NLC - Tinkle (drum module) - Build notes

These are my build notes for the nonlinear circuits "Tinkle" module.

Tinkle..... 4 channel preset drum module with a distortion circuit added on the summed output.

TINKLE & STOCHAOS 1

Links;

+ Wiki

+ BOM

I found the PCB you be very " tightly packed"... so use a soldering iron with a fine tip to avoid shorts.

The drums are based on the so called “Gamelan resonator” circuit developed by Paul DeMarinis in the early 70s for his Pygmy Gamelan device. 

They are quite simple but have a lovely gentle tone and are easily tunable using different capacitors and gain resistors. 

When building you can select capacitors to suit the pitches you want & different resistor values to vary the gain. 

It would be good to build a few of these to have some variations in pitch

The distortion circuit is fairly harsh but when the drums are feed high frequency clocks you can get some quite heavy sync’d riffs going.

Each drum sound has an output. The 4 sounds are summed together and

passed through the Razor circuit for a mixed output. Turn the pot to

minimum or max to turn off the Razor and get a clean summed signal. 

Parts

TL072 or TL082 4 Soic Tayda: A-1139

LM3900 1 Soic Mouser Part No: 595-LM3900DR .

LL4148 6 sod-80 Tayda: A-1213

--------------------

The 5V1 zeners limit the output of Razors to +/-5V, they do not

have to be installed.... order

Mouser have a lot of 5V1 zeners in stock, some numbers are

78-BZT55C5V1-GS18 or 78-BZT55C5V1 or 78-TZMC5.1

Tayda have a sod-80 5V6 zener which is close enough - A-6014

I'm using a VISHAY TZM5232B-GS08 from Tayda (X-4504)

Zener Single Diode, 5.6 V, 500 mW, SOD-80 (MiniMELF), 2 Pins, 175 °C, Surface Mount

There are 2 x 33K resistor marked..... replace these with10K

 

.

When building you can select capacitors to suit the pitches you want,

also need to select a resistor value to vary the gain. Some suggested

values are listed below, but feel free to experiment.

Each Gamelan circuit uses 3 caps & 1 resistor.

There are 4 of these circuits.

For comparison, Andrew has also included a pic of a Twin-T & bridged-T circuit.

These 3 circuits are resistor-capacitor networks which are placed in the feedback loop of an op-amp.

Twin-T Oscillators (sometimes known as parallel-T oscillators) are RC (resistor-capacitor) oscillators which produce a sinewave output .

Andrew built a Twin-T  (BongO) drum module many years ago which you might like to check out.

Bong0 - Bongo - Drum / VCO build notes

The Bridge-T oscillator is an even simpler form of RC network.

It's based on a bridge-T notched filter that is placed in the feedback loop of an op-amp, 

changing it from a notch to a bandpass filter.

The Roland TR 808 bass drum circuit is based around a bridged-T network oscillator.

-----------------------------------------

All the capacitors in the build, should be the same value for each Cx and Rx sets

the gain for that section.

For example, use three 2n2 capacitors for the three C1 pads and solder

200k into R1.

Cx   Rx

1n5 180k higher tone

2n2 200k

3n3 220k

4n7 220k

10n 200k

15n 390k

22n 510k

33n 680k lower tone

I think I'll mix things up

Gamelan 1 4n7 220k

Gamelan 2 10n 200k

Gamelan 3 22n 510k

Gamelan 4 33n 680k

It would be good to experiment with this values.

I reckon it would be fun to build a few of these 

modules with other combos.

I prefer the higher frequencies.

Each drum sound has an output. The 4 sounds are summed together and

passed through the Razor circuit for a mixed output. Turn the pot to

minimum or max to turn off the Razor and get a clean summed signal. 

The Razor is an unusual distortion circuit employing unbalanced comparators

to produce pulses that are mixed with the original signal. 

The Razor can be used as a separate module. Patching a signal into the input

disconnects the summed Tinkle signal.

.

final caps 7 resistors.

S1JL, Schottky, power

rectifier or 10R

Im using a 10R resistor

LL4148

headers to connect the 2 PCBs, & eurorack 10 pin power connector.

10 sockets, 1 x B100k pot

Pair this with a VCA/EG

-----------------------------------------------------------------------------------

You can find more NLC builds here.

http://djjondent.blogspot.com.au/2015/03/non-linear-circuits-ncl-index.html

---------------------------------------------------------------------------------------

Roast 4 - Kitchen Drum roaster

 I'll be roasting some beans from PNG (Papua New Guinea)

These are Arabicas from the Central Highlands in Simbu provence.

