Possibly, but not necessarily. The ADCs can be set to sequential or round-robin mode, and there is no information as to which has been selected. If they are operating in sequential mode then there should be a similar pattern from the 4 readings from each of the MCU showing, in effect, the waveform following the trigger. The period of this waveform for each MCU is Michael’s calculation of 0.002ms. If they are operating in round-robin mode then they should each show the same portion of the waveform, but over a period 4 times as long.
Possibly, but not necessarily. If the ADC are in free running mode then they will quickly get out of sync and the readings could be starting at any period within the 96-cycle range of the 48MHz clock for each MCU. But if they are running in one-shot mode then the cycle will start simultaneously for each MCU and any variability in the 48MHz rate will likely be unnoticeable. It depends on how the ADC are configured and that is not revealed in the code provided. The comment does say that the ADC is started by the trigger, and that might be one of the “'OPTION” settings but I don’t know enough about PicoMite to determine that. I am, however, highly suspicious of “'OPTION POWER PWM” because all the advice about using ADC on the RP2040 recommends turning PWM power off before using the ADC unless you are providing your own VRef (which OP isn’t).
Hi Jeff
Thanks for that. I had a look at the relevant bits of the data sheet but not too deeply. I was wondering what accuracy Hasham was looking for. The above quoted 0.804V to 0.811V only represents an error of 0.3% to slightly below 1%. Not too bad and could be a combination of instrument tolerance and measurement techniques.
Have you looked at his PDF linked above. It looks like the main variation is between MCUs NOT between ADC channels in any one MCU.
Cheers Bob
Hi Hasham
I just had a look at your linked PDF. Interesting.
Prior to that I had formed some questions which I will ask anyway.
Is the signal input and trigger input the same sine wave signal.
What instruments are you using for measurements and how are they connected.
What is the calibration status of instruments.
What is the source impedance of your test signal. Refer note below.
The last 3 questions will only affect the absolute measured value. They should not affect comparison measurements as long as all methods are the same.
I ask this as most instruments will have a tolerance usually expressed as a +/- percentage plus a number of Least Significant Digits. Example +/- 0.5% Plus 10 LSD. Your quoted absolute figure ov 0.804V and 0.811V would represent an absolute value error of 0.3% and a little under 1.0% compared to 1.2V which in the grand scheme is not too bad.
So just what sort of accuracy are you striving for.
Regarding your PDF.
This indicates the main variation (DC and Sine wave) to occur between MCUs. Variation within a single MCU I think would be quite acceptable. I don’t think there is anything you can do about this except to select matching MCUs. This could get a bit expensive.
Variation with frequency. This I cannot explain. This seems to be very similar within each MCU and approximately follows the DC pattern between MCUs.
Important. You are rectifying this AC sine wave and present a series of half (positive going) sine waves to the ADC and trigger are you not. As Jeff said. “If you are applying an AC signal to the ADC all bets are off”. You CANNOT apply a negative voltage.
Note: The source impedance of your test signal can be important when applying to high impedance inputs. If too high the connection of a test instrument (normally 10MΩ) will make a difference to the signal level. It is for this reason that an oscilloscope should be in the X10 mode unless your source is pretty low (50Ω).
Cheers Bob
PS. Could you post a proper schematic (not like the previous one) of exactly what you are doing about signal rectification. including component types. Your choice of diode here could have an influence.
Yes, both the trigger and input signals are sine waves, generated from the same frequency generator. I am currently using a LabView-based code for data collection. Could you kindly guide me regarding the source impedance (internal resistance) of the test signal? I aim for the signals measured by the 16 ADCs (across 4 PICO units, each with 4 ADC channels) to be below 0.8 V, as the trigger is set to the falling edge. At a lower frequency of 30 Hz, I observed voltages of 0.804 V and 0.811 V. However, as I increase the frequency, the voltage continues to rise, as detailed in the previously shared PDF file. My concern lies in understanding why this increase occurs, particularly because I plan to utilize this setup later for data acquisition from equipment where precision is paramount, and any deviation could compromise the results.
That has answered one question. What about the rest.
You have not even said what this signal generator is. The source resistance should be quoted in the user manual or specs. This can be measured and/or calculated but if you don’t know how to do that I cannot (and will not) conduct a course in electronics and measurement techniques in 5 minutes. If it is selectable on the instrument (whatever that is) you can set it for 50Ω or as low as possible and ignore it.
I am thinking that possible the sample is not taken at precisely trigger time. There will be delays in the Pico and each sample will take finite time. As you are sampling a sine wave this will be changing while the sample is being processed. As you say DC is OK and repeatable but that is not changing during sampling. That could be the reason the output changes more with frequency increase, the level will change more during the sampling time which does not change.
You have also made no attempt to show just how you rectify this sine wave and present the signal to the ADC inputs.
Like I have said before getting information is sometimes like pulling teeth. I am afraid little information will probably be reflected in little assistance. It will be for be anyway.
It looks like you are trying for Laboratory results with Hobby equipment. I don’t think that can happen. The Arduino, RPi etc range I believe is targeted for the hobbyist and experimenter or home type use. I think the absolute accuracy is questionable although for its design use is pretty good.
I think for laboratory results you will need laboratory equipment,
That PDF is not a schematic - it’s just another photo, and it doesn’t actually show how things are connected. It’s possible to make some guesses from the photo, but guessing is not very helpful for solving the problem. A schematic would show every connection, including the frequency generator and the CRO. It would also show what the comparator and trimpot are there for, as these have not previously been mentioned.
One photo that would possibly be interesting is the CRO setup, including the display showing the trigger occurring at 0.8V on a falling voltage.
Also, you still have not indicated how you are applying a sine wave output to an ADC that can only cope with positive voltages.
