#robotics Logs

Jan 14 2020

01:13 AM mrdata_ is now known as mrdata
10:00 AM rue_mohr: armyofevilrobots, hows roads out there?
10:01 AM rue_mohr: your hill is way worse than mine
05:51 PM rue_shop2: ok, instead of being within 1% of the value
05:51 PM rue_shop2: I'll go for 1/128th
07:08 PM rue_shop2: hmmm
07:08 PM rue_shop2: I want to redo the analog front end more and more
07:08 PM rue_shop2: I also want to finish this filter profiler
09:17 PM rue_mohr: so better was to just be +- 80 units
09:18 PM rue_mohr: as the precision of the readings doesn't change with their size
09:29 PM rue_mohr: hmm
09:32 PM Tom_L: how are you collecting data for this?
10:11 PM rue_mohr: serial port
10:11 PM rue_mohr: 4.630V, 0.105V, 3.972V, 0.580V, 6101Hz, 6099Hz, 47�, 47�
10:11 PM rue_mohr: ^^ those are the lines it sends back
10:11 PM rue_mohr: you enter the freq you want it to test
10:19 PM rue_shop5: I need to work out why the phase angle is out
10:21 PM Tom_L: numbers not accurate enough?
10:22 PM rue_shop5: -106 degrees, but the max should be -90
10:23 PM rue_shop5: 4.986V, 0.028V, 4.956V, -0.057V, 4.9Hz, 4.9Hz, -2 deg, -2 deg
10:23 PM rue_shop5: 4.984V, 0.068V, 4.289V, 0.643V, 859.9Hz, 859.9Hz, -44 deg, -44 deg
10:23 PM rue_shop5: 4.941V, 0.171V, 2.595V, 2.330V, 20000.0Hz, 20000.0Hz, -106 deg, -106 deg
10:25 PM Tom_L: maybe at 20khz the samples are too slow
10:27 PM rue_mohr: no, the front end uses averagers for the peaks, and there is a "zero crossing" detector that gets the phase
10:27 PM rue_mohr: altho
10:27 PM Tom_L: where you gettin the extra 16 deg from then?
10:27 PM rue_mohr: if there is an error between the two channels, I'd not have noticed yet
10:28 PM Tom_L: propogation delay
10:28 PM rue_mohr: maybe I should pipe it straight thru and see what it gives me
10:28 PM rue_mohr: the "zero corssing" is trickey
10:28 PM rue_mohr: for each channel, it takes the high and low peak, and divides it by two for a zero point
10:29 PM rue_mohr: and then compares the signal against that point
10:29 PM rue_mohr: I had to do it without involving a filter
10:30 PM rue_mohr: come to think of it
10:30 PM rue_mohr: I know one of my peak detectors is drifting, and that'll cause error
10:31 PM rue_mohr: maybe I have to wrap the analog front end board in a self calibration layer