PDA

View Full Version : In need of an Ultrasonic Transmiter driver circuit



Brian Carpenter
04-13-2007, 04:24 AM
and also a circuit for the reciever. I would like to use it with my sx20.

i found this image of a circuit but am not sure how to impliment it with the sx.

I want the SX to trigger a burst from the Transmitter and have another sx increment a counter when it recieves the sonic.

Any ideas?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Mike Green
04-13-2007, 04:39 AM
The 555 timer has a reset pin that is normally high and brought low to reset the device. You can simply tie this to an SX pin and output a low until you want to start the ultrasound pulse, take it high, then low again when you want to inhibit the pulse again.

The receiver includes a comparator (the CD3140). You can attach the output of that to an SX pin and have the SX increment an internal counter whenever it sees a high to low transition (which would turn on the relay in the original circuit).

Brian Carpenter
04-13-2007, 04:48 AM
Thanks Mike, You are always here and that is great

is there anyway to do away with any of the parts and use the sx for the tasks?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Mike Green
04-13-2007, 05:03 AM
Sure. You could use the SX instead of the 555 timer, just supply a pulse train to the driver transistors. Depending on the transducer you're using, you might be able to drive it directly from the SX's output pin(s).

You might be able to do without the comparator by using the SX's switching threshold. I really can't help you much here. I'm just not that familiar with the SX.

Tracy Allen
04-13-2007, 06:09 AM
The ultrasound transducer can probably be driven directly from the SX output, and will have a higher volume if each side is driven by pins in opposite phase (for 10 volts p-p amplitude on a 5v supply.) The schematic does not indicate the transducer, but I am assuming it is a piezo single frequency device, or else a tweeter for lower frequencies. There are tricks to get higher volume if need be.

The receiver probably will need gain, depending on distance. The circuit as shown would be fine, with output taken either from comparator or directly from the detector outputs. If a tweeter is used, it will pick up a lot of out of band noise that could trigger the circuit, so it may be that a narrowband transducer will better fit the purpose.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com (http://www.emesystems.com)

Brian Carpenter
04-13-2007, 06:40 AM
Thanks Tracy and Mike

I am using this one from the Jameco.com site

136654:
Transmit/Receive Set
• Capacitance: 2000pF 30%
• Transmitter: bandwidth 4.0kHz @
112dB, sound pressure level 119dB @ 401.0kHz
• Lead length/spacing: 0.39"
• Receiver: bandwidth 3.5kHz @ -71dB,
min. sensitivity 65dB @ 401.0kHz (R=3.9kΩ)
• Case size: 0.47"H x 0.62" Dia.

I am looking to get about 20 to 25 foot range between at about a 8 foot circle. Do you think that it is possible?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Bean
04-13-2007, 06:46 AM
I thought that ultrasonic transducers needed much higher voltages ? Something like 40V ? Maybe I'm wrong...

Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Educate your children to self-control, to the habit of holding passion and prejudice and evil tendencies subject to an upright and reasoning will, and you have done much to abolish misery from their future and crimes from society (http://www.quotationspage.com/quote/29139.html)"

Benjamin Franklin
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
www.hittconsulting.com (http://www.hittconsulting.com)

Phil Pilgrim (PhiPi)
04-13-2007, 06:56 AM
A couple things to consider:

1. Is the frequency response curve of the receiving transducer sharp enough that no additional bandpass filtering is needed in the receive circuit to eliminate noise?

2. Normally, ultrasonic receivers use a TVG (time-varying gain) amplifier, in which the gain starts out low when the "ping" is emitted and gradually increases as it waits for the echo. This helps to prevent overloading on nearby objects, yet provides enhanced sensitivity for distant objects. Analog Devices offers amplifiers for this purpose (AD600, AD602).

-Phil

Brian Carpenter
04-13-2007, 07:32 AM
Phil,

the reciever is on a different device and is not being used in the typical Ping Echo configuration (ie on the same board) it will be on a seperate SX and always listening for any signal

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

N0QBH
04-13-2007, 09:44 AM
I've got a Devantech model SRF04 that makes clever use of a MAX232 to convert the 5v pulse from a PIC into a 20v p-p drive for the transducer ping.

Brian Carpenter
04-13-2007, 10:42 AM
Tracy, By higher volume i am also assuming higher voltage. my supply voltage is 4.8v. What kind of tricks can you offer?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Tracy Allen
04-13-2007, 01:59 PM
The data stated by Jameco is incomplete. What conditions were used to produce a pressure of +119 db at 40khz? It should usually be stated something like this:
-- Sound pressure level: relative to 0db is 0.0002 microbar measured at 10Vrms, 30cm distance
-- Sensitivity: 0 db vs 1 volt per microbar.

Note that figure of 10 volts RMS and 30cm (30.48cm = 1 foot). The 119 db figure is very loud and might have been produced by a higher voltage and at a different distance, but the data specs do not say. It is not unreasonable to drive these things at 10 to 20 volts RMS, and although they can be driven harder to deliver "pings", I have been told that it is hard on them for continuous excitation. There are other types of piezo or film transducers that are driven at the 100 volt level. I have a system in operation that drives the transmitter in the differential mode directly from the pins of the SX running from a 3 volt lithium cell, and the receiver can easily pick it up in a room similar in size to the distance you are asking about, even indirectly after reflections from the walls.

