Hi,

First I should describe my project. It is basicly a range calculator. I am using 38khz IR led to activate TSOP 1138 sensors. I have 4 sensors and the IR light is coming to some of these sensors depending on the distance that the IR light is reflecting.

My question is, how can I obtain sound like a parking sensor. I want the beeps' frequency to be higher when the object is closer.

I thought of getting diffrent frequencies from each sensor by using op-amps but how can I set the circuit such that I can use only the higher frequency signal?

Or do you have any other ideas to design the circuit?

Thank you already.

 

I have a similar problem. Could you solve it

 

Hi,

First I should describe my project. It is basicly a range calculator. I am using 38khz IR led to activate TSOP 1138 sensors. I have 4 sensors and the IR light is coming to some of these sensors depending on the distance that the IR light is reflecting.

My question is, how can I obtain sound like a parking sensor. I want the beeps' frequency to be higher when the object is closer.

I thought of getting diffrent frequencies from each sensor by using op-amps but how can I set the circuit such that I can use only the higher frequency signal?

Or do you have any other ideas to design the circuit?

Thank you already.

What is your application? Are you looking for a parking sensor system such that as you pull into your garage it gives you an indication of how far from the wall you are?

hgmjr

 

Probably, a better solution would be a constant tone freq that's pulsed. The closer the object is the more rapid the rep-rate until you reach a danger point of a continuous tone.

 

How about using a PIC?

Different values for different sensors, giving differenet input, and then again different sound...

Just thinking loud...

 

The whole circuit could be easier with a PIC but it is forbidden. I have to all this by using op-amps and capacitors etc.

The application is that the closer you are to the wall, higher the frequency of beeps.

I can have 5V from each of IR sensors, but how can I set the frequency depending on how many sensors are active?

I couldn't understand what CDRIVE said, can you please explain it?

 

I think he meant something like this:

| = tone in one frequency
- = no tone/silent

At some range:
|---|---|---|---|---|---|---

At some closer range:
|--|--|--|--|--|--|--

Even closer:
|-|-|-|-|-|-|-

Close:
|||||||

Thats what I think he meant....

 

Yes, this is what I supposed to get, but I can't find a way to change the timing between two tones with respect to the sensors I mentioned. Any ideas?

 

The cart may be before the horse. Have you got a successful means of using the IR source and the sensors to measure distance? How are you doing that?

 

I may be way off, but how about using 555 timers to get the tone, triggered by the different values on the resistors to the different sensors....?

 

@beenthere
I am using a Schmitt trigger to produce a 38khz square wave for the IR led. By using another op-amp I turn that of after its 6 duty cycles, then I turn it on in 6 duty cycles. Because this is what TSOP 1138 requires. I don't think I have a mistake there but If you disagree I would appreciate your help.

@nerdeguetta
As I told, I can't even use 555 timers, they are all banned from project.

 

I am very skeptical that any measurement can be made using time-of-flight for light as the method. Light travels at approximately 1 foot per nanosecond. I don't think there is a practical way to discriminate nanoseconds without fairly complex electronics.

I think you will have better results using sound. Specifically the use of ultrasonics is a common way to tackle this problem.

hgmjr

 

@hgmjr

I would love to use a ultrasonic sensor, but as I said, the project is a IR project. I don't use time-of-flight anyway, I am using the angle of reflection to activate more sensors, so I calculate approximately the distance.

We are getting away from my question, I am looking for a solution for changing the time interval between two beeps by using 4 sensors whose output is 5V each.

 

I think you will have better results using sound. Specifically the use of ultrasonics is a common way to tackle this problem.

hgmjr

If I were doing this I would to want to employ the same basic concepts that a common depth finder uses but without a display. One of the key differences will be to shorten the ultrasonic burst (to the transducer) for very short echoes. Experimentation between a compromise in energy output (Burst Width) vs ability to measure very close objects. This is because the output burst must have ended before the echo is returned.

 

Sum the voltages from the four sensors and use the sum as the input for a voltage-to-frequency convertor circuit, or a voltage-controlled oscillator (VCO). There is a schematic for an audio-frequency VCO at http://en.wikipedia.org/wiki/Voltage-controlled_oscillator . I hope you can also use an FET. The LM324 and LM358 datasheets also have VCO schematics. For all of the other basic opamp circuits you need, including a multi-input summing amplifier, download AN-20 and AN-31 from national.com .

OK, I did a little digging and this sounds like exactly what you need, and uses only opamps: Look at the circuit at the top of Page 9, in http://www.national.com/an/AN/AN-31.pdf . If you connect the second opamp's + input to a positive voltage instead of ground, it lowers the output frequency. So you'd have to subtract the sum of your sensor voltages from the positive supply voltage (or something close to it), in order to get a rising tone as more sensors catch the reflections. I did a quick LTspice simulation, using OP275 opamps for both opamps, with +/-17.5V supplies. I had to change the resistor values to get it to oscillate, changing R5 to 1K, R1+R2 to 22k, and R3+R4 to 50k. Changing the second opamp's + input voltage from 15V to 0V changed the output frequency from 2 kHz to over 10 kHz. Doubling R3+R4 cuts those frequencies in half. With the positive supply voltage on the + input, there was no oscillation. But use a capacitor in series with the output, because it goes to DC near the positive supply voltage in that case. You'll need to change the supply voltages and resistor values to suit your specific application, and might need to divide-down your sensor voltages depending on the power supply voltages that are available (so their maximum sum fits within the positive supply voltage).

Cheers,

Tom

 

Are you saying that you can't employ discrete transistors either?

 

As I see it, you are aiming a narrow beam IR LED at an angle to a reflective surface, say 45 deg. Close and clesest detector will have greatest output, move back abit & 2nd detector lights up; so you need a voting circuit to detect strongest signal. Use PNP transistors as switches, controlled by loudest detector to change timing resistors on a 555.Each detector has its own tone.

 

Amplitude discrimination is out as TSOP1138 has built in AGC.

 

@goote

Thank you very much, your second suggestion was a kind of information that I needed, I will try it in the next lab class.

thanks to everyone who was interested.