TDC1000-C2000EVM: Signal detected in air but not in water

Hello,

Could you attach those scope shots again? They did not upload properly. 

It would be helpful to see the TX signal, START, STOP, and COMPIN scope plots, for each test in question. 

It's difficult to say why you are seeing these readings in your testing without seeing the respective scope shots. 

Are you able to test the transducer in a simple level sensing setup with the transducer placed in the tank wall so that it is electrically isolated but physically contacting the water? This may be a good starting point. 

A small, closed 3mm transducer, driven by 3.6Vpp, may have difficulty transmitting and receiving through the silicone.

Regards,

Gabriel

Hi Gabriel,

I am using a 2-channel PicoScope so the scope shots will only indicate the START and COMPIN signals. The only settings changed in the GUI from default was TX_FREQ_DIV = "Divide by 8" , NUM_TX = "8 Pulses" and PGA_GAIN = "21dB".

For the shot in air, the graph in the GUI updates at a value of ~42us which I've also marked with a dotted line in the scope.

For the shots in water, the transducer placed on the tank wall (expected 33us) and in the middle of the tank (expected 15us) but both scopes showed the same reading.

If the 3.6Vpp signal is not enough to drive the transducer, we would expect the scope shot in air to be the same as the scope shots in water instead of obtaining a signal?

Thanks for the assistance!

I agree that the signal would be weaker in air, so perhaps the drive voltage is not the issue.

Further questions:

- What is the target material for the air test and the material of the tank for the water test?

- Are there any differences in the hardware/transducer setup between the air and water test other than one is placed in water? (silicone in both cases?)

- Is the transmit signal about the same in both cases?

- Have you gone through your experiment again with another transducer and replicated the previous results?

- What type/brand of silicone are you using?

Regards,

Gabriel

The material of the tank is plastic. For the air test, the transducer is just removed from the water in the tank and left hanging above it. However, even when the transducer was moved toward/away, the reading still shows ~42us.

No. The setup for both tests are identical. The transducers are sealed with silicone in both cases. Multiple transducers were also tested with the same result.

Yes. The transmit signal is the same when examined for the oscilloscope.

Yes. Another transducer was prepared and tested with the same results.

I am using Selleys Silicone sealant as the silicone is just for waterproofing.

Hello,

I am trying to replicate and debug the issue here with the OOB 1MHz transducer. I will provide an update tomorrow. 

Regards,

Gabriel

Hi,

I'm sorry for the delay. 

Using a 1MHz Steminc transducer I replicated your results for the test in air with and without silicone, and your test in water with silicone. Similarly, I could not get a clean signal from the COMPIN pin in these scenarios, but I saw normal operation and a clean signal after gluing the transducer to my acrylic water tank as done in snaa266a.

Transducer held in water:

Transducer glued to the outer tank wall: 

As seen in previous post 790615  we have encountered this issue before from customers, but cannot provide proven advice for solving the issue at this time.  I would recommend experimenting with different types transducers for water coupled applications and contacting the transducer manufacturer about the issue and further transducer related questions.

Regards,

Gabriel

Hi Gabriel,

Thank you for the response!

I have tried of several other silicone sealed transducers (2.2MHz, 1.2 MHz, 665kHz, 480kHz) as well as adjusting the settings in the TDC1000-C2000 GUI to try and obtain a signal without gluing it onto the container.

I was able to obtain correct oscilloscope images using the 480kHz piezo and with an update frequency of 100ms in water:

5cm

8cm

Could you provide some insight as to why this particular setup worked? I have listed some observations about this below

1) Trigger update frequency varied from 100ms-1.0sec. Only 100ms gave proper oscilloscope image.

2) 2.2 MHz, 1.2 MHz, 1 MHz, 665 kHz piezos were not affected by adjusting update frequency

3) TX_Freq_Div at 2/4/8 gave good signals (best at 8) but signal disappeared at 16/32. Theoretically, I would've expected best signal at 16 since freq = 8mhz/16 = 500khz to be closer to resonance

Thank you for these updates. We are starting to actively look into this issue, but it will require some experimentation and time to fully understand. As the Thanksgiving holiday is this week, I will have to get back to you later next week with an update. 

Regards,

Gabriel