The threshold of human hearing is considered to be around 0.0002 microbars. 119 db is around the volume of a rock concert or jet engine and you might consider it quite deafening; if your ear were sensitive at 40khz. The range of hearing covers some 6+ orders of magnitude. Sound pressure over much of the range will be proportional to the voltage drive to the transducer. db = 20 log V2/V1 = 20 log P2/P1.

These transducers are strongly resonant and thus are very efficient at resonance. Conversely they are very ineffient far from resonance. That makes them also very selective as receivers, so for example a 40khz receiver transducer will pick up practically nothing of the noise or ambient sound outside of its narrow bandpass. It can however pick up the stray energy from noise sources that do fall in the bandpass, such as jangling keys or hissing noises.

One way to increase the output level and also to widen the bandwidth slightly is to put an inductor in series with the transducer. It is an electromechanical resonant system. An inductance of 3.3 to 10 millihenry is typical. On the receiver side, the inductance would go in parallel with the transducer. Other ways to increase the drive to the transducer is to use something like a CD4050 or the MAX232 to boost the drive voltage. I've tried small audio transformers, but the results have been disappointing.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com (http://www.emesystems.com)

Post Edited (Tracy Allen) : 4/13/2007 7:04:50 AM GMT

Brian Carpenter
04-15-2007, 10:23 PM
Tracy, Mike, and Bean



If i wanted to use the Max232 chip to increase my voltage from 10v p-p to 24v p-p would i just send the oscilating 2 pins from the sx into the input side of the max232 and then take the output side of the chip to the Ultrasonic transmitter. Is it really that simple?



edit: maybe it is just 20v p-p


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Post Edited (Brian Carpenter) : 4/15/2007 3:31:32 PM GMT

RocketScientist
04-16-2007, 04:07 AM
A year or two ago I considered using that circuit and fortunately trashed it. Mainly, the 555 performance is very sensitive to temperature fluctuations. And that whole diode power amp to the transducer sucks!! I would like to take some of the advice given and take it one step further:

Build your transmitter with a push-pull FET driver network. If you plan on using under 100volts on your transducer then you can buy the FETs as logic gated from Allied, Digikey, Jameco, etc. That means you can make 100v 40Khz waves with 5v from your Sx processor. I wrote my SX code to create 1 burst every 4 ms of 150V 40 KHz waves. I also made a burst pattern of 3-40 KHz with 1ms of wait. Measured with oscillascope my 40KHz wave was incredibly accurate by the Sx. Not only did the SX maintain wonderful accuracy I could program the chip to create virtually any pattern of burst-silence from 1 wave repeating to continuous. The reason you should use a push-pull FET network is that it is highly efficient - temperature won't mess up your timing source and maintian a crisp power wave on your transducer. Traditional transistors will mess you up that way - you will get fluctuating results with higher loads. If you want to use power greater than a hundred volts, you have to put a MOSFET driver between your SX and push-pull network. The MOSFET driver will take your Sx's square waves and make them taller to the tune of 35v +. This project was very neat for me because you ended up with a transmitter you had total control over switching up to 600V RMS of power with any burst pattern (poor transducer) driven by the SX. What more could you ask for? The quality of the wave was then dependent on the transducer quality.

If parallax is interested in obtaining a tested schematic of this please email me. I just have no use for it anymore as I moved on to another project.

Post Edited (RocketScientist) : 4/15/2007 9:12:27 PM GMT

Brian Carpenter
04-16-2007, 04:21 AM
i am interested in the schematic

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Tracy Allen
04-17-2007, 05:55 AM
Yes, it is that simple. You could also use just one SX pin by using one of the rx inverters. Feed the single SX output to both R1in and T2in, and feed R1out to T1in. Then T1out will be at +10 volts when T2out is at -10 volts and vice versa, as one SX pin is made high and low.

That said, it is not something I have tried myself and I fear the result might be a bit disappointing. The outputs won't be the full +/- 10 volts. The transducer is like a 0.01uf capacitor and the current flowing is directly proportional to that capacitance and the switching frequency and the voltage. At full 20 volts peak to peak and charging that 0.01uf at 40khz, it comes out to about 8 milliamps by my back of envelope calculation. I don't have the figures on the load curve for the MAX3232 handy, but I think the voltage would be more like +8/-7 volts, not the full +/- 10 volts. If you have one around, it would be worth a try.

Question: do you have a higher power supply voltage available, like 9 or 12 volts? It would make more sense to power the push pull circuit directly off that if so. The CD4049 is a hex buffer chip that is good for that sort of thing.



Brian Carpenter said...
Tracy, Mike, and Bean
If i wanted to use the Max232 chip to increase my voltage from 10v p-p to 24v p-p would i just send the oscilating 2 pins from the sx into the input side of the max232 and then take the output side of the chip to the Ultrasonic transmitter. Is it really that simple?
edit: maybe it is just 20v p-p

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com (http://www.emesystems.com)

Phil Pilgrim (PhiPi)
04-17-2007, 06:13 AM
Brian,

Take a look at the Micrel MIC4428 (http://www.micrel.com/_PDF/mic4426.pdf). It's sold as a MOSFET driver, and it's designed for switching high-capacitance loads at high frequencies. Vcc ranges up to 18V. It has two drivers in an 8-pin package: one inverting, the other non-inverting. By connecting the inputs of both drivers to the same uC pin, you can get a high P-P output with very little skew. DigiKey carries them.

-Phil

RocketScientist
04-18-2007, 09:13 PM
I'll post the schematic. Give me a few days. We just moved and I have to dig it out. I'm sorry about the delay.