dstroy0
Zeroes and Ones
Yeah! We found him yowling in the bushes out back.
Growing stuff and doing things
- Thread starter dstroy0
- Start date
Yeah! We found him yowling in the bushes out back.
A buddy of mine is building an arcade cabinet with rotating screen so it can play cames like centipede and defender. He's got the motor on limit switches now but has to use trim to get it level. He's thinking about using an accelerometer. Thought you might have an elegant solution.
dstroy0
Zeroes and Ones
I think the best solution for this would be an absolute position sensor.
Absolute Rotary Encoders | Dynapar
Absolute encoders work in situations where accuracy for both speed and position, fail tolerance, and interoperability matters more than system simplicity. Click here to learn more about our most popular absolute rotary encoders, situations and alternatives or to talk to a specialist now!
www.dynapar.com
dstroy0
Zeroes and Ones
Using an accelerometer is slightly more involved
Using An Accelerometer for Inclination Sensing
Cell phones now commonly change screen orientation as the user rotates the phone.
www.digikey.com
edit: 100% an accelerometer is a solution to this problem too, just a little more software engineering involved
dstroy0
Zeroes and Ones
Either way, an absolution position sensor mechanically connected to the object in motion would be easier to implement in software than an accelerometer. But installing the accelerometer is a lot easier than installing a rotary encoder unless you were planning to install the encoder because then there's somewhere to pick off the motion.
dstroy0
Zeroes and Ones
There's also lidar, he could stick a target on the sides of the monitor that will present to the sensor. Might run into the leveling problem still though.
High precision indoor positioning by means of LiDAR
The trend towards autonomous driving and the continuous research in the automotive area, like Advanced Driver Assistance Systems (ADAS), requires an accurate localization under all circumstances. An accurate estimation of the vehicle state is a basic requirement for any trajectory-planning...
ieeexplore.ieee.org
We had that very discussion and didn't know how the accelerometer would handle being mounted at an angle. The math is there to calibrate it, just wasn't sure how. He also looked into it a little since we talked and apparently there are other software hurdles too.
Thanks.
dstroy0
Zeroes and Ones
youre welcome, no problem. I’m curious about the project it sounds cool.
You would dig it. I'll get some pictures. The cabinet was cut on a his 3D router table. He went through several emulator (pc and raspi based) methods but he found this thing called a flipboard that could handle the flip in aspect ratios without letterboxing.
hello little guy!
dstroy0
Zeroes and Ones
Lemon royale
looks like bumping the temps up to 84f brought the leaves up to 86f. I’m at 85% strength jacks 3.6a-2.4b-1.8g epsom. I’m not seeing any P problems, plant is a uniform green color. The root chamber is around 15-20f cooler than ambient, so having temps at 70f wasn’t doing me any favors for P uptake. I added more mg because I saw some lightening between veins on lower leaves.
looks like bumping the temps up to 84f brought the leaves up to 86f. I’m at 85% strength jacks 3.6a-2.4b-1.8g epsom. I’m not seeing any P problems, plant is a uniform green color. The root chamber is around 15-20f cooler than ambient, so having temps at 70f wasn’t doing me any favors for P uptake. I added more mg because I saw some lightening between veins on lower leaves.
dstroy0
Zeroes and Ones
I've got "over the air" OTA updates working for the esp32 environmental sensors. It connects to wifi and I can send it updates that way instead of connecting it to my computer. Next step is getting it working over HTTPS, it's built into the library I'm using and it should be trivial to implement. ESP HTTPS OTA - ESP32 - — ESP-IDF Programming Guide latest documentation
dstroy0
Zeroes and Ones
This is what I'm using to calculate VPD, and get my leaf surface temperatures.
I'm only able to get like 14-15 bits of resolution out of the ads1115, and it seems like reducing samples per second on the gray market ads1115 itself in setup gets rid of some noise.
I'm only able to get like 14-15 bits of resolution out of the ads1115, and it seems like reducing samples per second on the gray market ads1115 itself in setup gets rid of some noise.
