at about 105dB.

 

[BobK]

First, a tweeter cannot be rated as 4 or 8 Ohms, it is one or the other. Which is it?

To put 200W into an 8 Ohm speaker requires 40V RMS with is a whopping 112V p-p at about 1.8A.

For a 4 Ohm speaker it is 28V RMS or 78V p-p at about 2.5 Amps.

You can find integrated circuit amps that will do this. But, the transformers for the power supplies will cost more than the amplifiers, so if you can find a amp of these specs already built, I suspect it will cost you less.

Bob

 

[ned707]

Thanks. It sounds like I may need to find a low cost car amp (or such) due to the cost of transformers, for this application.

On a different note, I may need to find a different tweeter. The specs on this one seem sketchy:

http://www.audioelectronicwarehouse.com/index.php/pro-series-speakers/tweeters/pet-1103.html

I emailed (only way to contact them) their company but haven't heard back. I suspect they are waiting on a response from the mfc. Guess I shouldn't complain, they just sell them.

I searched for a data sheet but came up with nill. Any suggestions on a tweeter that can handle a solid 22KHz and a lot of dBs (105+)? (again cost effective... yes I know there are some $600+ ones out there)

 

[BobK]

I re-read your OP and I think you do not need the kind of power you were talking about. The sensitivity rating of a speaker gives the output in Db at 1 meter with 1 Watt input. Each 6Db after that will need double the power. So if you find a tweeter rrated at 90Db at 1W it will put out 96Db at 2W and 102 Db at 4W. So it is unlikely you will need anything like 200W to produce 105Db.

Note that the rating is at 1 meter from the front of the speaker on axis. How far away do you need to produce this sound level? And / or over what kind of area?

Here is a speaker seller that I have used. They are very knowledeable and can probably help you out in picking a speaker driver.

http://www.madisound.com/

Bob

 

[ned707]

Good to know! Thanks.

Did some math with the 1W/1m rule and I'm thinking either 32 or 64 Watts should provide sufficient reach for my application. For the super tweeter that I'm eying that should theoretically reach out to roughly 20 feet and provide the range of dBs that I've been hoping for. Ideally the wave coverage would be 180 degrees. 90 degrees would also be OK. I'm guessing the coverage is determined by the frequency and shape of the speaker housing. For my tweeter of interest it appears to spout to 90 degrees or so, but whether the sound really follows that path or not is beyond me. I think I should plan to incorporate 2 of these tweeters for starters, and if that doesn't function well enough, perhaps add 2 more later.

Would it be feasible / cost effective to build an amp that can supply 32, or maybe even 64 Watts (X2) using amp ICs? Perhaps there's an already built amp that does this?

I appreciate your time.

(I checked the site you are recommending. Only saw the other piezo as a possible alternative... but it's a tad weaker).

 

[twister]

So this begs the question, what are you doing?

 

[KrisBlueNZ]

You could look at H-bridge drivers like L293, L298, L6201/2/3, L6205/6/7 from STMicroelectronics. These provide a complementary drive at several amps and operate up to around 50V. The output seen by a tweeter conneted across the H-bridge is actually a bipolar square wave, which will probably damage the tweeter if used directly because of all the high harmonics, but it might be possible to use the impedance of the tweeter (which is not just its inductive impedance but also includes the effects of the physical movement of the cone) with one or more capacitors, to form a tuned circuit that will produce an effective drive much closer to a sine wave. This would probably only be workable if you're using a constant (or fairly constant) drive frequency and either driving continuously, or at least not turning it on and off at a high rate. I haven't had any experience with this, but I can try to make some further suggestions if you can't find anything with a Google search. It's a somewhat specialised area and looking at prior art would be very helpful.
Edit: PLEASE give a detailed description of what you want to do, and part numbers for any components you have decided on or are considering using.

 

[BobK]

The intensity is going to fall off as the square of the distance, so the intensity at 20feet on axis will be about 1/50th of that at 1m, i.e. -34db So I think you are going to have a lot of trouble getting 105db at that distance, and 180 degrees is also very difficult. You need to know the actual radiation pattern of the tweeter, they tend to be pretty directional.

Bob

 

[ned707]

Twister:
its a training system for some noisy dogs. The yard is about 50 x 30 feet.

Kris:
Thanks. I'll take a look at those suggestions. The function generator I'm planning to use can output several different wave forms (including sine). Are you thinking a sine wave would be safer to drive a tweeter? I'm thinking the frequency should be constant, and it should only output through the tweeter for 1 second or so after detecting each dog bark. I'll need a control switch from a mic to detect a bark and that will allow the signal to be fed into the amp.

Bob:
Are intensity and SPL different? They seem to both be measured in dBs. Since the tweeter only outputs for about 1s, could a large capacitor serve as an amp? Discharge on a bark, charge between barks? I have some big ones that I pulled out of a PC power supply, or they might have come out of a large CRT. BIG MOMMAS!

I realize there are products out there that do what I'm trying to build (retail for $50), and I have 2 of them already. They work pretty good too. These products seem to output ultrasound waves in a very directional manner however, so I would expect the tweeters I chose to do the same. That's another reason why I want to build my own: to have the flexibilty to add more tweeters where needed. Needless to say, the products I have contain a single tweeter.

Edit: wanted to add that I've opened up the product one I bought and it has a very small, flat looking speaker inside, and doesn't appear to have a fancy amp (certainly no heat syncs or anything). Just a small pcb, tweeter and a mic.

