Eurobricks Dukes
  • Content Count

  • Joined

  • Last visited

Everything posted by Toastie

  1. @evank Well, then here we go ... Born '62, my first LEGO set was #323 - the push along train (I still have). No Technic sets for me until entering my dark ages in the late 1970's - resurfaced from that period in 1996 and "found" #8485 (Control Center II) dead cheap on sale. But for me, the best Technic set of all time is #9719 (LEGO Mindstorms Robotics Invention System 1.0). It changed everything in my LEGO world. Best, Thorsten
  2. @evank: Do you want to break out the real Technic, or should I? Best, Thorsten
  3. Toastie

    3rd Party Controllers and Powered Up!

    Hehe - OK. Now: First party is TLG, right? And third party is ... whoever exists in non-first party land other than ... me/everyone? Is that right as well? So second party is - personal party, correct? If so, this is definitely the way to go! All the best, Thorsten
  4. And that is a blast! I recently "returned" to my beloved Sinclair ZX81 and Spectrum, the latter having a RS232 port - they all live here in my attic - along with a PCW8512 (Joyce), an IBM XT and an Atari 1040 - all from around or before 1985 - almost 40 years old (next year is going to be THE big celebration data party up here - OK me, a bottle of Captain Morgan, and the other oldtimers) - and all going strong, as you said. And talking happily to my Win11 laptop. 9600 baud - the >real< light speed and true benchmark in serial data transfer world ... You should dive into LEGO Interface A world - some here on EB know much more about than I do! That is so cool; 1986 LEGO Technic ... controlled by either of one of the above oldtimers - or a Win11 laptop. Sooo cool. Have fun! All the best, Thorsten
  5. Woo-hoo!!! OK, getting sentimental - but: One of the things that drive me, is getting "good old" and "apparently new" stuff together. And the serial port (now in the incarnation of USB) is the way to literally talk to - ancient teletypes, in a decent way. Or the very first computers. I love the original handshake approach, the original serial interface implementation (25 wires ...). You know that 2 + GND wire (+ XON/XOFF in software) is just a matter of hardware speed, but essentially same old thing: When you are busy, I won't talk. And vice versa. And now you have the LEGO Interface B at your fingertips! Simply wonderful. Have all the fun with it! Or: Ready Player One! All the best, Thorsten
  6. Congratulations! That is very good news! I actually believe that you can use a Win11 computer with such a dongle and a terminal program for starters! Also, the VS C# program, Tom is providing on his website ( should work! Very cool - I love it. All the best, Thorsten
  7. Hi Lars, can you share that schematic as well? For sure the UART in the SAB chip gets confused with the serial input, but I would not know how to trace that to a faulty power board without knowing how that is wired. What is the primary power supply you are using? Should the ripple etc. a problem: Do you by any chance have a stabilized 12V DC power supply, let's say around 1A? That would remove all ripple, and the diodes take care of polarity issues. However, first it would be good to have a look (sketch is enough!) at the power delivery section (the small board) before making any further suggestions. Best, Thorsten
  8. Hi Lars, you need to use an image host platform (e.g. bricksafe etc.) for storage and then c/p or embed the image links into your messages. Best, Thorsten
  9. That IS a nice document! According to the datasheet of the Siemens SAB chip ( you need to supply 5V DC as well; is that active? Best, Thorsten
  10. Hi Lars (@LH4PI), could you provide that link, where this Italian guy is mentioned? The search function of EB is not of much help to me in this case ... I don't own an interface B - but before you go diving into the innards of the IB electronics, particularly the Siemens SAB 80515 chip ... in 95% of all my RS232 "troubles", it was the serial port/cable wiring. I did not really look into the details yet, but are you sure that the 2/3 crossing and 5 (GND) straight through (5/5) is all there is? Could very well be, but true null modem cables also cross 4/6 and 7/8 (for 9 pin hardware handshake). LEGO was always creative and did some tricks to make you use their very cables, as others did not work. @BatteryPoweredBricks: Is the 2/3 crossing the only thing happening in that cable? It does not mean anything should there be more wires used, but, well it could be. TLG did something nasty it on the totally dumb Mindstorms IR tower as well; this one only uses 2/3/5 for all functionality, but without a 7/8 bridge within the tower, the original LEGO software refused to work. All the best, Thorsten
