0
\$\begingroup\$

I have an ultrasonic sensor, operating in continuous mode, that is providing a 0-10 V DC output signal. It is connected to an ADC. We are using a cheap power suppy to power the sensor. The ADC has its own wall plug supply. The node where I connect the the ADC input to the sensor output reads the DC voltage just fine, but it also has an unwanted frequency component which causes my ADC readings to fluctuate. I will be isolating components, and using coupling caps and/or RC filters to attempt to isolate the noise, but I would like to know if anyone can identify this type of noise immediately so I can get rid of it. I am suspecting one of our power supplies/bad ground right now. It appears that this 100 Hz noise is the lowest frequency. Any ideas what this noise is, or other ideas about how I can stop it? Thank you in advance!

noise on the scope

\$\endgroup\$

2 Answers 2

0
\$\begingroup\$

Continuous mode is probably just automatically repeating pulse mode. It is much easier to design ultrasonic sensor which alternates between transmission and reception and measures time delay before hearing its own pulse, than any continuously transmitting sensor. And frequent enough pulses will work for most applications.

The 100 Hz frequency could be simply the repeating frequency of ultrasonic pulses sent (it gives 3.3 m roundtrip distance, maybe enough for unwanted reflections to be attenuated compared to the specified 25 cm max sensing distance). It is possible that power rail voltage drops or peaks on transmitter start/stop because of extra load and internal regulator for analog output is unable to filter it well enough. You can try to add more capacitance to the power rail close to the sensor. If your sensor and ADC has common ground and single-ended sensing, it can be ground current which makes this peak too. Make sure, there is no ground current flowing between ADC and sensor when it switches on transmitter. Or use a differential measurement.

100 Hz could be some coupling of mains too, but the frequency is bit off (mains frequency is usually pretty exact). But you can easily check if your pulses are in sync with the mains.

\$\endgroup\$
2
  • \$\begingroup\$ Europe can be off by as much as 150mHz before extreme countermeasures are taken. Twice as much as the measurement shows. At less than 10mHz off nothing is done. \$\endgroup\$
    – PlasmaHH
    Feb 13, 2018 at 8:25
  • \$\begingroup\$ You were right. That was the sensor cycling through a measurement cycle that was appearing at the output. I found the best practice was to sense, then turn off the sensor before getting the ADC to read because the sensor holds its output steady when not measuring. With supply coupling, and a RC filter from the sensor output to the ADC, I am getting steady ADC level readings at 16 bit resolution. I appreciate the help and pointers on what the noise signal could be. \$\endgroup\$
    – Fillups
    Feb 14, 2018 at 2:48
1
\$\begingroup\$

It appears that this 100 Hz noise is the lowest frequency.

100 Hz will be the ripple frequency from the output of a bridge rectifier used in a power supply that takes 50 Hz AC as its input source. Below is a picture of 120 Hz ripple when the input supply is 60 Hz: -

enter image description here

Here's what half wave and full-wave (bridge) signals look like when aligned to the input AC waveform: -

enter image description here

Half wave tends to produce a ripple at the same frequency as the incoming AC but full-wave (bridge) produces double the frequency because both halves of the AC waveform are combined in the rectification process to make DC.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.