 

[BobK]

Intensity and SPL are the same.

I am reasonable certain that the units for $50 are not putting out anything like 105db at 20ft. Where did you get this number?

Bob

 

[roltex_rohit123]

is it that you are going to zap the dog's ears if he is barking? good idea!! hope we get something like this for our ministers!! why not use dog whistles?

 

[ned707]

I figured 105dB from a chart I saw on wikipedia. It says 105dB would be borderline sound discomfort, proportional to a jackhammer. 105dB may be overkill to a dog (I'm no vet), but id rather design for higher power and be able to tone down the amp, as needed, than to design it too weak and end up with a useless $100+ paper weight.

I found a kit audio amp that can do 2x30W max / 2x15W RMS. Handles frequency to 60KHz.

http://www.amazon.com/gp/aw/d/B0002Z82YO/ref=aw_d_detail?pd=1

I'm wondering if this will get the job done.

Edit: why does this thing need a transformer for input voltage? Why not just a dc power supply that I can plug in from an extension cord? Id prefer to use a simple dc power supply. Is it unusual to find such an amp?

 

[ned707]

is it that you are going to zap the dog's ears if he is barking? good idea!! hope we get something like this for our ministers!! why not use dog whistles?

Yes, and automatically... 24/7. It trains the dog not to bark in the sound field. After 1 or 2 barks they figure it out and stop barking, or will change their bark to be lower, so as not to set off the device.

I've tried whistles (actually a droid app). Not affective unless you plan to spend a lot of time in your yard. Also, this isn't a good way to train the dog. You need something more persistent. Every dog is different, "so mileage may vary".

 

[BobK]

I would worry that you might be harming the dog's ears. If it can produce 105db at 20 feet, think what happens when the dog is 2 feet away. You have about 135db now. In fact, it might even be harmful to human ears at that level.

There is no way the commerical product you describes is even within a factor of 1000 of what you are trying to do. If you insist on doing this, which I am beginning to think is a very bad idea, start small and work you way up to higher power. Start with 1W or less.

Bob

 

[ned707]

I would worry that you might be harming the dog's ears. If it can produce 105db at 20 feet, think what happens when the dog is 2 feet away. You have about 135db now. In fact, it might even be harmful to human ears at that level.

There is no way the commerical product you describes is even within a factor of 1000 of what you are trying to do. If you insist on doing this, which I am beginning to think is a very bad idea, start small and work you way up to higher power. Start with 1W or less.

Bob

The minimum distance will be about 10 feet. But your concern is well taken. That's why I'm wondering if there is an alternative, more strategic way to drive the tweeter, such as with a capacitor instead of a full blown amp.

As with anything experimental, I'll start with the minimum and work my way up. It's safer for the subjects and cuts down on electricity costs.

Thanks for all your help.

 

[CocaCola]

In fact, it might even be harmful to human ears at that level.

There are real life commercial devices (many military and law enforcement only) that produce extreme levels of ultrasound frequencies with the sole intent on disrupting the human body, so IMO there is a potential for negative side effects if you keep pushing for more power...

 

[twister]

I had a big lab that went with me a lot in my truck. He could bark so loud that it would make your ears ring, in the truck. I wondered what his own ears were doing! ha He was a good dog though. Whenever he barked you knew to look outside, because someone was around.
I would like to have one of those ultrasound things for my neighbor's dog that barks day and nite, for no reason.

 

[CocaCola]

I would like to have one of those ultrasound things for my neighbor's dog that barks day and nite, for no reason.

I tried one of the commercial ones the effect is really dependent upon the dog... My neighbor has several 'rodent' dogs that yip and yelp like insane idiots at nothing and everything all day and night when they are out... After installing the device on the side of my garage facing them, you would see the dogs I mean 'rodents' ears perk up after every bark, and then they would immediately start barking at the device itself as if they entered a feedback loop... It did curb the neighbors bigger dog that clearly found the noise annoying and headed in the opposite direction when the 'rodents' were stuck in their noise loop...

 

[ned707]

I tried one of the commercial ones the effect is really dependent upon the dog... My neighbor has several 'rodent' dogs that yip and yelp like insane idiots at nothing and everything all day and night when they are out... After installing the device on the side of my garage facing them, you would see the dogs I mean 'rodents' ears perk up after every bark, and then they would immediately start barking at the device itself as if they entered a feedback loop... It did curb the neighbors bigger dog that clearly found the noise annoying and headed in the opposite direction when the 'rodents' were stuck in their noise loop...

The trick I've discovered on the commercial products is to get rid of the 9V battery (generally used) ASAP and solder on (as needed) an appropriate DC power supply, connected to wall power. 9V/600mA has been working great on my current setup. The dogs soon realize that the buzzer ain't wearing down like it used to. Strong and steady, day in day out The problem is that you need at least 2 of these to have any good results, and they generally run $50 each (sale price too).

 

[CocaCola]

The trick I've discovered on the commercial products is to get rid of the 9V battery (generally used) ASAP and solder on (as needed) an appropriate DC power supply, connected to wall power. 9V/600mA has been working great on my current setup. The dogs soon realize that the buzzer ain't wearing down like it used to. Strong and steady, day in day out The problem is that you need at least 2 of these to have any good results, and they generally run $50 each (sale price too).

Mine was running off a wall wart 24/7