  11. Toastie

    7710 with Red and Black motor

    12V voltage - I could not resist, Emanuele - Alessandro Volta, the brilliant Italian physicist ... The current guy is from France - also brilliant, but wait - you are right: It is the current driven by a 12V voltage. Ha! You just left out a word like induced or driven ... I get it! All the best and say hello to the family! Yours, Thorsten
  12. TLG's tricksters - always in for a special treat, aren't they? Particularly when it comes to cables and electronics Well, one of these lill' sub-D 9 pin female/male null modem adapters + your straight through serial extension cord will do as well. Amazon has them and many other sources (Ali Ex for a fraction of the Amazon price of course) Best, Thorsten
  13. Toastie

    PF Lights on 9V Motor

    Oh yes ... here is to TAE (no Nobel Prize for him, but he made it work for everyone, and that is much more important ). Any tiny tungsten wire will light up either way; that is the simplest and most efficient solution. I love it! If you want to go LED - the bridge rectifier is your friend. Will drop the voltage by about 1.2V, but who cares at 9V supply voltage. Best, Thorsten
  14. Can only agree with @Vilhelm22 ... I am speechless. See, in the photograph with the three attached locos I am not sure, but the other photographs certainly don't look like LEGO but the real thing. Wow. This is perfection to the very end. Congratulations!!! All the best, Thorsten
  15. Dear All, one of my favorite themes (in many regards) is Steam Punk. Steam-propelled futuristic technology, in shiny brass, gold and silver, leaning towards Jules Verne's "fiction". Not the dirty dark, maybe original punk. I am not sure whether the shiny stuff is Steam Punk at all; this genre has taken on so many forms and interpretations … Whatever, it may well be that my affection for Steam Punk comes from my admiration for Thermodynamics – in chemistry of course So guess what happened, when in late 2023 I became aware of the #85007 set “Steam Punk Railway Station” from Pantasy/China? Yes, I got sort of positively nervous. It quickly turned out that the title of the set is misleading: They use the railway station shown on the front of the box and the two instruction booklets simply to hide the Steam Punk locomotive along with one carriage. I almost freaked out, when I saw the train. And finally I passed out when I saw that the train is of the suspended monorail type … for more than two decades I was dreaming of a suspended LEGO monorail train – read more about that here:;do=findComment&amp;comment=3680627. With respect to the Dark Side: Pantasy is using GoBricks pieces (as per advertisement on the box) – and these have reached the 100%(+) LEGO quality level, no doubts. OK, as expected, has been discussed on EB. Very small injections points (barely visible), perfect colors, perfect clutch, same “softness” as the latest LEGO bricks and plates have, metal gold and silver pieces all over the place. On their website (, they talk about their approach, which is a quite nice read. Building that set was pure and utter fun. It was as if the Pantasy management said to the designers: “Folks, go Steam Punk nuts – no restrictions, just freak out and have fun. When you need more pieces to make it even more glitzy and punky – just do it”. 2723 pieces, BlueBrixx asks for €140, you can get it much cheaper – it is currently almost everywhere sold out, though. No surprise. This set is a blast in every regard. Motorization of the engine From day one, I wanted to motorize this model – it appears to be designed for display only. There are a couple of challenges to do so, though: For one, any propulsion system needs to turn the wheels on any of the four roller brackets the locomotive and the carriage are attached to. Second, available space is rather restricted, as the locomotive is essentially one “solid” piece of snotted ABS – except for the magic glass boiler: (At first, I thought: OK, TLG's well known color vomit coding here as well – but then was told that I was totally wrong: It is right here, where the Steam Punk magic generates the power to propel the locomotive: Inside the glass boiler. Attached is a 10 cylinder V-motor connected to the 8 blade pearl gold propeller on its back side – which actually propels the train. Well, there is no color coding at all in the entire set, except for the boiler brackets. So yes, these colors very well may represent the Steam Punk magic). Third, the elevated track Pantasy used is of the roller coaster type, JK Brickworks also employed for their suspended train ( and thus may be no good choice for motorizing a 700+ g heavy Steam Punk train; this will certainly will fail in roller coaster