first weeks commit

README.md

+# Inductive Encoder
+
+## HTMSTMAA Week 7 - 4/15/21
+
+This week I switched gears from working on the dynamometer and set out to start communicating with the [LDC1101](https://www.ti.com/product/LDC1101) from TI. The LDC1101 is an inductive to digital converter that uses an external [LC tank circuit](https://en.wikipedia.org/wiki/LC_circuit) to measure changes in impedance (Rp measurement mode) or resonant frequency (LHC mode) of the oscillator, relative to the change in presence of conductive or magnetically permeable materials respectively. Despite the existence of plenty of inductive to digital converters on the market (many of which are 2 or 4 channel devices in a single package), the LDC1101 is the only one I could find that uses SPI, and has a sample rate fast enough for high speed rotary encoding. These encoders are a bit of a bear to configure, and require multiple spreadsheets and a pretty thorough read through the datasheet. With that said, there aren't many other sensing options that provide 24 bits with noise this low, and with SPS + communication this fast. 
+
+Over IAP I put together a closed loop stepper driver using the TMAG5170 (a 3-axis hall effect sensor) in place of the popular AMS hall effect sensors. Ultimately that sensor's SNR was just high enough to prevent resolving more than two positions between 1.9 deg full steps, while also being speed limited in terms of update rate. While this sensor will ultimately live on the back of that board, for a first pass on starting communication, I designed and machined a simple breakout board in Altium, and milled it in house.
+
+![breakoutLDC1101.png](./images/breakoutLDC1101.png)
+
+Because I have been programming so much on the SAMD51 lately, I figured I would go ahead and do a first pass communicating with the board on that micro. The LDC1101 requires a high frequency (in the MHz with a recommended value of 16MHz) external clock signal for LHC measurement, to serve as a reference when counting oscillations of the tank circuit. I reused some of the dyno code and adjusted it to run at 50% duty cycle at 15MHz as shown below, keep in mind TC4 is a 120MHz clock. This did require some digging through the datasheet, and I ultimately discovered the period register which was relevant after switching from NPWM WAVEGEN mode to NFRQ. With a working 15MHz CLKIN signal I started probing the LHC data registers which were all returning zero.
+
+```c
+TC4->COUNT8.CTRLA.reg = TC_CTRLA_MODE_COUNT8 | TC_CTRLA_PRESCALER_DIV1 | TC_CTRLA_PRESCSYNC_GCLK;
+TC4->COUNT8.WAVE.reg = TC_WAVE_WAVEGEN_NFRQ;
+TC4->COUNT8.PER.bit.PER = 3;
+```
+
+After more snooping around with a scope, it appears that my clock is turning off after I open up USB communication, which is bizarre considering I wasn't seeing this happening with the dyno. That's as far as I got this week, but I do plan to double back and investigate if that timer is also turning during serial communication for the dyno (this would be bad considering it determines back torque on the motor, which I am very reliant on being steady during sweeps). For this project, considering I am not relying on any external libraries, I will switch over to an STM32 micro (also what I've used for the closed loop stepper driver that this will ultimately be implemented on).
+
+![firstTest.png](./images/firstTest.png)
+