Hallo zusammen!
In diesem Thread: https://abbuc.de/forum/viewtopic.php?p=36044#p36044
habe ich eine Beschreibung und den Status des Projekts zum Zeitpunkt der Einreichung beim Abbuc HW 20025-Wettbewerb vorgestellt. Hier möchte ich euch über den Fortschritt und die Herausforderungen bei der Entwicklung der Benutzeroberfläche informieren. Da der Wettbewerb noch nicht abgeschlossen ist, kann ich nicht alle technischen Details verraten, denke aber, dass ich die Entwicklungsphasen und Möglichkeiten zur Fehlerbehebung aufzeigen kann. Ich werde auch versuchen, die im Gerät verwendeten Komponenten zu erläutern, damit sie später in anderen Projekten verwendet werden können.
Dieser Text wurde automatisch aus dem Englischen übersetzt – bitte verzeiht mir also etwaige Unklarheiten oder Rechtschreibfehler.
Mehr über TasmARI könnt ihr im oben verlinkten Beitrag lesen. Kurz gesagt: TassmARI ist ein einfaches Gerät, das über den SIO-Anschluss mit einem Atari verbunden wird und über WLAN mit IoT-Geräten über das MQTT-Protokoll kommunizieren kann. Dadurch kann der Atari als Controller für unser digitales Zuhause genutzt werden. Das Herzstück der TasmARI-Schnittstelle ist ein Modul, das auf dem ESP32-Mikrocomputer läuft – einem sehr beliebten Mikrocomputer mit umfangreichen Anwendungen in den meisten IoT-Geräten. Ich habe mich für ESP32 entschieden, weil mehrere sehr fortschrittliche Betriebssysteme für diesen Mikrocomputer entwickelt wurden, darunter Tasmota. Aus der Kombination der Wörter Tasmota und Atari entstand TasmARI. :-)
Blockschaltbild
Wie im Blockdiagramm dargestellt, verfügt unsere Schnittstelle über ein OLED-Display, eine LED-Anzeige (acht LEDs für den PCF8574-Controller und zwei für die SIO-Übertragungsanzeige) und vier Tasten auf der OLED-Anzeigeplatine. Der Zugriff auf Display und Tasten erfolgt über den I2C-Bus. Die Platine enthält außerdem mehrere Puffer, die als Pegelwandler dienen, um die Spannung der Atari-Signale an die ESP32-Eingänge anzupassen.
PCB-Visualisierung erstellt in KiCAD:
Und die fertige Prototypplatine, gestern geliefert und schnell zusammengebaut:
Erster Test (ohne Atari)
Ich plane, dieses Wochenende mit den Tests zu beginnen!
Br,
pancio
TasmARI - Atari als IoT-Familienmitglied
Moderatoren: Sleeπ, andymanone
- pancio
- Beiträge: 295
- Registriert: 31.08.2021 07:31
- Has thanked: 263 times
- Been thanked: 265 times
- Kontaktdaten:
Re: TasmARI - Atari als IoT-Familienmitglied
Hi,
The silent hero of the episode are: strapping pins :-)
I didn't manage to accomplish much with the project, but I'd like to share a few conclusions and discuss a few problems I encountered during implementation. The first Tasmari RC1 board was a bit disappointing because I didn't anticipate that the ESP32 had several strapping pins, the state of which affects how the ESP32 boots. I supposedly knew this beforehand, but somehow I overlooked one of them, and my Tasmari wouldn't boot correctly. Of course, I managed to get the board working (I had to cut and rewire a few traces), but I still had to redesign the circuit. That's how the RC2 version was born...
The board arrived a week later, and I assembled two of them at once. I was very surprised that the new version wouldn't start again... It turned out that while swapping several GPIOs, one of the strapping pins was being forced into an incorrect state by the Atari's PROCEED signal. The solution was to cut the trace and remove the signal... The module booted.
So, success! But not entirely :-)
The principle I adopted for communication between Atari and Tasmota OS is very simple: a simple program in Berry listens on the serial port and everything that comes in is interpreted as a command for Tasmota.