curves. The original (four individual) roller brackets look like this ( render) In conclusion, there wasn’t any space other than within the boiler for motorization, but I had no clue how to get any gears, axles, rubber bands to the wheels of the roller brackets. A medium PUp motor may fit in there, but the next challenge would have been PUp cabling and powering. The carriage also provides some space, but no LEGO hub fits in there, even after hefty modding, at least in the lower league I do my builds. But then I remembered a post not mentioning the elephant in the room, a thread @zephyr1934 created back in early 2021 on EB (, which deals – in addition to the Elephant in the room – also with the Circuit Cube BLE hub and Cubic motors from Tenka ( Furthermore, @Asper reported on using these (@HoMa called “critters” ) for train automation of his four wide trains, whereas @Ts__ showed a prototype truck for train displays. I looked at the dimensions of the motors (the Cubic motor is 2x 2x4) and the hub (2x 4x4) – and … BOOM – brain currents got ballistic. Furthermore, the wiring between the motors and hub is done with Dupont connectors. Tenka uses 2-wire sockets receiving the corresponding two pins of the hub outputs and motor inputs; however, these sockets are too large to fit through a Technic hole. One can easily make single socket wire ends, which then do fit very well. After some tinkering, MLCad suggested: It may work. My plan was to a) motorize the 10 cylinder V-engine from “within” the boiler section, b) propel the entire train with a motor mounted on a 2-axle monorail bracket assembly, c) route the cables somehow to the hub, which is to be mounted in the carriage in a way that one can easily turn it on and off – and d) easily recharge the hub … so far, a) to c) have been accomplished; d) is on the to-do list, see below. Here we go: New roller brackets for the locomotive and the carriage, and the entire driving motor assembly (render and photographs below). I used two stacked 9V train wheels arranged in the same direction – this way, the (custom) rubber band on the driving wheel has good grip. The rubber O-ring on the powered axle for enhanced traction is one I found in the lab – the white LEGO O-rings do not provide sufficient grip, even on the 12V rails “with teeth”. Here is what I use as my monorail track (no Dark Side elements to be seen, all pure LEGO): Photograph below: One Tenka Cubic motor “inside” the boiler, with an attached axle extending all the way to the front, rubber band connecting to the 10-cylinder motor driving axle. (Custom) cables attaching to the motor are routed through Technic holes in the boiler bracket and locomotive frame. They extend to the Circuit Cube hub mounted in the carriage. Pantasy designed the boiler essentially as an all-four-side-studded central structural element. All other parts attach to this element and, at the same time, the “side panels” secure the whole model from falling apart when lifted through snotted connections. You can remove the two large “side panels” - and are left with this when they are put together (and yes, good clutch power really shows): This is the entire engine … ... and the lightly modded carriage: The photographs below are showing the Pantasy railway station, designed as a terminal station. Some folks were complaining that the four platforms (labeled 2 to 5, prints of course - there are no stickers in this set) don’t make any sense, as the monorail runs perpendicularly to the apparent platforms. Of course this makes sense : Below are four tracks terminating at the station: (track 2) 4.5V track – coal delivery for steam locomotives on track 4 and maintenance; (track 3) 9V track – passenger and freight trains; (track 4) 9V track – real steam passenger trains; (track 5) 9V track – freight trains. The monorail is of course floating above all that; when the suspended train is leaving, all these tracks can be serviced ... duh. I thought this is the whole point … The Pantasy Railway Station … Station “unfolded”: Have a look inside: Ticket counter (right) and barista coffee stand (left) … when you push the 3-elements revolving door, the golden (Technic break) disc and the clock hands are spinning as well – this calls for additional motorization … Train 85-007 ready to depart … no surprise, should James Bond be on board … After arrival at the other terminal station: This is the Circuit Cube doing all the work; charging is from the top (see below, though), the on/off switch is easily accessible from the back of the car. To-Do list: I need to find a way of controlling the Circuit Cube hub with an ESP32. This has been demonstrated by @Asper; he has provided his code on GitHub ( Steffen used a tiny M5Stack Atom and