Amplify signal level from microvolts to 5Volts

[johnxavier]

My project is to detect the object distace by using ultrosonic sensors. The signal from the sensor is in the order or microvolts. Can I use OP-AMP to amlifiy the signal to give to microcontrollers. Please suggest the part number of the OP-amp (which is easily available) I can try to amplify microvolts input to 5V output without distortion.
regards,
JOHN XAVIER

 

[Lewish]

Hi John, assuming that you have 5 microvolts and you want 5 volts, this is a million to one amplification. Far more than could be done with a single amplifier. Even if you could do it, it would be an uncontrollable oscillator. Even if you limit your gain to 1000:1 you will probably only end up with oscillators.
So, I recommend that you do 3 series amplifier stages, each with a gain of 100. I like either a TexInst OPA2137 or a Analog Devices OP275. You will need a supply voltage of at least +/- 9 volts. +/- 12 volts is better.
To connect, you put a 1Kohm resistor to the - input pin, and a 100Kohm resistor between the - input pin and the output pin. Then connect the + input pin to ground. Put your transducer output to the first 1Kohm resistor and the output of the first amp to the 1K resistor of the next stage. Put the output of the last stage into your microcontroller.
If you have too much voltage at the microcontroller, reduce the 100Kohm resistor in one stage only.
Good luck.

 

[xnuke]

Remember, when measuring distance, you are looking at an analog signal. Your desired output as you described would then be in the range of 0-5 Vdc, not just a 5 Vdc input.

With three inverting stages as described by Lewish, you're going to end up with an inverted signal. It might be nice to use an inverting buffer to keep a positive voltage input. Hook up the output of an op-amp to its own - input pin. Hook up the output of the third op-amp stage (described in the post above) to the + input pin. Then, hook up the output of the buffer to the input of the circuit you're using to interface to the microcontroller.

There are a couple of ways I can remember offhand to input voltage data to a microcontroller.

1. Use an ADC that is polled periodically by the microcontroller and puts the input on the data bus.

2. Use a voltage to frequency converter to input on an available interrupt line, then count the pulses for a given sample time. Control the sample time by controlling the enabling of the V/F converter with the microcontroller.

After the data is in the microcontroller, scale as necessary to correlate the the actual distance sensed.

 

[johnxavier]

dear lewish & xnuke
Thanks a lot for your ideas. As lewish said I was presiously tried with 1000 times amplifiaction with LM324 op-amp I ended up with oscillation. I will have to go for 100 amplification.
I have generated only +5V supply in present design. Can I use directly use +5V and other pin as GND. This will help in directly giving the signal to controller. What is the impact on +-9V. or +-12V power supply.
Actually the signal in micro volts is in frequncy range at 40Khz. I am using the above amplifier to increas the signal strength. After that the 40Khz. waveshaped(+5V) signal will be given to microcontroller.
As xnuke suggested I need a F to V converter to convert the 40Khz. in to DC level. If 40Khz. signal in micovolts level appears it should give a high. Can you suggest the F to V converter part number or any circuit idea for F to V convertors.

 

[elvik]

Hi Johnxavier,
It is a bit unclear, is the detector's uV is the main echo burst or consequent one. If I understand correct, you need ignore some reflections from surrounding obstacles, that may come earlier and be stronger, than the echo you need.
In other words, I am not sure that your project will be operable without some sort of Automatic Gain Control, providing the same output in wide range of input signals.

 

[johnxavier]

Hi elvik
The AGC is not required in my application because all the obstacles nearby has to be monitored. A 40Khz. UV signal is transmitted and the reflected signal from the object is received. By monitoring the time gap between the transmitter sinal and received one the distance is calculated.
Thanks for the reply
Rgds,
JOHN XAVIER

 

[Lewish]

Hi John, with the Opamp connect to +5V and GND, your useable signal voltage range is only about 1.2 volts to 3.8 volts. An Opamp can not swing its output to the rails. Thus if you use +/-5V as your supplies you can swing the output from -3.8 to +3.8 volts. Or, 0.0 to +3.8 volts if your signal doesn't go below ground. If you really want to stay with a single ended supply, then use at least a +9 volt and GND connection. Then, AC (capacitor) couple your signal and bias it up to about 4.5 volts above ground. This is done by connecting the +input pin of the Opamp to the junction between two resistors, the ends of which are connected to GND and +9 volts. 10Kohms is a good value to use for the resistors. I have done echo distance measurement using 7KHz as the signal from the transducer. It is possible to be quite accurate, but accuracy suffers with temperature changes.

 

[repoman]

If you are reading an analog signal and want a microcontroller to do ANYTHING with it, you need an ADC circuit, or a microcontroller with one built in, like Microchip's PIC 16C67x series. The output is in 8bit BCD. Micro controllers will properly not react to an analog signal.

 

[buzz41]

lewish -

An Opamp can not swing its output to the rails. Thus if you use +/-5V as your supplies you can swing the output from -3.8 to +3.8 volts.

A pleasant surprise for those of us who had not done much opamp shopping recently. There are now a variety of rail to rail opamps at little premium. Inputs and outputs are spec'd within 100 mV or less from rail. Some get a little limp as they approach even closer to the rail voltage, but especially for slower circuits, it may not matter. The rail to rail feature can make things simple.

 

[xray]

Hey John

Check this thread reference: thread181-27567 in forum: 222
or do a lookup on ultrasonic in a keyword wearch.

regards;

cy

 

[xray]

Hey John:

That post reference was incorrect the ultrasonic keyword look up and in their is a post for a simplified ultrasonic circuit, title. The device I was reffering to is a simple one bit adc. It effectively gives 5000v/v gain with one amp and a comparator. I can not boast about the effectiveness of the circuit it was a proffesor design. anyways if your interested email me for a schematic and equations.

regards

cy

 

[bondoman]

Couldn't he use the 1.2-3.8 volt output of the Opamp as the input to a transistor and let the transistor switch the levels from 0-5V.

jmtcw

 

[tiki]

Hi johnxavier, try this link:

http://www.ece.concordia.ca/~dave/project/elec.html,

a very "low level" ultrasonic transducer kit. Or: Maxim A0713.pdf (design showcase).
For accurate distance measurement you will need high input impedance, high bandwith, low noise amps for amplification (40kHz*100V/V = 4MHz, >10 times more may be needed for accuracy) and a linear phase bandpass for filtering the incoming burst. Bessel filters you will find at LT-application notes and datasheets. You may also use a JFET input stage and a following low input impedance OPA stage. Also the signal evaluation is relatively complex. This task will not be done in days!
tiki