From eee9156186830a429ea5cef3552f15d181b398e4 Mon Sep 17 00:00:00 2001 From: David Preiss <davepreiss@gmail.com> Date: Thu, 6 May 2021 17:32:37 +0000 Subject: [PATCH] Update README.md --- README.md | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 9dd407a..9dcb9e3 100644 --- a/README.md +++ b/README.md @@ -20,15 +20,21 @@ Before mounting and testing the encoder I wanted to test some parameters of the  -It's worth noting that inductance numbers were definitely asymptoting off around 200kHz (at 150kHz we had dropped to only 1.133 from 1.044 -Two explanations I thought of could be: -Because the coils are unshielded (unlike the SM component) at high frequencies we are casting flux further and therefore being more effected by pieces of metal on the order of 10's of mms away -Because the top and bottom coils are only 0.8mm apart and highly parallel, at higher frequencies we are getting more of an effective parallel plate capacitor effect +It's worth noting that inductance numbers for my coils were definitely asymptoting off around 200kHz. For example here are three inductances measured at three different frequencies, both with min target interaction: +* At 50kHz - 17.768 uH +* At 150kHz - 1.133 uH +* At 200kHz - 1.044 uH + +Two explanations I thought of for frequency dependant inductance could be: +1. Because the coils are unshielded (unlike the SM component) at high frequencies we are casting flux further and therefore being more effected by pieces of metal on the order of 10's of mms away. *Later I seem to have disproved this by introducing larger pieces of metal around the coils at and not seeing appreciable changes in inductance between frequencies. +2. Because the top and bottom coils are only 0.8mm apart and highly parallel, at higher frequencies we are getting more of an effective parallel plate capacitor effect (although this also doesn't seem correct as you think the capacatance effects would start to dominate at higher frequencies and cause more delta and less asymptoting). I also experimented with just measuring Rp directly, which the impedance analyzer I had accesss to allowed. The problem here was that I am limited to 200kHz peak frequency by the analyzer, and so I would have to interpolate way into higher frequency space with what's behaving (at least < 200kHz) as a second order polynomial, and I think that skin effect starts going off and doing nonlinear stuff at higher frequency, so I don't think the interpolation shown below is really safe to use.  +And lastly below is a picture of the encoder setup for the most part (I was soldering on different caps so wound up removing one of the screws that's in the way). It appears I have 26105 ticks of resolution, but am still sorting through parameter tuning necessary to bump that off and starting to trade some it in for SPS. +  ## HTMSTMAA Week 8 - 4/22/21 -- GitLab