used the Visual Studio Code IDE with the Platform.IO extension described by him here: I’d like to use the same program I wrote for my Crocodile; this was done using the Arduino IDE and Cornelius Munz’ Legoino code ( The reason is simple: It works with 4 optical sensors I want to use again for stopping at both terminals and acceleration/deceleration phases. We will see how that shakes out For charging the Circuit Cube, I need to make a custom USB cable (space … need a flat top USB-B plug). Make a longer video showing how powerful the Circuit Cube stuff really is. It is simply amazing. Video wise, I only have this less than 30 sec clip: First, the 10-cylinder engine fires up, then the train departs, totally loses focus, stops somewhere , V10 stops as well as it needs to kick into reverse (yes, the motor goes into reverse, this is a Steam Punk motor, no gear box here – you can see that by carefully looking at the pistons, they go down and up upon return, not up and down when it started), train reappears, totally out of focus, stops and gains focus, motor shuts down. I suck at taking videos, I know. Steam Punk Train.mp4 (Updates will go into this thread) All the best, Thorsten
  16. Thank you very much @Asper! Well, it would be a good project as is, when the monorail would be a full circle or whatever shape but closed - then the Tenka app would suffice for running the train in a controlled fashion. But I simply don't have the space for that closed loop. However, with the help of your code for controlling the CicuitCubes via an ESP32 based microcontroller, it eventually may become a good project. My plan is to have it (forward) accelerate out of the main station just to decelerate into the "remote" station and then go in reverse to the main station. I did something similar with my PUp/ESP32 controlled Crocodile using IR proximity sensors. In this new setup, I like to try these small forked light barrier type sensors - for fun. So far I am programming the ESP using your BLE includes in totally BASIC-style C/C++ - global variables all over the place, just to mention one no-go ... so far I can control the CC connected motors individually using the LEGO PUp remote; the ESP is waiting for two H-bridge motor drivers to spin the gears on the center support post and the door of the station. We'll see. All the best, Thorsten
  17. Welcome to EB! Google fails me to find a model 710 12V LEGO transformer - it finds 740, 741 ... but then others here may know better! As far as I am concerned, any adjustable 12V DC power supply providing high enough amperage will do the trick. If you have a Voltmeter, I'd check the output voltage first. On the other hand, when you have a 110V -> 220V adapter at hand, suitable for driving inductive loads(!) (a transformer is an inductive load and some older voltage converters don't like that, they were designed for driving resistive loads only, such as old incandescent light bulbs), that sounds also good to me. Best regards, Thorsten
  18. Toastie

    Research Lab - Materials Chemistry

    Chemistry largely IS black magic. PChemists use thermodynamics, kinetics, quantum mechanics, modelling etc. to touch it up as natural science That is why I like it PChem so much (and hopefully none of my colleagues are on EB ) Best, Thorsten
  19. Damned, forgot: The TLG website talks about attempts to do exactly that. I believe - just from reading between the lines of this TLG text: "In 2021 we had an exciting breakthrough and created a protype brick made from PET plastic bottles. However, after two years of continuous development it did not deliver the overall carbon reduction required to realise our ambitions" ( They continue with: "We’re proud of our achievements with the prototype and will apply the learnings as we continue to develop new materials and explore other ways to make our bricks more sustainable." OK, my conclusion is: There is a breakthrough, and they are proud. But they do not use regrinded/recycled material to make LEGO parts. Which is absolutely fine - as >we< want that the parts "endure decades of play and be passed down from generation to generation." (same reference). See below for CO2 balance ... It simply does not work that way. Chemistry-wise. Well, there is also Thermodynamics telling you: Yeah, can be done. But look at your energy balance. Hint: "The first and second law". Well, that happens, when you "invent" (TLG did certainly not, they simply use it) a material, that is way down on the thermodynamic scale with regard to "endurance" and need ultimately clean and defined starting materials. Recycling without indefinite energy supply means: Degradation. Not tolerable in LEGO world. Don't you agree? With very best regards, Thorsten