This time, the Tasmota operating system surprised me... it turns out that in serial-bridge mode, which is used for communication between the Atari and Tasmota OS, it intercepts and attempts to execute everything that appears on the console, while I assumed these would only be messages coming from the Atari. Interestingly, when I use the ESP32 module without the Tasmota PCB, the problem doesn't occur! I again began to suspect strapping pins.
It also turns out that the problem disappears after 10 minutes and the Atari can take over communication with Tasmota OS! That's a clue – I already know that logging from the serial port can be disabled via the SerialLog option, but it's not very pretty... I'll have to look into this a bit more.
After reading a few pages of the ESP32 documentation, I rebuilt the wiring diagram, added some pull-up resistors, designed the RC3 version, and am waiting for the boards – they should arrive this week. We'll keep fighting :-)
Oh! I forgot to mention what other difficulties I encountered... anyone who has ever tried to insert a Sikor SIO socket will understand! THESE ARE NOT SOCKETS FOR NORMAL PEOPLE! :-) It's so difficult to insert into standard PCB holes that after the entire operation, I had cut my hands and damaged two SIO sockets (and they're not cheap!). So, I decided to add a standard pinheader to the RC3 version to make it easier to check the SIO pins and also to save on the expensive Sikor SIO sockets...
Will be in touch!
pancio
The silent hero of the episode are: strapping pins :-)
I didn't manage to accomplish much with the project, but I'd like to share a few conclusions and discuss a few problems I encountered during implementation. The first Tasmari RC1 board was a bit disappointing because I didn't anticipate that the ESP32 had several strapping pins, the state of which affects how the ESP32 boots. I supposedly knew this beforehand, but somehow I overlooked one of them, and my Tasmari wouldn't boot correctly. Of course, I managed to get the board working (I had to cut and rewire a few traces), but I still had to redesign the circuit. That's how the RC2 version was born...
The board arrived a week later, and I assembled two of them at once. I was very surprised that the new version wouldn't start again... It turned out that while swapping several GPIOs, one of the strapping pins was being forced into an incorrect state by the Atari's PROCEED signal. The solution was to cut the trace and remove the signal... The module booted.
So, success! But not entirely :-)
The principle I adopted for communication between Atari and Tasmota OS is very simple: a simple program in Berry listens on the serial port and everything that comes in is interpreted as a command for Tasmota.
This time, the Tasmota operating system surprised me... it turns out that in serial-bridge mode, which is used for communication between the Atari and Tasmota OS, it intercepts and attempts to execute everything that appears on the console, while I assumed these would only be messages coming from the Atari. Interestingly, when I use the ESP32 module without the Tasmota PCB, the problem doesn't occur! I again began to suspect strapping pins.
It also turns out that the problem disappears after 10 minutes and the Atari can take over communication with Tasmota OS! That's a clue – I already know that logging from the serial port can be disabled via the SerialLog option, but it's not very pretty... I'll have to look into this a bit more.
After reading a few pages of the ESP32 documentation, I rebuilt the wiring diagram, added some pull-up resistors, designed the RC3 version, and am waiting for the boards – they should arrive this week. We'll keep fighting :-)
Oh! I forgot to mention what other difficulties I encountered... anyone who has ever tried to insert a Sikor SIO socket will understand! THESE ARE NOT SOCKETS FOR NORMAL PEOPLE! :-) It's so difficult to insert into standard PCB holes that after the entire operation, I had cut my hands and damaged two SIO sockets (and they're not cheap!). So, I decided to add a standard pinheader to the RC3 version to make it easier to check the SIO pins and also to save on the expensive Sikor SIO sockets...
Will be in touch!
pancio
- pancio
- Beiträge: 295
- Registriert: 31.08.2021 07:31
- Has thanked: 263 times
- Been thanked: 265 times
- Kontaktdaten:
Re: TasmARI - Atari als IoT-Familienmitglied
Hi,
Unexpectedly, the last few weeks have been very difficult for me, both at home and at work. I regret not having had time to share everything I wanted to about the project – I hope to make up for it in the future.