1500m +

For a PNG Arabica from Simbu Province, the bean is typically a washed, high-altitude coffee with bright acidity, medium body, and floral/citrus notes. The roast profile can vary depending on your desired flavor balance.

https://www.conwaycoffee.com.au/

Different roast levels will give different flavours.

Light roast – highlight floral, tea-like, and citrus notes.

Medium roast – balance body and acidity, bring out caramel sweetness.

Medium-dark roast – reduce acidity, emphasize chocolate and nutty tones.

This is my planned roast.

To compensate for any inaccuracy in my roaster's temp I'll charge it at 200C and aim for a turning point closer to 100-110C. ... basically adding 10C to each of the above temps.

Here is my planned  roast curve.

The Kitchen air roaster is a drum version but the temp control isn't accurate.

I'm using a simple thermometer and probe.

It can't measure the bean temp so I'll need to make some educated guesses.

Below is my plan of attack.

---------------------------------------------------------------------------------------

Charge temp

For a 200g drum roast of PNG Arabica (Simbu Province) targeting a medium roast, a good charge temperature is typically between 185 °C and 195 °C (365–385 °F), depending on the roaster’s thermal momentum and airflow setup.

Here’s a breakdown for context:

🔹 185 °C (365 °F) → gentler start; longer Maillard phase, more sweetness, softer acidity.

🔹 190 °C (374 °F) → balanced approach; good development and clarity for PNG profiles.

🔹 195 °C (383 °F) → faster roast, slightly more body and caramelization, reduced floral notes.

In the future when I get a better roaster, I'll be more accurate.

For the moment, this is just a learning exercise.

For a small 200 g batch, heat transfer is quick — so I’m starting around 188–190 °C,  & 

watching for a turning point around 1:30–1:45.

-----------------------------------

Turning point

The turning point marks when the bean temperature bottoms out after the initial heat absorption.

Since 200 g is a small batch, heat transfer is fast — Im trying to avoid charging too hot, or the turning point might happen too soon (<1:20), causing a “front-loaded” roast with harsh acidity.

I'm aiming for a smooth, declining Rate of Rise (RoR) afterward — aiming for a steady climb through drying (~8 min total before first crack).

----------------------------------------------------------

The roast turned out as follows:

Start weight: 200g

End weight : 187.7

Water loss: 12.3.... 6.15%

It falls into the extremely light roast category.

I think the readings I'm getting from my probe don't accurately give me an idea of what the bean temp is.

It's much lower than the readings on the probe.

I think I'll add maybe 15C to the temp next roast

+ Roast 5 - Kitchen drum roaster - PNG Arabica - medium/dark roast

Time               Temp         

0 ....                200c charge 

1.30                169 TP

2                     163

2.30                153

3.00                148

3.30                144

4.00                143

4.30                146

5.00                150.... end of yellowing phase

5.30                155

6.00                163

6.30                170

7.00                175

7.30                179

8.00                181 ... end of browning phase

8.30                184  FC. start development phase

9.00                190

9.30                196

10.00              202

10.30              END roast temp 207C

---------------------------------------------------------------------------------------------------------

Coffee index. Click here :

https://djjondent.blogspot.com/2024/03/coffee.html

Roast 3 - Auto-Drum Roaster. (Air Roaster) Colombian Huila

 This is a Heller Air Fryer that I'm experimenting with to use as a coffee bean roaster.

It's a cheap kitchen appliance. ($100AUD) ... really designed for roasting food but could work.

The drum looks sturdy. But I don't think I could easily put a probe inside the drum to measure bean temp, but a simple k-probe is easy to place to measure the drum temp.

Controls are very basic & I don't think accurate , but I'll use the thermometer to get a better indication of the ambient temp inside the chamber.

I'm using 200g of a Colombian coffee.

Starting weight :200g

Finishing weight: 183.9g

difference :16.1 .... 8.05% loss  .... thus its  light roast !!!

Time 

 0 min         Charge Temp: 195C

1.30 min     Turning point 110C

5 min .........drying phase 160C

8 mins ........Browning phase 182.6C

9 mins ..........1st crack  205C

10 mins         ... Drop 

11 mins..........cooling

11 mins total

I think I could go longer or raise the temp after 5 mins to get a darker roast.

Still Lets see how this lighter roast tastes.