  20. And what about the owner of the machine? "They" (TLG's research department) can surely fine-tune, couldn't they? I'd do it, if I were the owner. Secret recipes provide IP. For sure! But again, that is part of the research of the owner of the machine, isn't it? And there we come to custom engineering. Precision is only one aspect, size and location of injection points, as you point out, is another. But these designs are again in the realm of the owner of the machine, right? I may piss off people here, but I had a "web-look" at the GoBricks site. These uncountable blocks of metal resembling "molds" ... I am asking, because I am really curious (nerd here!). I gather my information from the web for sure, no real-world experience how molding really works. Yes, it takes a little longer, as I am not fine with one reference only, and I like to see some consistency, coherence and logic. The only field, I feel somewhat confident, is chemistry, polymers, companies in that field. And OK, maybe some >limited< machining experience; we need to make optical and vacuum parts that rely on the micron scale and below in our research. Way more importantly: Thank you a lot for your insights, which I sincerely value! All the best, Thorsten
  21. Hi Thomas, thank you very much, I really appreciate that - from a master builder you are (and a very polite individual of course). As we are in the Circuit Cube thread, I'd like to share just a few experiences with regard to BLE access to the cubes. Legoino for sure has it (potentially) all, but in this case it is >completely< overblown: All the numerous service routines even the "small" LEGO City hub provides is breathtaking and Legoino handles them all. However, CC's are aiming at a totally different target: small, very small efficient builds or better: driving mechanisms, which simply don't need all that Legoino overhead. Steffen's (@Asper's) really slick code is perfectly matching with that target. I do not understand >any< of the BLE related code Steffen provided us with, nor do see any parallel to Cornelius' code other than, as Steffen pointed out, gathering information from Legoino. Steffen's code feels (have to use "feel" because I can't rationalize it) to be so on target, it is fun to simply read the code lines. Moreover, when leaving this "deeper" code layer and resurfacing to the "user level" (Steffen's main.cpp) - it still remains being slick and very elegant. As said, I am in awe. Steffen's code should get a name - Circuino would be lovely, IMHO. But it does not compile in the Arduino IDE (file extension .ino); I believe this is caused by the different BLE libraries used(???). Steffen uses "BLEDevice.h" (from library ???), Cornelius the NimBLE-Arduino library. I bet, Steffen's code will run within the Arduino IDE as well, when changing a few (or more ^^) things. Circuino - sounds like the magic you experience in a circus ... absolutely appropriate name! What I am saying is that yes, VSCode/PlatformIO are certainly (much) more powerful than the Arduino IDE, but the latter - this is my view only - seems to be more in line with some one-time, small breadboard solutions (I know, you can do so much more with it). What I am really saying is: Any chance, that Steffen will provide us with Circuino? Wait - someone in the ... what was it ... CaDA Forum, talking about retro-brighting, offered me a "Kasten Bier", if I could tell him how to bright-up some LEGO pieces, which so far, remained unfavorable attempts of doing so. So, Steffen: Would a Kasten Bier do it? (OK, I know, probably not) All the best, Thorsten
  22. Yes, true, saw that also before starting to wonder into monorail world. The two things for me though were: How do I >not< need a support every single 4.5/12V rail connection? My monorail had to cross this LEGO water region ... and secondly: How do I mimic the Wuppertaler Schwebebahn approach, with this doubly flanged wheel running on top of the rail? In other words: This was not about using 4.5/12V track for a monorail, but rather how to get this Pantasy Steam Punk train moving in Schwebebahn fashion ... Best, Thorsten