Just a quick update today regarding the change in the execution module. I decided to use the ESP32-S3, which has more features – especially more GPIO pins, which aren't specifically assigned to specific functions like the basic ESP32. The ESP32-S2 allows for greater operational freedom and easier development. Unfortunately, there's also a downside: the ability to play Atari sounds via the SIO_AUDIO input is gone (the ESP32-S2 doesn't have DACs). I'm not worried about this, though, because I can always add an external DAC via the I2S bus – which I'll do if the prototype works :-)
The current PCB is designed to support three different displays (two with buttons) – this will allow for greater flexibility in your own designs. I also dropped the LED bar indicators. Instead, I added a separate power supply :-)
BR,
pancio
Unexpectedly, the last few weeks have been very difficult for me, both at home and at work. I regret not having had time to share everything I wanted to about the project – I hope to make up for it in the future.
Just a quick update today regarding the change in the execution module. I decided to use the ESP32-S3, which has more features – especially more GPIO pins, which aren't specifically assigned to specific functions like the basic ESP32. The ESP32-S2 allows for greater operational freedom and easier development. Unfortunately, there's also a downside: the ability to play Atari sounds via the SIO_AUDIO input is gone (the ESP32-S2 doesn't have DACs). I'm not worried about this, though, because I can always add an external DAC via the I2S bus – which I'll do if the prototype works :-)
The current PCB is designed to support three different displays (two with buttons) – this will allow for greater flexibility in your own designs. I also dropped the LED bar indicators. Instead, I added a separate power supply :-)
BR,
pancio
- Dateianhänge
-
- TasMARI2.0R0.png (267.77 KiB) 260 mal betrachtet
-
- Beiträge: 763
- Registriert: 17.08.2021 21:16
- Wohnort: Reutlingen
- Has thanked: 917 times
- Been thanked: 275 times
- Kontaktdaten:
Re: TasmARI - Atari als IoT-Familienmitglied
hmpancio hat geschrieben: ↑13.10.2025 16:31Just a quick update today regarding the change in the execution module. I decided to use the ESP32-S3, which has more features – especially more GPIO pins, which aren't specifically assigned to specific functions like the basic ESP32. The ESP32-S2 allows for greater operational freedom and easier development. Unfortunately, there's also a downside: the ability to play Atari sounds via the SIO_AUDIO input is gone (the ESP32-S2 doesn't have DACs).

Welcher wird jetzt verwendet? S2 oder S3
Which one is being used now? S2 or S3?
Greetings
Wolfgang
Online
- andymanone
- Beiträge: 4207
- Registriert: 17.08.2021 12:34
- Wohnort: Berlin - Suburban
- Has thanked: 1043 times
- Been thanked: 1671 times
- Kontaktdaten:
Re: TasmARI - Atari als IoT-Familienmitglied
War wohl ein Typo.
Der S2 hat zwei DAC´s und der S3 nicht mehr.
Gtx.,
andY
Der S2 hat zwei DAC´s und der S3 nicht mehr.
Gtx.,
andY
Ressortleiter Hardware - Save the past and use the future!© andYManOne / Demozoo / “SillyVenture 2025 Winter Edition Invitro - my Tribute Cover to Buddy+PPs cool ATARI 8-Bit version!”
- pancio
- Beiträge: 295
- Registriert: 31.08.2021 07:31
- Has thanked: 263 times
- Been thanked: 265 times
- Kontaktdaten:
Re: TasmARI - Atari als IoT-Familienmitglied
It's typo - should be S3. I'm using this module (clone): https://docs.espressif.com/projects/esp ... -reference
S2 has only one single core.
EDIT: https://openelab.io/blogs/learn/ultimat ... d-esp32-s3
BR,
pancio
S2 has only one single core.
EDIT: https://openelab.io/blogs/learn/ultimat ... d-esp32-s3
BR,
pancio
Wer ist online?
Mitglieder in diesem Forum: 0 Mitglieder und 1 Gast