Hello,This is a long shot, but I wonder if anybody can help me here. Is it possible to convert the amplitude of a sinusoidal burst signal into a stable 4-20mA response? If yes what device would allow so?When controlling high frequency ultrasonic transducers (e.g. for NDT) the response is made of multiple sinusoidal reflections.Would it be possible to convert the amplitude (or the RMS of the amplitude) of high frequency (e.g. >1MHz, <10MHz) sinusoidal burst reflections into 4-20mA standard via an analog RF device? Further, would it be possible to do that in case of multiple reflections? Does the nature of this type of signals introduce any complication in such conversion?Best Regards
Generic ways to deal with this:
1) Implement analog peak detector and scale it to a proper voltage range to be used as an input for the 4-20mA loop driver such as AD5750. For other parts, please have a look here: https://www.analog.com/en/products/amplifiers/specialty-amplifiers/industrial-4-20ma-driver.html
2) depending on the number of samples required per period, an ADC with high sampling rate might be required. Once sampled, the RMS or the peak values of the converted input signal can be derived and passed to a microcontroller. This information may then be forwarded to a 4-20mA industrial DAC such as AD5420 or AD5410.
Do you have an idea as to how fast does the amplitude of the signal changes? and how fast do you want to see the change in the 4-20 loop?
thank you for your reply and the suggestion. To answer your question, I am interested in measure at the rate of 1Hz and that is the rate at which the amplitude changes and approximately the conversion time that I would like to guarantee also in the 4-20mA side.
isantos said:depending on the number of samples required per period, an ADC with high sampling rate might be required
What do you think is a treshold for the number of samples? When processing signals at high frequency the digitisation rate should be quite high, but I can keep this down by reducing the number of cycles of the input burst.
If I use the AD5750, can I use any classical data logger to test the functionality of this solution?
Further, I would like to ask the following just to make sure that I can use the output from the suggested device. Will the current output be constant?
If the amplitude rate of change is 1 Hz, AD5750 should be able to respond very well.
You can start with sampling frequency as 10x the input frequency, more samples will give you better precision as you will be able to detect the "peak" of the input signal even with few periods of the input signal.
If the input frequency and waveform is constant and defined, let's say a sine wave, you may opt for a lower sampling frequency then use the sampled data to get the approximate "peak" value of the input signal.
You may use the usual data logger to measure the 4-20mA output from AD5750, or you may use a load resistor to convert it to voltage as long as it is within the output loop compliance voltage range.
Output current is constant as long as VIN is constant.