Toastie

Google says: ""Supra" is a Latin term meaning "above," "over," or "beyond" ... ". That was easy. Another one?

@Gunners TekZone Thanks - and yes, the BT connection clears up some mess, particularly when you feed Int.A as well as the Arduino/Mega from the same 9-12 V power supply. That doesn't even violate the strict LEGO philosophy of isolating everything on the Int.A power side incl. GND from the PC power side, as air is rather non-conductive :D Also, the Dupont type connection to 9750 (I am always using when prototyping) is a bit shaky when moving the equipment around. The circuit diagram of 9750 says "+4V" as feed voltage for the power MOSFETs of the 6 outputs and the permanent 4V output. So I guess 3.7 - 3.8V is fine. My three 9750's are all in the same range as yours on the permanent 4V output. Best Thorsten @Bliss Thank you - well, I am trying to do it as clean as I can, but that does not mean much! THAT sounds also really cool! Yes, Wifi seems to be everywhere ... don't the Shelly people use customized ESP32s to use and extend Wifi range at the same time for all their wireless stuff? There are some Shellys (Wifi power plugs) in the house, as the Huawei EMMA controller for our PV system knows how to talk to them. "No sun, no juice" control :D How about using the Arduino "UNO Rev4 WiFi" board with RA4M1 and ESP32-S3 controllers for that purpose? There are also ample of LEDs on the board itself, as shown for example here: https://www.reichelt.com/de/en/shop/product/arduino_uno_rev4_wifi_ra4m1_esp32-s3-353108?country=de&CCTYPE=private&LANGUAGE=en EDIT: And it has BT as well ... All the best Thorsten

@Bliss, all, after finally getting it (9750 inverts the sensor readings on 6/7, pulling low when active, thus 18/20 outputs go high and vice versa), I finalized my Arduino Uno R3 LEGO Blockly interface. Dedicated to @Bliss! Yes, has been done on bread boards etc. - here is the Arduino prototype board version. I found these a couple of days ago - double-sided, pre-tinned - 1$ per board. If I had known this before ... ;) The board, some call "shield", is nothing more than just an 8 LED carrier with an HC05 Bluethooth socket. I simply cut out the HC05 shape and shortened the board a bit, as there were concerns about touching things below (= the Arduino Uno). I also removed the ICSP header pins (not really required) to make space for the 20 pin header connecting Uno and Int.A. This header could also go on the component side of the prototype board, but it is then a bit "tougher" to solder pins 6 - 20 to the I/O ports of the Uno. And - it seems easier to get the combo into a case ... we'll see. Uno and prototype board attached, again, with HC05 (purely optional): It works with the Bluetooth adapter (outputs 0-5 turned on with LBlockly, input LEDs turned on with a short on Int.A's 6+7). Note, there is no USB connection, just a 9V power plug: It also works w/o HC05, just using the USB connection: Finally, some uninteresting details on soldering the thing together: The plan (made in PowerPoint ^^) And the soldering side: That is it! Now I have a BT/USB Int.A interface for LEGO Blockly, which works without Int.A connected - regarding the outputs. I can just carry the lill Uno with me when traveling ;) Input wise, bold switches, as the 4.5V touch sensors work as well (but logic is inverted). Doesn't get any better! All the best Thorsten

I am so sorry, see my edited post above. Everything works as it should. I attached my Int.A to the Arduino and it works fine!!! Regards, Thorsten

@Bliss Foremost: Thank you so much for being so helpful, supporting, and responsive!!! Good question ... my reasoning is (and that may prove completely off): When the LEDs are driven by a separate output, anything connected to the two inputs of the Arduino does not have to drive the "input (10k with pull-up) + LED". I am using blue 5 mm LEDs, which do not need that much current to light up nicely, but still, some mA are flowing. This would be no problem for the 4.5V switch, but who knows what new 4.5V sensor may come as custom additions . According to my lill' circuit diagram I made some time ago, the digital pins of 9750 go H (EDIT: wrong!) LOW when a 4.5V sensor is pressed, or the 4.5V light change sensor "sees dark": The TCLogo "Getting started" manual suggests the same: "The optosensor actually acts like a “dark” sensor. When it sees “dark”, it reports true. When it sees light, it reports false." I guess it is the other way around, so using the internal pull-ups should mimic 9750's behavior? I shall try that out later tonight. EDIT OF EDIT: Finally got out one of my Int.As (9750). And got the TC Logo manual open. Here we go: When I use this program with a touch sensor attached to input 6 the following is happening: Press = closed => Int.A's input 6 LED lights up, as expected. But also ouput 0's LED on both, Int.A as well as Arduino light up. So all is good, right? These double negations drive me crazy ... here is the updated "logic" table: * Bright means something is reflecting the IR light "better than something else". Example: With daylight present in the room the sensor LED is on = dark (although light is around). Making it even darker by putting a black box around the sensor keeps the LED on. Putting a white LEGO plate or anything else reflecting IR light, e.g., putting a fingertip directly onto the sensor "eye", turns the 9750 LED off = bright. This may be counterintuitive, but the main function of the 4.5V light change sensor is the usage with this b/w sectored technic disc right in front of the sensor. One may also use a 4.5V light brick, as it emits +90% IR light in addition to the visible part. In this case the sensor will detect "bright" even from a distance (+10 cm) and "dark" when something is between sensor and light brick. The object interrupting the light beam should not be too close to the sensor, e.g., 2 mm, as then the light of the internal IR diode of the light sensor may be reflected, mimicking "bright" conditions. It would be nice to have identical hardware/software logic. 9750 is detecting LOW on its inputs (sort of pull-up wired), this turns on the indicator LED as well as the opto-coupler LED. The latter though pulls the sensor lines 18 (20) low, as these are pulled-up as well (2.2k). It should be fairly straight forward to implement that behavior per software as well, wouldn't it? These are just ideas. When you have a LEGO model using touch/light sensors, it would be pleasant to just using the same software logic in TCLogo/LEGO Lines as in LEGO Blockly? It seems to exactly do that as is !!! Sorry for the confusion!!! Thanks again and all the best Thorsten

