Shared Projects by dewhisna

The Optical Interrupter Shield V2 board, for ProMini PiDuino HAT, was designed to interface four off-the-shelf optical reflective sensors, such as my QRE1113 Sensor Board, used in raw-sensor mode, to a Raspberry Pi via an Arduino ProMini PiDuino HAT. This board expands on the V1 design with a better on-board debouncing circuit and four sensor inputs instead of only two. It also adds level-shifters to work with either 5V or 3.3V, as driven by the I/O Ref voltage pin.

ERRATA: To work better with the QRE1113 sensor, as this board was originally designed to do, change R3, R9, R15, and R21 from 10K to 47K. This change is necessary to boost the gain of the QRE1113 sensor by about five times, which is needed to get the sensor out of the linear region and into the switching region of operation. Without that change, the sensor will lack enough sensitivity to function adequately for use in the quadrature encoder application it was intended (and probably most any other application too).

PID Loop version of the AN3813K Brushless DC Motor Driver Arduino Shield for use with VCR Cylinder Motors. PID Loop Control done via on-board ATmega328 coprocessor. NOTE: This is a ATmega328PB. NOT a ATmega328P! The ‘B’ version micro is needed for the dual I2C interfaces as one is a master and one is a slave. The master communicates with the on-board DACs used to drive the AN3813K chip and the slave communicates with the main processor via the Arduino header pins.

See the AN3813K BLDC VCR Cylinder Motor Driver Arduino Shield V2 project for the PLL Version of this board. Both versions are viable solutions depending on preference of Analog PLL circuit or Digital PID via a microcontroller.

All smaller surface mount discrete parts are 1206 footprint except for the two 24pF capacitors for the crystal oscillator circuit for the micro (C307 and C308). Those two parts use 0805 footprints.

Note: Populate either C1-C3 with R1-R3 or populate C12-C14 without R1-R3. C1-C3 is a low-side filter and C12-C14 is a high-side filter. Only one set is needed. Preference is generally for C12-C14. In either case, those capacitors should be bipolar.

ERRATA: Atmel/Microchip recently decided to change the ATmega328PB (and its variants) to no longer support full-swing oscillator drive and therefore won’t drive the standard AT-cut crystal that this board was designed to use. In fact, I even discovered a bug in their die that if you use a crystal that almost starts but doesn’t quite, it can get into a mode where it switches back and forth between the internal oscillator and the external crystal and if you try to do flash reprogramming with it running in that mode, it will drive the charge pumps too long and fry the flash cells, destroying the chip. I accidentally fried one of my ATmega328PB chips this way.

Therefore, do NOT populate the Y301 crystal, the C307 or C308 capacitors, or the R312 resistor. And instead, set the low-fuse byte to 0xE2 to enable the internal 8 MHz RC oscillator at full speed. There is no good way, as it is currently designed, to attach a low-powered ceramic oscillator or even an external oscillator module to this board.

Perhaps a future version of this board will be spun to allow better oscillator options and higher speed operation. But, in all honesty, 8MHz is plenty fast, as the original design of this circuit, some eighteen years ago, was with a 68HC11 running at 4MHz, which since it was CISC instead of RISC, was less than an eighth of the speed of this ATmega328PB at 8MHz, and it was plenty fast.

Breakout board for the common and widely available STM32F103C8T6 “sticks”. Provides easy connection for both I2C channels at both +5V and +3.3V with level converters. Has dual serial FTDI ports with level shifters, MicroSD Card slot on SPI #2, Waveshare compatible SN65HVD230 CAN connector port, and breakout connectors for Analog, PWM, and SPI #1.

Assembly Note: Be sure to trim the bottoms of the ST-Link connector pins on the STM32F103C8T6 board before soldering it to this breakout board. If you don’t, those pins will likely collide with C1 and perhaps R20 and R24. Trimming them, however, gives plenty of clearance.

DO NOT BUILD This circuit has serious design issues rendering it virtually useless. The biggest problem is that the sensor input circuit’s filter time-constant is greater than the PLL frequency, causing it to never detect lock. It comes closer to working without C2 and C7 installed, but even then has such a small window of detection that it’s useless.

A better variation would be to remove C2 and C7, ditch the entire HC7046 PLL and 74LVC1G123 debounce circuit, tie the input side of R8 and R15 to +5V, and just use the 74HCT14 output signals directly. But that basically means throw this board away and just use the sensors with a couple of schmitt trigger inputs for filtering.

For a usable version of this circuit, see the original Modulated Reflectivity Sensor Driver instead.

This QRE1113 Sensor Board can be used as a Line Sensor, quadrature encoder pickup, object detector, etc. It supports three different modes of operation, configurable, via component selection, as: Raw Sensor, Analog Output Sensor, or Digital Output Sensor.

When assembling, select the desired board mode type of raw, analog, or digital depending on your application. The first value listed for each component is the ‘raw’ mode value. The second is the ‘analog’ mode value. The third is the ‘digital’ mode value. “NC” means “No Component”. For example, for a digital mode board, use R1=150 Ohms, R2=220 Ohms, and R3=0 Ohms (jumper), and RC1=10nF (or 0.01uF) capacitor.

All surface mount resistors and capacitors are 1206 footprint for simple, easy hand assembly.

The Pinout in Digital and Analog modes is:

• 1) 5V+ Power In
• 2) Ground
• 3) Ground
• 4) Output

The Pinout in Raw Sensor mode is:

• 1) A (Anode)
• 2) K (Cathode)
• 3) E (Emitter)
• 4) C (Collector)

This board is similar to the two SparkFun Line Sensor Breakout boards https://www.sparkfun.com/products/9453 and https://www.sparkfun.com/products/9454, only more generic in that the same board can be configured either way or used as a raw sensor connection, such as for use with my Modulated Reflectivity Sensor Driver Shield.

The part selection, as detailed on the silkscreen, is identical to the SparkFun design except for the current limiting resistor for the sensor’s Infrared LED. I selected 150 Ohms and the SparkFun design uses 100 Ohms. The difference is approximately 25mA vs 38mA through the LED. The LED is rated at a max of 50mA. According to sensor datasheets, the peak wavelength output characteristics are achieved at 20mA. The higher current may help for slightly longer distance sensing, but will consume more power. You decide which one you wish to use.

For schematics, refer to the SparkFun schematics at the above links for the Digital and Analog modes. The Raw Mode is simply the four pins of the sensor routed to the connector.

The connector chosen is a 4-pin JST XHP series, 2.5mm connector. However, given the minimal difference between 2.5mm and a 0.100" or 2.54mm pitch connector, you should be able to also use a regular 0.100" pitch header connector if so desired.

NOTE: The 10K value shown for RC1 for ‘analog’ mode is low enough to keep the sensor’s phototransistor in the linear region. To push it into the switching region, use a 47K resistor instead to increase the gain nearly five times and make it function adequately for most applications. This is an example where Pololu got it right, but SparkFun didn’t, in my opinion. But, it depends on your application and whether you want the phototransistor in the linear region or switching region.