  23. Ouuu - Flensburger OK? You are talking about white pieces, right? Colored pieces are so much more difficult, as it depends on the "pigments" or "dyes" they used for coloring the polymer - you can easily oxidize these to colorless as well ... OK, when using H2O2 for retro-brighting, you don't need any heat - but UV light ("sunlight" slowly does it, but sunlight has more than 90% of radiation causing heat rather than H2O2 dissociation) or UV LEDs - they do it. Next: As high as possible concentrated H2O2 solution. The higher the concentration, the lower the UV exposure time required - otherwise H2O2 begins degrading the backbone of the polymer itself, not just the reactive sites causing yellowing. It seems there are two threads currently (actively, there are numerous non-active) discussing retro-brighting ( and this one - which may be prone to off-topic moderator flagging Vanish needs heat though to start producing active oxygen (= O atoms). Well, it is used in washing machines - and there is not much light in there. And sure enough: a) When the removal = saturation of active sites in the polymer is not deep enough, re-yellowing is swiftly occurring, b) when the polymer (as e.g. my Atari case) seem(!) to have many remaining double bonds - it simply will just start again. Depends on the ABS formulation they used when making the polymer. Some formulations are prone to yellowing (my Atari), some others are not (my IBM), and as far as I am concerned, this is due to the presence of non-aromatic double bonds in the outer molecular layers of the polymer. Hmmm - that blurb does certainly not qualify for receiving the Kasten, does it? Best of luck though - chemistry is fun, weird, but mostly experimental trial and error. Still in 2024. Best, Thorsten
  24. I believe this whole retro-brighting approach has so many aspects, it is really tough to tell, what works in what "case". I just wrote a way too long post (and I bet I will get a lot of flak for that content) in the Technic forum/CaDA (;do=findComment&amp;comment=3681951), mostly touching on ABS as such and (scroll down - yellowing and retro-brighting. The entire process is based on oxidation - Vanish = "O2" oxidation, H2O2 = "OH" oxidation. The yellowing is - as far as I am concerned after extensive literature surveys - breaking double bonds in the ABS polymer, leaving a single bond and free or better delocalized electrons. These cause the absorption of a portion of the "white light" shining on them - leading to yellowing (white = no absorption of "white" light). And they can be "removed" upon oxidation = making a new bond with "O2" or "OH" which does not absorb any of the visible light wavelengths. In other words: Whatever active oxidation compound can penetrate through the ABS material to "remove" or better "bind" the free/delocalized electrons. H2O2 is quite good, but you need to crack it first into two OH radicals, which is easily done with UV light. Vanish is activated by heat to yield oxygen "in statu nascendi" = O atoms rapidly recombining to O2 (and then they are "dead" with regard to oxidation power). So the O atoms initially produced upon heat mediated decomposition of Vanish need to find these free electrons before recombining to O2. H2O2 does not decompose into 2 OH radicals with heat, but with UV light. In other words: H2O2 needs UV light to activate, Vanish heat. Mixing both + heat + UV light may accelerate the overall oxidation process. I'd use as high as possible concentrated H2O2 aqueous solutions + UV light (sun/LEDs). Problem is: The authorities don't like to let us easily get 80% H2O2 solutions - you can make so many really bad/cool things with it. Just look up "The Blue Flame" for example, which is totally cool. NASA is also fond of it. And: The 80% solution works fast - and then may bite into the plastic as well. As said, retro-brighting is an entire philosophy among the retro-folks (computers, game pads, whatnot) Best, Thorsten
  25. Well, I'd like to challenge that, knowing a little about ABS, yellowing of plastic material, regarding reaction mechanisms - it is part of my daytime job. I am not saying that I know anything to the extent that it is "true". No way. However, there are few things, that the chemical industry delivers to customers that I know of for sure. First: ABS is characterized to the "last digit", when it comes to "arriving at desired properties". Desired as in: You, the customer, provide a set of characteristics you want the ABS polymer to have after molding it with a temperature and pressure programmed adjusted process. The (usually large scale supplier) offers you a list of possible compositions. Then the trial and error process begins on your side - let's call it fine-tuning. This is up to you. In most cases, though, you get the (almost) right composition, depending on how "good" your list of your envisioned characteristics is, and the supplier said: Can be done. The molding process is up-to you, but the supplier of the molding machine certainly has a list of predefined temp/pressure profiles, at least for the later generations of machines out there. As ABS is all over the world since so many decades, companies have assembled good-sized databases of what mix does behave in what way when molded in a specific process. Making such a molding machine is one thing, knowing what it does with different starting materials