TDA7317 5-band Stereo Digital Controlled Graphic Equaliser eq_options.py

Python script Download here. Setup python below: 

      pip3 install pandas
      pip3 install numpy
      python3 eq_options.py

Run it by typing the commands above. The first two just install dependencies and you only need to run them once. Most Linux Distros probably have Python3 installed already. For Windows, install Python first, tick the box to add to PATH. Version 3.10 is what I've tested with. Then use the same commands in the command prompt (CD to the directory where the file is). Last command you may need to use python instead of python3.

Return to main article

Code below: 

#*****************************************************************************************************************
# Copyright Daniel Clarke https://electro-dan.co.uk, 27th September 2022
# Free to use and adapt but NO guarantees or support
#*****************************************************************************************************************/

import pandas as pd
import numpy as np
from decimal import *

qLower = 0.6
qUpper = 0.7
gainLower = 11.4
gainUpper = 12.6
freqLower = 60
freqUpper = 65
# Example frequency ranges for 10 band
# 240, 260
# 950, 1050
# 3850, 4150
# 60, 65
# 7900, 8100
# 120, 130
# 480, 520
# 1950, 2050
# 30, 32

e6 = [10, 15, 22, 33, 47, 68]
e24 = [10, 11, 12, 13, 15, 16, 18, 20, 22, 24, 27, 30, 33, 36, 39, 43, 47, 51, 56, 62, 68, 75, 82, 91]

# Create capacitor dataframe with e6 values from 1e-11 (100pF) to 6.8e-7 (680nF)
e6list = []
for m in range(11, 7, -1):
    for v in e6:
        dict1 = {"c1":Decimal(v) / Decimal(10**m), "c1nF":Decimal(v) / Decimal(10**m) * 1000000000}
        e6list.append(dict1)
dfC1 = pd.DataFrame(e6list, columns=['c1','c1nF'])
dfC2 = dfC1.copy() 
dfC2 = dfC2.rename(columns={'c1':'c2', 'c1nF':'c2nF'})

# Create resistor dataframes with e24 values from 1000 (1k) to 51000 (51k) and 2000 (2k) to 9100 (9.1k)
e24list = []
for v in e24:
    dict1 = {"r1":Decimal(v) * Decimal(100), "r1k":Decimal(v) / Decimal(10)}
    e24list.append(dict1)
for v in e24:
    if (v <= 51):
        dict1 = {"r1":Decimal(v) * Decimal(1000), "r1k":Decimal(v)}
        e24list.append(dict1)
dfR1 = pd.DataFrame(e24list, columns=['r1','r1k'])
e24list2 = []
for v in e24:
    if (v >= 20):
        dict1 = {"r2":Decimal(v) * Decimal(100), "r2k":Decimal(v) / Decimal(10)}
        e24list2.append(dict1)
dfR2 = pd.DataFrame(e24list2, columns=['r2','r2k'])

#print(dfC1)
#print(dfC2)
#print(dfR1)
#print(dfR2)

# Cross join all sets
dfP = dfC1.merge(dfC2, how="cross").merge(dfR1, how="cross").merge(dfR2, how="cross")
print(str(len(dfP.index)) + " results to explore")

# Calculate centre frequency
dfP["freq"] = 1 / (Decimal(2) * Decimal(np.pi) * np.sqrt(dfP["r1"] * dfP["r2"] * dfP["c1"] * dfP["c2"]))
# Reduce to target frequency range
dfP = dfP[dfP['freq'].between(freqLower, freqUpper)]
#dfP = dfP[((dfP['freq'] >= 15200) & (dfP['freq'] <= 16800)) | ((dfP['freq'] >= 240) & (dfP['freq'] <= 260))]
print(str(len(dfP.index)) + " results with matching frequency range")
print(dfP[['c1nF', 'c2nF', 'r1k', 'r2k', 'freq']])

# Calculate Q
# Reduce to target Q range
dfP["q"] = np.sqrt(dfP["r1"] * dfP["c1"] * dfP["r2"] * dfP["c2"]) / ((dfP["r2"] * dfP["c1"]) + (dfP["r2"] * dfP["c2"]))
dfP = dfP[dfP['q'].between(qLower, qUpper)]
print(str(len(dfP.index)) + " results with matching frequency and Q range")
print(dfP[['c1nF', 'c2nF', 'r1k', 'r2k', 'freq', 'q']])

# Calculate Gain
dfP["gain"] = 20 * np.log10(((dfP["r2"] * dfP["c2"]) + (dfP["r2"] * dfP["c1"]) + (dfP["r1"] * dfP["c1"])) / ((dfP["r2"] * dfP["c1"]) + (dfP["r2"] * dfP["c2"])))
# Reduce to target gain range
dfP = dfP[dfP['gain'].between(gainLower, gainUpper)]
print(str(len(dfP.index)) + " results with matching frequency, Q and Gain range")
print(dfP[['c1nF', 'c2nF', 'r1k', 'r2k', 'freq', 'gain', 'q']])
Content date October 2022