@Bliss I am done with my Arduino LBlockly "shield" - will show some pictures here soon. Quick question before I do that: The 4.5V light change sensor needs an actual Interface A to convert the small voltage changes it causes when it sees bright or dim reflections of its built-in IR diode. However, the 4.5V touch sensor is a bold on/off switch. This sensor does not require 9750 to induce a 0/1 or 1/0 change on the Arduino digital input pins 8/9. Would it be possible to define two more digital out pins (12/13 would be nice) and have them showing the status of the two input pins 7/8? I did the same thing on my Arduino interface code (with @maehw's mod) - just to visually check whether there is similar input action as on 9750: My Arduino interface running your V2 code has currently 6 LEDs for the outputs. Would be cool to add two more for the inputs. I can also try that myself (void pollInputs seems to be the place), but then we have multiple V2 versions which I don't like ... All the best Thorsten

My Excel 2013 agrees on that: Thorsten

Quick-fix: Sticky notes + attached print-outs of new sets, with special attention to the color of the sticky notes blending into wallpaper design. Looks goofy in TLG's view of living rooms/home offices and so on, though. Permanent fix: Snap frames/front loading poster frames. They come in all varieties, including brushed/glossy stainless steel. I would not care that much about fading colors on the printout (in contrast to the quick fix!), as this approach calls for frequent replacement of the latter. Best Thorsten

And no shelves required!

and because they have less Technic nowadays, which is always a bit challenging for the outside worlder Best Thorsten

Yes, noticed that, both versions work very well! I shall make the Arduino ↔ IntA cable and then thorough sensor testing (light change, touch) is planned. EDIT: I just ordered 10 of these bare Arduino prototyping boards (for about 1€/piece - no, not Ali, Germany ;), which a bit of luxury - usually I use what I find in my workshop. This damned 2 mm pitch on the output port side of the Uno drives me crazy though. Who on earth came up with that??? These board take care of that. And I can do some LED soldering on the board. And then make the cable, again with an appropriate plug. It is all about the looks - and as the LBlockly software is a) looking really nice and b) works perfectly well for me, this needs some attention! All the best Thorsten

@Bliss Back home, tested, works Couple of notes: Instead of the USBtoTTL converter (FTID) for HC05 configuration, you can also simply use the Arduino itself and the Arduino IDE (use the IDE's serial monitor window, with a blank Arduino script) to set the parameters for the HC05 in AT mode. That is what I usually do and learned it back in the days from here: https://lastminuteengineers.com/hc05-at-commands-tutorial/ Note however, that for the AT mode, Arduino D0=RX goes to HC05 RX (3.3V level, use voltage divider), and Arduino D1=TX goes to HC05 TX. And this is where I screwed up (again, did that before): I did not change back to BT mode (HC05 as COM port): Here, as you wrote, Arduino D0=RX goes to HC05 TX, and Arduino D1=TX goes to HC05 RX (via voltage divider) Simple as that ... sorry for causing any confusion!!! All the best Thorsten

Oh, nice! I have to try that later today, as I am traveling. Yes, I did all that as well - the fixed Baud rate for AT mode of the HC05 is 38400, and an Ardiuino (or any other terminal program works fine. I am using these BT adapters quite often, they are highly reliable. On one of my 9750, I have modified the 20 pin socket, so it delivers power to an Arduino Pico/HC05 combo. What version of LBockly is working with V1 and V2? I don't have to use different LBlockly versions, right? All the best, Thorsten

@Bliss26-1013 I guess I broke something ... I upgraded to the latest LBlockly version (2026-04-26-1013). Then downloaded the V2 version for the Arduino (this is only one file and no includes, correct?). Now I am getting the "Connection cancelled or device not responding" on both, USB and HC05 COM ports, in other words no more IntA experiments are possible at the moment. Is there a way to revert to Arduino V1 (OK, just use the Arduino IDE and upload the V1 script) - but how do I know which version of LBlockly still worked with the V1 Arduino version? And how may I download that version? I really don't care about V1 - V2 is fine for me as well, but as said, I can't connect even through USB using LBlockly version 2026-04-26-1013. Also, HC05 is just a nice to have for me, but nothing essential!!! If it does not work with the HC05, that is OK! Chances are that I am just too dumb to use the files correctly. All the best Thorsten

I used the V1 version at 115kBaud, but it does not work with the HC05 - same result: "[16:58:00] Connection cancelled or device not responding." Something is going on, as the blue blinking LED (ready for connection) goes dark for some seconds, and then resumes blinking. I also don't have a level shifter, just the 1/2.2kOhm voltage divider. So 115000 Baud may be too much ... does it work on your side at lower Baud rates, such as 38400? EDIT: I did do an update for LBlockly on your website before testing - does the new LBlockly version work with the V1 version for the Arduino? Or does the new version require the V2 Arduino version? Best Thorsten P.S.: Going to digital pins on the Arduino for IntA is really cool!