E-TKT: anachronic label maker

Andrei Speridião 2022 - http://andrei.cc

📰 news - August 2022

• E-TKT is a finalist at the Hackaday Prize 2022, in the Hack it Back category;
• Project of the week on Maker Update #289, by Donald Bell;
• It has also been featured at Hackaday, Make and Hackster;
• E-TKT is certified by OSHWA;
• A manufactured PCB is being developed and the files are already committed, but they are still a work in progress. Aka: not stable and with issues. I'm working on that at this moment. Meanwhile, please refer to this readme for reference on the previous model;
• Huge thanks to Wesley Lee for the assistance on the PCB design and manufacture!

🎯 To keep up with detailed updates, please stay tuned on the Hackaday page logs 🎯

table of contents

• what is it?
• why?
• how does it work?
• features
• components
• 3D printing
• to do
• process
• libraries
• license

🙃 what is it?

étiquette f (plural étiquettes)

"ticket, memorandum, attach, stick, pierce, sting, to be sharp, goad, puncture, attach, nail"

1. tag, label
2. prescribed behavior

E-TKT is a DIY label maker that mixes both old fashioned and contemporary technology to create something as simple as... Labels!

🤔 why?

TLDR: curiosity, technical challenge and of course organizing stuff.

The initial spark for this project came from an ordinary manual labeling device that I bought to organize my workshop. I was VERY UPSET when I noticed that it was basically rubbish, getting broken and unusable after just a few operations. This little plastic hook (circled in red) grabs a gated cog (the white plastic part below) to feed the tape forward. I tried to fix and reinforce it, but the failure was unavoidable because there seems to be a critical flaw in the project: it is too weak for the lateral resistance imposed.

I knew I could easily buy a brand new modern electronic thermal label printer (such as Brother), but something about the vintage embossed finish fascinated me. Also, these new models use expensive proprietary cartridges that create a lot of trash after use.

After all, even though the pressing mechanism was poorly made, the characters' carousel was pretty sturdy and sharp. So I thought: can I build over its principles to create a functional device? If so, then why not make it physical-digital?

...why bother mixing an archaic printing method with current automation and connectivity features? Because why not? It was a technological blind spot to be explored!

...anachronism 🤓

I've later realized that it was what I’ve been attracted to. According to Wikipedia:

“An anachronism (from the Greek ἀνά ana, 'against' and χρόνος khronos, 'time') is a chronological inconsistency in some arrangements, especially a juxtaposition of people, events, objects, language terms and customs from different time periods”.

The point is: even though the process is digitalized, the resulting label is totally old school and there is no easy way of telling if it wasn't made in the traditional way.

⚙️ how does it work?

An ESP32 commands the label production and also serves an on demand web application to any device connected in a local network.

Neither internet is needed, nor installing any app.

The whole process of connecting the E-TKT machine to a local network and then launching the app is aided by a small OLED display that provides instructions and a dynamically generated QR code with the URL, according to the IP attributed by the WLAN.

The web app provides text validation, special characters, a preview of the exact size of the tape, an option to select the desired lateral margins and also real-time feedback during the printing (also present on the device screen). There are also commands for attaching a new reel, manually feeding and cutting the tape.

The label production itself uses the same mechanical principles as the original machine did, but is now automated. A stepper motor feeds the tape while another selects the appropriate character on the carousel according to a home position acquired by hall sensor. Then a servo motor imprints each character by pressing the carousel to the tape. That happens successively until the end of the desired content, when there is a special character position to cut the label. A light blinks to ask for the label to be picked.

💡 features

🌟 highlights

• No need for internet, app installation, data cables or drivers;
• Use from any device that is connected to a local network and has a web browser: desktop, tablet, smartphone;
• Compatible with ubiquitous and cheap 9mm generic tape (DYMO-compatible);

📱 web app

• Instant preview: what you see is what you’ll get;
• Real time check for character validity;
• Label length estimation;
• Margin modes: tight, small (1 space each side) or full (max input length);
• Buttons for special characters: ♡ ☆ ♪ $ @
• Reeling function: for when a new tape reel must be installed;
• Manual commands: feed & cut;

🔌 device

• Minimum label size to allow for picking it up;
• OLED screen + LED feedback:
• Instructions for configuring the wifi;
• QR code/URL for easily accessing the web app;
• Real time progress;

🧩 components

The estimated cost is around $70 (USD) without shipping, as of May 2022.

⚡ electronics / schematics

Power

• 7-12v to be provided by an external power supply with at least 35w to deal with servo peaks current while pressing the label. It supplies the stepper drivers directly.
• 6v out of the L7806 step down and is provided for both the servo and hall sensor.
• 3.3v is provided by the ESP32 WROOM board (as in its logical ports).

Logic Level

• As the ESP32 uses 3.3v logic, we need this conversion for parts that are running on higher voltage (servo and hall sensor).

Press

• The servo uses 6v, higher voltages tend to damage it.

Carousel

• NEMA stepper, driver and a hall sensor to match the position origin.

Wifi reset

• A tact button that when pressed while booting, clears the saved credentials.

Feeder

• Reduced stepper motor along with its standard driver.

User interface

• Running on 3.3v an I²C OLED display and two LEDs (no need for resistors).

🧵 3D printing

16 parts in total, using approx 200g of PETG filament.

📐 All the .stl provided are in the correct orientation for optimal printing:

• A_bottom
• B_wall (support)
• C_wall_track (support)
• D_pillar_1
• E_pillar_2
• F_pillar_3
• G_pillar_4
• H_pillar_5
• I_top (support only in the K_top_tapefeed area)
• J_top_screenholder
• K_top_tapefeed (support)
• L_caroulsel_cube
• M_carousel_hallholder
• N_carousel_coupling_1
• O_carousel_coupling_2
• P_press

🎛️ Settings

• Layer height: 0.25mm
• Infill: 20%
• Wall line count: 3 + 1 (Cura usually adds the later together with the infill)
• Top/bottom layers: 4

🛠️ to do

• [ ] Bugfix
• [ ] Manufactured PCB

🌑 process 🌘🌗🌖🌕

• Using Arduino mega, communication via serial.
• Carousel homing with infrared sensor + one missing “teeth” led to inconsistent results and misalignment.
• Testing with SG-90 servo, but it was too weak to press the tape.
  
  

• Experiments using ESP8266 with self served app + receiving commands, still isolated from the main functionality.
• Wifi manager.
• First sketches for the web app user interface.
  

• Stronger MG-996R servo resulted on successful tag printing.
• Hall sensor for homing with precision.
• Carousel direction on clockwise only to avoid tape screwing.
• Migration for the ESP32.
• Dual core tasks: one serves the network/app and the other controls all physical operations.
  

• More compact device, with smaller stepper motor for the carousel.
• Tested using NFC to open web app on smartphone, failed (security standards?).
• OLED display to help on configuration and feedback.
• QR code as an alternative to quickly open web app.
• Improvements on web app usability and visuals.
  

• Implemented special symbols.
• Few printing optimizations.
• Documentation with texts, schemes, photos and video.
  

🔤 label printing improvements along the way

🤸 all the 3d printed iterations

📚 libraries

Framework: Arduino

Environment: NODEMCU-32s @3.3.2

• waspinator/[email protected]
• madhephaestus/[email protected]
• ottowinter/[email protected]
• alanswx/[email protected]
• olikraus/[email protected]
• ricmoo/[email protected]

⚖️ license

MIT @ Andrei Speridião

If you ever build one, I would love to know ;)

...for more: andrei.cc