Ultrasonic Transceiver Problem ( No Echo )

[rneraslan]

Hi, I am working on Ultrasonic Transceivers for my range finder project. My main goal is to achieve this project with one sensor.
Before this, I've achieved range finding with Receiver / Transmitter Sensors separately. It worked well but as I said I need to finish this with transceiver sensor.

So, here is the my schematics that I draw for 1 sensor.


My driver circuits works fine. It generates around 100Vp-p. But the problem is there was no any echo signal coming through amplifier circuits. I tried to bring another ultrasonic transmitter sensor that generates 40khz 20Vp-p closer to this single transducer circuits. Still, there was no any signal after R13 & D2 protection stage on my osciloscope.

Here is the another info. Amplifier circuits also works fine. How do I know? When I stop generating high voltage from Driver circuit and remove the R13 & D2, I got very clear echo signals from output of 1st amplifier.

Here is the scope view

Blue: My Burst trigger signal. From Signal Generator ( LABEL : PWM )
Yellow : Bursted Voltage x100 ( TP14 )
Purple : 1st Amplifier output ( TP3 )

What am I doing wrong? Why can't I see any echo? Why my protection circuits ( R13 & D2 ) blocks the echo signals?

Thanks

[tatus1969]

just an idea: the transducer is basically acting like a capacitor. If that ends up fully charge at the end of the burst, then it may be still busy with discharging itself through R13/D2, attenuating the received signal.

Also: you drive the transducer through a transformer. The transducer "sees" its leakage inductance in parallel to it. Unless you manage to manage to obtain resonance at 40kHz with transducer and that inductance, it is short circuiting it when receiving.

EDIT: not leakage inductance, but inductance of secondary winding, because we can assume that the driver transistor is switched off.

EDIT2: that is a lot of capacitance for IC3B, that can easily lead to stability problems and parasitic oscillation. I would at least add 100R to its output. But there is also no need to make the nodes there such low impedance.

[rneraslan]

So you are suggesting that I should correct the resonance with transducer?
Btw, what is for R15?

And I've just noticed that my reference voltage is GND. Should I give 2.5 offset voltage for secondary winding?
Is this scheme OK ?

Right now, I dont know what to do..

[tatus1969]

So you are suggesting that I should correct the resonance with transducer?
Btw, what is for R15?

I'm just saying that your transformer short-circuits the transducer in receive mode. Maybe put a resistor (500 ohms?) in series with the secondary, remove R15.

Another thing: what is the distance range that you want to measure? From your oscilloscope diagram, it takes more than a millisecond for the transmitting signal to decay. That is enough for the acoustic signal to travel 0.35m. You cannot easily measure distances less than half of this because you'll have transmitting and receiving signals overlapping. What is your "big picture" concept of how to be able to work with one transducer? How do you want to handle the resonant decay after transmission?

And I've just noticed that my reference voltage is GND. Should I give 2.5 offset voltage for secondary winding?
Is this scheme OK ?

Why not? Your receiver is AC coupled.

[rneraslan]

Actually, I have a tolerance for distance due to transmission overlapping. And It's around 40cm which is enough for this resonance decaying. I am interesting measuring long distances like the 3m ground to 40cm.

Why not? Your receiver is AC coupled.

Opamp is feeding with Single Supply, isn't that a problem for ac coupling?

[rneraslan]

And this is the circuit from referance manuel from Transducer.

[tatus1969]

Opamp is feeding with Single Supply, isn't that a problem for ac coupling?

Left of C3, everything is referenced to GND. Right of C3, everything is referenced to VCC/2. Should be fine like that.

[tatus1969]

And this is the circuit from referance manuel from Transducer.

As you can see, they use a tuneable transformer there, probably to be able to tune the secondary into resonance at 40kHz with the transducer. That is not visible from your schematic. What transformer do you use? Did you tune the circuit?

[Audioguru]

The minimum supply voltage for the AD712 opamp is 9V but yours is only 5V.

[rneraslan]

The minimum supply voltage for the AD712 opamp is 9V but yours is only 5V.

I am sorry for my wrong schematics, forgot the correct it. I use MCP6024 instead of AD712

And this is the circuit from referance manuel from Transducer.

As you can see, they use a tuneable transformer there, probably to be able to tune the secondary into resonance at 40kHz with the transducer. That is not visible from your schematic. What transformer do you use? Did you tune the circuit?