Code:
void setup(void)
{
ads1115_0x48.begin(0x48); // Initialize ads1115 at address 0x48 3.3v max input
ads1115_0x48.setGain(GAIN_ONE);
ads1115_0x48.setDataRate(RATE_ADS1115_128SPS);
}
/*
mathematical constants
*/
#define ENUMBER 2.7182818284590452353602874713527 //euler's number
/*
thermistor settings
*/
#define THERMISTORNOMINAL 100000 // resistance at 25 degrees C
#define TEMPERATURENOMINAL 25 // temp. for nominal resistance (almost always 25 C)
#define NUM_SAMPLES 5 // how many samples to take and average
#define BCOEFFICIENT 3950 // The beta coefficient of the thermistor (usually 3000-4000)
/*
0x48 ads1115
series resistor values
measure resistance before installing and record it here
*/
#define AIN0_R 100000
#define AIN1_R 100000
#define AIN2_R 100000
#define AIN3_R 100000
/*
0x48 ads1115
thermistor 1-4 temperature calibration offset in degrees Celsius
*/
#define THERMISTOR_0_TEMP_C_OFFSET 0
#define THERMISTOR_1_TEMP_C_OFFSET 0
#define THERMISTOR_2_TEMP_C_OFFSET 0
#define THERMISTOR_3_TEMP_C_OFFSET 0
void get_thermistor_temperatures()
{
//thermistor voltage divider series resistors
const uint32_t SERIES_RESISTOR_VALUES[4] = {AIN0_R,
AIN1_R,
AIN2_R,
AIN3_R
};
//temperature offset in degrees Celsius
const float offset[4] = {THERMISTOR_0_TEMP_C_OFFSET,
THERMISTOR_1_TEMP_C_OFFSET,
THERMISTOR_2_TEMP_C_OFFSET,
THERMISTOR_3_TEMP_C_OFFSET
};
static uint8_t num_ads_samples = 0; //sample index
static uint16_t samples[4][NUM_SAMPLES] = {0}; //2d sample array
static bool calculate_temp = false; //flag is assigned true when num_ads_samples is assigned 0 (rollover)
float average[4] = {0}; //array to hold sample averages
float steinhart[4] = {0}; //array to hold temperatures
static uint32_t timer = millis(); //sample delay timer
if ((millis() - timer) >= 5UL) //non blocking delay
{
if (num_ads_samples <= NUM_SAMPLES) //only take a sample if the samples[R][C] array C index value is less than the NUM_SAMPLES macro
{
for (uint8_t i = 0; i < (NUM_SAMPLES-1); i++) //count to NUM_SAMPLES - 1
{
samples[i][num_ads_samples] = ads1115_0x48.readADC_SingleEnded(i); //start a single ended conversion, if this is not set to blocking in setup this will block
}
num_ads_samples++; //increment the index
if (num_ads_samples > NUM_SAMPLES) //if the index is greater than the number of samples we are supposed to take (last sample)
{
num_ads_samples = 0; //reset the index from 6 to 0
calculate_temp = true; //we've got enough readings to average that we can calculate the temp
}
}
timer = millis(); //reset the timer
}
if (calculate_temp == true) //if we have the samples
{
for (int i = 0; i < 4; i++) //outer for loop is where the temperature calculation happens
{
for (int j = 0; j < NUM_SAMPLES; j++) //inner for loop
{
average[i] += float(samples[i][j]); //use inner for loop to add up all the samples from one ads1115 channel
}
average[i] /= float(NUM_SAMPLES); //average the samples
// convert the value to resistance
float resistance = ((31500.0 / average[i]) - 1.0);
resistance = float(SERIES_RESISTOR_VALUES[i]) / resistance;
// convert to temperature
// magic number 273.15 is to convert from kelvin to degrees Celsius
steinhart[i] = resistance / float(THERMISTORNOMINAL);
steinhart[i] = log(steinhart[i]);
steinhart[i] /= float(BCOEFFICIENT);
steinhart[i] += 1.0 / (float(TEMPERATURENOMINAL) + 273.15);
steinhart[i] = 1.0 / steinhart[i];
steinhart[i] = steinhart[i] - 273.15;
tx_payload.leaf_temp_c[i] = steinhart[i] + offset[i]; // deg celsius + calibration offset
//c to f (0°C × 9/5) + 32
tx_payload.leaf_temp_f[i] = (tx_payload.leaf_temp_c[i] * (9 / 5)) + 32;
}
for (int i = 0; i < 4; i++)
{
for (int j = 0; j < NUM_SAMPLES; j++)
{
samples[i][j] = 0; //assign zero to samples[R][C] array elements
}
}
VPDcalc(); //calculate vapor pressure deficit with new leaf temperatures
calculate_temp = false; //we just calculated, time to take more samples
}
}
void VPDcalc()
{
/*
Calculate VPD and convert to kPa
SVP formula, e is euler's number T is temp in celsius
SVP = 610.78 x e^(T / (T + 238.3) x 17.2694))
*/
float SVP[5] = {0.0, 0.0, 0.0, 0.0, 0.0};
float vpd_kpa[5] = {0.0, 0.0, 0.0, 0.0, 0.0};
float rh;
float temp[5] = {0.0, 0.0, 0.0, 0.0, 0.0};
temp[0] = tx_payload.air_temperature_c;
for (uint8_t i = 0; i < 4; i++)
{
temp[i + 1] = tx_payload.leaf_temp_c[i];
}
rh = tx_payload.air_relative_humidity;
for (int i = 0; i < 5; i++)
{
float SVPexp = ((temp[i] / (temp[i] + 238.3)) * 17.2694);
float SVPpow = pow(ENUMBER, SVPexp);
float SVP_t = (610.78 * SVPpow) / 1000.00; //convert to kpa
SVP[i] = SVP_t;
if (i == 0)
{
//VPD = SVP * (1.00 - rh / 100.00)
vpd_kpa[i] = ((1.00 - (rh / 100.00)) * SVP[i]);
}
else
{
//LVPD = LSVP - (SVP * (rh / 100));
vpd_kpa[i] = (SVP[i] - (SVP[0] * rh / 100.00));
}
}
for (int i = 0; i < 5; i++)
{
tx_payload.vpd_kpa[i] = vpd_kpa[i];
}
}
Skunky Dunk Farms
Cannabinoid Receptor
Ahh an orange guy!
dstroy0
Zeroes and Ones
ESP32 environmental sensor code
I still need to add the differential pressure sensor and TDS sensor functions but I think that's all.
Documentation wise I need to write up the initial calibration procedures but that wont take long.
edit: it's 6mb but most of that is from the RF24 and RF24Network libraries
RF24 doc
RF24Network doc
I still need to add the differential pressure sensor and TDS sensor functions but I think that's all.
Documentation wise I need to write up the initial calibration procedures but that wont take long.
edit: it's 6mb but most of that is from the RF24 and RF24Network libraries
RF24 doc
RF24Network doc
Newly Optimized RF24Network Layer: Network Layer for RF24 Radios
nrf24.github.io
Did you end up with a thermistor you can clip to the leaf or are you doing air temp conversion?
