zygot Posted February 12, 2018 Share Posted February 12, 2018 21 hours ago, xc6lx45 said: Here is some experimental data. It's completely taken with "on-board tools" using the XADC, no external filtering. Anybody can repeat the experiment with a patch cable. So the analysis accompanying the quote above implies one thing and this quote: 21 hours ago, xc6lx45 said: First, any DAC requires a reconstruction filter (lowpass) implies something different. I'm just trying to get you to discuss what I think that you should have mentioned from the start. So connecting your "DAC" output to a low bandwidth high input impedance ADC appears to perform one way. Connecting the same pin to any random load might appear to perform quite differently. No doubt you are aware of the discussion on using the PMODAMP2 using a similar approach on a different thread. In that case the relevant reconstruction filter and low impedance driver are included in the design. I just think that if you are presenting something for an undefined purpose you should discuss the relevant issues. I'm not one to throw a pocket knife into a crib and say "have fun kid.. and don't hurt yourself" Of course had you not added the "analog" function generator feature to your tool there would not have been much discussion. Anyone who has had to select a commercial DAC or ADC for a particular purpose has been introduced to the kinds of games that vendors play to sell product that might not be appropriate for that purpose. BTW, my idea of a project for this particular forum is one that the user can modify and learn something from... but the rules aren't clearly stated and there isn't a place for something like this.. Link to comment Share on other sites More sharing options...
xc6lx45 Posted February 12, 2018 Author Share Posted February 12, 2018 === 0v4 released === Voltmeter: Option to extend the ADC input voltage range (and raise the probe impedance) by using series resistors Voltmeter scaling slightly tweaked (fullscale changed from 3.3 V to 3.32 V calculated from the schematic) Added Octave function to load binary ADC data to the installer On my board, there is a ~50 mV offset to both ADCs. in other words, leaving the input open shows 50 mV. I may add some calibration in the future to get rid of it. EDIT: Attachment deleted because of quota. Please pick a later version, there are no known downsides Link to comment Share on other sites More sharing options...
xc6lx45 Posted February 14, 2018 Author Share Posted February 14, 2018 === LabToy 0v5 === Voltmeter rewrite, additional readouts Main page buttons raise sub-windows Actually quite a bit has happened under the hood. The internal architecture now supports feeding any number of sinks from a FIFO data stream (the ADC connects to both the voltmeter and the file writer). On the surface, only the voltmeter changes. Please note the tooltips! It now implements true moving-average statistics over a 200 ms window. This particular length suppresses both 50 Hz and 60 Hz power line hum, and responds still near-instantaneously. The screen refresh rate is actually higher than the window length, the two are independent. New readouts for: VMax: Highest sampled voltage in the last 200 ms window time interval VMin: Lowest sampled voltage in the last 200 ms window time interval VRMS: "True RMS DC average" over the window (put this number into Ohm's law and it will give the correct power dissipation over some resistor) VRMS, AC: "True RMS AC average" over the window (put this number into Ohm's law and it will give the correct power dissipation over some resistor that is AC-coupled e.g. using a sufficiently large capacitor or transformer) The latter two should give similar readings to an "expensive" (true RMS) multimeter. Speaking of "expensive", on my board the readings are about 10 % off... Until now, the results are calculated rigidly based on component values taken from the schematic, with no regard to tolerances. Maybe it's time for some calibration algorithm, stay tuned... The following are drafts for a future online help for the voltmeter. It's first-semester stuff. Hope I got my equations right Well I guess it's one of those exams few people want to face a 2nd time... PS: For best results with the voltmeter, set a high ADC sample rate e.g. 720000 in the "capture" window. CPU load goes up, but so does bandwidth. Link to comment Share on other sites More sharing options...
xc6lx45 Posted February 14, 2018 Author Share Posted February 14, 2018 Hmmm looks like I've exhausted my upload quota for today... 0v5 link on google drive: https://drive.google.com/file/d/1PfloGnO0D80ZMiJ5eGi0AKRI1IgXWe9s/view?usp=sharing Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.