is another. There is soft ABS, hard ABS, more dense ABS and so on and on; enclosures are made of ABS (many parts of my beautiful IBM XT from 1985, which has not yellowed a tiny bit; also my Atari 1040, which has yellowed to the extent that you could say it was spray-painted), and myriads of other things. And bricks, of course. You know all that, I know, I just want to put things into perspective. I have read that TLG page you referenced a thousand times, I believe. Starting next winter, the chemistry department at my university offers a B.Sc. course "Sustainable Chemistry". In preparation for that, I was looking around quite a bit, and my beloved hobby offers a lot of things to discuss with regard to "sustainable". Well, ABS and all the other polymers TLG lists on that page are persistent (non-bio-degradable) and simply accumulate. It's the whole purpose of using them, they talk about that on the very same page: "... to endure decades of play and be passed down from generation to generation." It is what we want. The only chance they (and competitors) have with regard to sustainability is energy/CO2 saving when making the pieces - if they, or better we, do not want to give up that they endure decades of play. You are absolutely right: ABS is only one of the highly persistent polymers, TLG and competitors are using. For ABS, they list bricks; I could be wrong, but Technic beams looks very much like being made of ABS as well, among other Technic pieces. Axles would be dumb; the torque, TLG themselves will stress these with during QA, will tell the ABS supplier: No sorry, there is no ABS formulation that can withstand that torque. Well, so on to POM: TLG says: " hard and stiff material that is also flexible and strong". Hmmm. Hard and stiff = strong, but not that flexible. Tested it: Upon bending a 10L Technic axle, not much flexibility is seen (perpendicular to its main axis) before it disintegrates. OK, yes, I know, it is a page, marketing has made. And that is OK. POM though is another non-biodegradable ... etc. etc. Of course not. We don't want to look at rotting LEGO. And all the other polymers are not. They are made from different monomer formulations to match with the desired characteristic (highly flexible = PE, transparent = MABS, polycarbonate works as well, but is not listed under transparent but hinges). And finally the yellowing. Man, so many articles delivering so many hypotheses ... I suggest though to drop the flame retardant argument, at least for bricks made after - let's make it 2000. The yellowing was presumably attributed to release of bromine, a brownish liquid at room temperature. That sounded good, but since 2000 "all" these bromine containing flame retardants (biphenyls and the like) suitable for usage in polymers have been phased out. The odor of bromine is, as the greek name bromos says, well, stinky. I bet you could smell bromine if a plastic piece yellows due to release of bromine. I guess the literature has settled on breaking remaining double bonds present (not in the aromatic ring systems) in ABS, which are then not easily saturated again (it is a solid material). The then "free" electrons - they are not free but delocalized - shift the absorption of light towards higher wavelengths = towards the red. And all that in the upper molecular layers of the polymer. In fact, when you ablate the upper layers (scratching is not good, it mechanically changes too many things) of yellowed ABS with nanosecond UV pulse laser radiation, the irradiated surface becomes swiftly white again, as the upper layers are non-thermally removed. Yes, did that in the lab, but it just proves that there was "something" changing in the upper molecular layers, which we already know. This also goes along very well with H2O2 mediated whitening, some call retro-brighting: H2O2 in aqueous solution (the higher %age the better) + UV light generates, among other things, OH radicals. These love electrons, delocalized or not, and thus saturate the formerly broken double bond by addition. Which removes the red shift and thus becomes "white" = non-absorbing again. However, only in the very upper molecular layers - if not the uppermost - since water or water/H2O2 solutions can't penetrate deeply into a polymer (on the timescale of treatment!). Maybe two or three layers, but that's it. Which also leads to the conclusion that it can't be bromine causing the yellowing, you would smell it, as it would sit in the upper molecular layer. Well, that is what I arrived at, when plowing the literature for quite some time and doing some (non-conclusive) experiments. Who knows, it may be all garbage - maybe others have other information/references. I would love to hear of! All the best, Thorsten