I would like to replicate the experiment described in Sensors & Transducers, Vol. 240, Issue 1, January 2020, pp. 11-18, but I am unsure about where to source or how to build the “ripple free precision rectifier” PCBs the authors used in the paper. They did not mention the model, manufacturer, nor the performance specifications of the PCBs, and because I’m much more of a chemistry expert than an electronics expert, I’m unsure about how to proceed. May I ask for your help with either identifying the PCBs they mention or with specifying the proper components to build them? Any guidance would be appreciated. Thank you!
Welcome to the forum.
From page 14 of the linked article:
The electronic part of the proposed measuring system as shown in Fig. 1 is constructed on a vero-board
That is the trade name for Vero Technologies Ltd’s line of prototyping stripboard.
I don’t see any further information in the article indicating which model they used.
Here’s Vero’s prodcut list:
Generically these types of prototyping PCB’s are:
and:
AFAICT, DigiKey does not carry Vero’s products but they carry similar products from many manufacturers with more choices for the copper pattern and often better prices.
Here’s some articles about using stripboard/perfboard:
2 Likes
Hello Paul,
I wasn’t referring to the specifications of the perf board that the circuit was soldered to. I would like to understand how they built a “ripple free” filtered rectifier. What type of diodes should be used? How much capacitance will be needed to make a “ripple free” output? How small do the peak-peak heights need to be to be “ripple free”? What type of capacitors should be used?
Hi MAP-6022,
Well, of course there is no such thing as a completely ripple-free precision rectifier, so one must decide how much filtering is good enough. Larger capacitance values will reduce the ripple, so one can get to “good enough” with sufficiently large values, though if the capacitance is too high, there may be insufficient frequency response to maintain stability.
A “precision rectifier” implies that one rectifies an AC voltage without suffering the voltage drop incurred by just using rectifiers (typically about 0.7V for a standard rectifier). To accomplish this, one would typically use op-amps to correct for this voltage drop across the rectifier.
Take a look at this video I found from Professor Fiore which explains one method of creating a precision rectifier with a couple of op-amps. Note that the circuits shown are not the only way to accomplish it, but he does a pretty good job of giving you the idea. He covers a bit about capacitive filtering in the middle portion, but the same would apply at the end of the video.
To account for whatever ripple remains after filtering, I would recommend you over-sample and average your readings. Taking 8 or 16 readings and averaging them should give pretty good results.
Here’s an additional tutorial from Analog Devices, MT-211, which covers a similar circuit.
Thank you! I’ve been researching unity-gain difference amplifiers and absolute value circuits formed from various op-amps, so I’m happy to see that your advice is pointing me the same way. The MT-211 tutorial was also really helpful.
However, one thing that I’m concerned about is the low, 50 mV excitation signal they’re using in the paper. Do these op-amp circuits work well with low voltage, low-frequency (1 kHz) signals? For example, I know that I should avoid traditional, filtered full bridge rectifiers because their diodes often require at least ~0.7 V to conduct in the forward direction. Are there certain op-amps that you would recommend for this application? I was thinking about using the AD8277 based on what this article was saying.
So, it does not appear that using the low 50mV signal would be a problem for the op-amps referenced above. However, you would likely get much more accurate measurements if you amplify the signal somewhere along the signal chain because an ADC, especially if one used a relatively low resolution one found in most microcontrollers, will do better measuring a volt or two rather than low millivolts.
You could amplify the signal prior to the rectifier circuit or after it. My gut tells me that doing this up front seems to make the most sense.
Nice find on the AD8277 circuit. I assume you meant to reference this article describing a precision rectifier circuit using that part. I was unaware of that note. I had never seen a precision rectifier circuit which didn’t require adding diodes to the circuit. The integrated resistors are also a big advantage, as they will be very tightly matched, which significantly improves circuit precision.
Excellent. Thank you for the recommendation to amplify the signal. What do you feel would be the most straightforward way of doing this? I’m thinking about taking advantage of the +5 VDC output of the Arduino and incorporating another two amplification circuits.
In the article you correctly identified (sorry about the copy/paste mistake!), it mentions in the “Improved Absolute Value Circuit” that “if the signal to be rectified is very small, a pull-down resistor at each op amp output can improve the circuit performance at 0 V.” Would it still make sense to make this improvement given the limitations of the built-in ADCs of the Arduino?
I greatly appreciate the guidance you’re providing me. Thank you.
Hi @MAP-6022,
Naturally the best way would be to sample the reference signal as well as the output of the opamp with the variable capacitance feedback, directly with two fast ADCs and do the rectification in digital domain (that is, a simple cycle-to-cycle peak detection). That way you could omit the analog differential amp and rectifiers completely and get very accurate results. In fact, the reference signal could also be generated in digital domain simplifying the circuit further.
Or another way, have a capacitance to frequency converter and measure the frequency deviation due to capacitance change. Then you do not need to worry about rectification at all, and neither need any ADCs.
Wondering, do Glucose molecules have electric polarity? Then the frequency of the AC signal may affect to the result.
Cheers, heke
1 Like