I have an Ultrasonic Range Finder kit that sells on ebay etc. I did tear down It's transducer and transformer and applied my schematics. I supposed that the transformer is tuned for this transducer

[tatus1969]

I supposed that the transformer is tuned for this transducer

Have you verified it with a scope??

I have worked with these in the past, they are fully sealed and not very sensitive. Make sure not to touch the active surface with anything, that brings them out of resonance.

[rneraslan]

I supposed that the transformer is tuned for this transducer

Have you verified it with a scope??

I have worked with these in the past, they are fully sealed and not very sensitive. Make sure not to touch the active surface with anything, that brings them out of resonance.

1- I am not quite sure how to be sure that it*'s tuned? I uploaded scope image which Yellow one was transducer Driving signal. What can you say about that?
2- What do you mean that they are fully sealed
3- I do touch it's surface sometimes by accident but nothings change.

[tatus1969]

1- I am not quite sure how to be sure that it*'s tuned? I uploaded scope image which Yellow one was transducer Driving signal. What can you say about that?

Your transmitting output voltage did not look like it would be resonant, it looked rectangular. Try to achieve resonance, you should expect something sinusodial then. I'm not sure if the driver transistor should stay in ON state while receiving, as this enables utilizing the transformer's leakage inductance to resonate with the transducer. And that is what you can vary with moving the insert. Maybe have a look how the signals look like in the original board.

2- What do you mean that they are fully sealed

Is it like this one?

[rneraslan]

I have an update.. I changed circuit like this


But the output of the 2nd opamp didn't satisfy me.


First gain is 44, second is 100. But the amplitude of amplified echo is so small that I could barely process that signal.And also there was a lot of noise. didn't expect that.

And yes I use something like this transducer

[tatus1969]

As mentioned, this type of transducer has reduced sensitivity compared to the "open" ones like these

http://www.theengineeringprojects.com/wp-content/uploads/2015/05/sensor.jpg

Also, these are resonant types (they actually have two poles, check the datasheet). Make sure you hit its working frequency precisely.

If you cannot further reduce noise floor (better opamp?) you need more sensitivity or more output power.

[Niklas]

Have you tried to move the anti-parallel clamping diodes on the opamp input to the left of the input coupling capacitor? As the are located now, they will pull the input voltage close to GND. The transducer is referenced to GND, but that corresponds to V- on the opamp.
Also the opamp can be on the limit in terms of gain margin. 10 MHz GPBW with 100x gain gives you only 2.5 times margin at 40 kHz. Maybe you should try something faster, for instance 60-100 MHz?

[rneraslan]

Have you tried to move the anti-parallel clamping diodes on the opamp input to the left of the input coupling capacitor? As the are located now, they will pull the input voltage close to GND. The transducer is referenced to GND, but that corresponds to V- on the opamp.
Also the opamp can be on the limit in terms of gain margin. 10 MHz GPBW with 100x gain gives you only 2.5 times margin at 40 kHz. Maybe you should try something faster, for instance 60-100 MHz?

Should I add 2.5V offset to high side of secondary transformer windings?  Like this


I couldn't understand that why should I use faster op-amp. Isn't that enough for amplifying 40 khz at gain 100? MCP6022 GPBW looks like handle as I understand, please correct me if I am wrong.

[BrianHG]

Since you appear to be force-feeding the PWM, if U$2 is not properly tuned, the device will still transmit, but, the reception will be really weak.

Can you double the length of your transmit pulse then provide a close-up of the transmit waveform, then the receive waveform, with vertical timing cursors on you scope for each?

[Niklas]

No, just move the clamping diodes to the point between C3 and R13. The opamp + input is kept at 2.5V and seen as a virtual GND. C13 will remove the DC offset and place the AC part of your signal around the virtual GND.

I had a similar issue in a similar application, but with 1 MHz excitation and resonance frequency. There the acoustic coupling was much better in a liquid so the bandpass filter gain was approx 5x. Even with a 20 MHz GPBW the filter type needed a faster opamp to not act as a low pass filter. Just by swapping the opamp, the transfer function started to match the calculations. TI has a technical paper:
www.ti.com/lit/an/sloa035d/sloa035d.pdf
where they recommended 40dB of gain margin at the corner frequency, in section 3.3, in your case around 40 kHz. Adding the dBs, you will then need >=80dB at 40 kHz.