The temperatures in the drum chamber are certainly  alot lower than the bean temps.

I think I'll raise the temps next attempt.

My temp meter and probe.

This was my initial plan

---------------------------------------------------------------------------------------------------------

Coffee index. Click here :

https://djjondent.blogspot.com/2024/03/coffee.html

NLC fourier module - DIY build notes - part 3

 This is the 3rd & final part of my build for the nonlinear circuits Fourier module.

This Eurorack module tackles the task of waveform & additive synthesis .

Links

Wiki

BOM

Youtube

The build notes for Part 1 & 2 are here:

Fourier - build notes... part 1

Fourier module DIY-- part 2

Headers next.

25  mono sockets

B100K linear taper pots

LEDs 

+ LEDs 

I'm using clear bipolar LEDs.

soldering the LEDs . almost there.

first tests.

I just love it when the module just works the first time.

Thanks Andrew.

KOLOR - LED pixel strip controller for Eurorack

 These are some teaser videos of the new eurorack module "KOLOR" from Elkelektroik.

What is it?

Its a eurorack module that controls LED light strips, blocks, etc

With it you can syncronise LEDSs to add visuals to live modular performances

The module itself runs on standard euro +/- 12V.

The target LED strip however needs it's own power.

Its DC +5V or +12V depending on what voltage your strip requires.

It's compatible with multiple LED strips.

Control can be either manual (via the knobs) or CV patched into 6 jack inputs.

The controls are:

1. Red /hue

2. Green/saturation

3. Blue/value

4. width

5. pixel number

6. preset pattern

Ed will be releasing this on Kickstarter soon.

For more info:

https://www.elkelektronik.com.au/kolor

Insta:

https://www.instagram.com/elk_elektronik/

I'm changing some of the presets & altering the speed with a LFO

I'm using a Erica Synths Pico system III

the module uses a WS2812 LED strip.

This is a programmable LED strip where the control circuit & the RGB chip are combined into one package.

Each LED is addressable. You can daisy chain the LEDs and control them from a single module.

There are lots of different LED formats.

They can be square, rectangular, circular, strips etc.

...These modules are pixel controllable.

Ie, you can control right down to the pixel level.

A bit of random noise.

A bit about the power source for the LED strip.

Euro power is precious, and LED strips can suck up lots of power.

So to preserve the valuable juice for your other modules, Ed has added a second power input just for the LEDs

The very bottom on the left has the input for the power for the LED strip/circle etc.

This can be +5V DC or +12V DC (depending on your LED strip requirements).

The Green rectangular jack on the bottom right is where you connect the LED strip.

There is also a overload switch.

here is a closeup of the Power section for the LED strip.

You need to connect the 3 wires to the LEDs.

DIN = Data in

+5V & Gnd are self explanatory.

On the other side of this LED tile you will see the Dout , +5V and Gnd connections.

These allow you to chain tiles.

NLC Fourier module DIY - part 2

 This is part two of my Fourier module build.

Part 1 is here:

• NLC - Fourier - build notes... part 1

It follows the basic principles of Fourier Synthesis, and being able potentially to design your own waveforms. There aren't many modules that do additive & Fourier synthesis so I'm looking forward to finish this.

The pots (or inputs) allow control of 7 sine (s1-s7) and 8 cosine (C1-C8) signals...... supposedly

being harmonics of the initial signals. “Supposedly” because, the waves

are not pure sinewaves, rather stepped pulse waves.

This module would be well paired with a oscilloscope so you could see the waveforms that you sculpt.

Andrew also reckons that this module needs to be patched into a low pass filter.

Links

Wiki

BOM

Youtube

Nonlinearcircuits Builders Guild on FB:

https://www.facebook.com/groups/174583056349286/ 

074 opamps

resistors ... looks like the resistor matrix is complicated but there are only

three values... 

there are 109 x 10k, 23x 27K, 55 x 15K .

1k resistors on the second PCB

The TL072 opamp & diodes ...LL4148

Im using a S1JL for the power protection.

The two 12k resistors set the output gains for the 2 channels. 

I'm using 12K.

"At some settings the outputs may seem a bit quiet, though generally you

will see 10V p-p. If you want the outputs hotter, install a higher

value, say 22k" (Andrew F).

I'm using bipolar red/blue 3mm LEDs, so 3k3 is my value for RL.

These leds were the same used on the NLC sloth modules.

Final caps & resistors are on.

to be continued....