keybrickone

We're currently busy setting up the shop on our website to make at least the "first served" units available for sale (should be ready in the next few days). Our conclusion from the Kickstarter campaign is, that there is genuine interest in the product, even if there's much room for niche needs that we won't be able to fulfill. Price is definitely an issue, but we're in a difficult spot with the quantities involved currently. Our main cost drivers at this stage at: regulations (standards confirmity, environmental and recycling), the case itself (currently accounting for about 50% of the BOM - and we have little wiggle-room for cheaper supply there as we use a Polyjet process for finish, material and durability reasons - not your typical home FDM print definitely) and the custom battery pack. The 2-3 Keybricks will be available in the comming days as I said. This again boils down to cost. Building in a charger solution that is capable of dealing with the noisy power source that is pickup wheels requires far more attention than just "swapping out the chip with one that can handle up to 12V". The rectification and filtering necessary to make that power usable to such a finicky sink like a LiPo battery without damaging it would not even fit onto the available board space and further increase the cost for a feature that only a part of the customers is likely to use. I like the idea of having a possiblity to do so, but I guess it'll be more economical to have a "addon brick" that takes the suppy from the rails and offers a nice and clean USB output. That's more versatile (could be used with anything USB basically) and a LOT cheaper for anyone not needing the feature. As for the USB-C topic: as I already mentioned numerous times, implementing USB type C is much more involved than just swapping out the connector. Implementing just a 5V compatible input stage would cost about 15 times what the current micro USB solutions cost (in terms of BOM cost). Implementing an input that can deal with the USB-C-PD for up to 12V stuff would skyrocket the costs. Implementing a non-compliant solution is not an option because we don't want to take responsibility for fried electronics of our customers. As for cost, I agree that the board itself is not terribly expensive, but the engineering (both electrical and software) and all the required changes (at least: connector, USB CC-line controller, V-Bus disconnect mosfet, a lot of passives, eventually another charger design if higher input voltages are a thing as well as probably the next bigger MCU in the lineup because the software stack won't fit in the reamining flash space) are. Now if anyone here is willing to put down anywhere from 8 to 14€ more to get that connector, tell me. I have my doubts. For anyone concerned with the diying-out of the connectors; you can technically buy "dumb" Type-C to MicroUSB adapters online. These are definitely non-compliant but will work with Keybrick as long as you choose an equally dumb power source. We won't be shipping such a device though (for aforementionned reasons) The main goal of the campaign was twofolds: see if the product can generate some traction at least and get feedback. Realistically there were two possible outcomes: either fail with a clear marging (which happened) or blast past the goal if a miracle happened. The 10K are definitely not enough to scale it up to a point where we get some comfortable wiggle room on the price (at least 5x the sum for that). We currently plan on doing what you suggested in your last sentence: slowly ramping it up if we can make the enthousiast happy. What we clearly heard was the interest in a solution for the Technic hub, so that's something we'll be looking into soon (including the Type C this time as there's a lot more room - in every sense - to implement it properly).

I have to agree that the formulation my teammate chose at that point was below par. In fairness though you are leaving out some context in the quote, which was a response to someone (rightfully) calling out the slow progress we made about 10 days into the campaign. Now as to our understanding of whose job marketing is, I think you're in a difficult place to exactly judge what went on behind the scenes. I'm not complaining and definitely don't want to sound whiny, but let's simply say that we faced more scepticism and gatekeepers than we anticipated, plus shipping out the review units was delayed significantly (they were shipped about 18 days late vs our initial plan) because of a manufacturing issue with the cases (which again, can be seen as our fault, launching the campaign earlier than we should have perhaps). A significant chunk of our advertising buget went in this direction with little return sadly. Seems to be that only a fraction of kids play enough for the battery thing to be a real issue (or people don't bother, idk) and the price probably exacerbates the point, making Keybrick a niche product (which was expected). As to why the kids are in the video: it's simply how the idea was sparked and we felt it was an adequate story to tell. Thanks for being rude. I'm not quite sure if, but I'm assuming that you already consider Keybrick in it's current state to be at least slightly overpriced for this part, if I'm wrong, please skip my following statement. It's not only about what is technically doable, but about economical viability. Swaping out that connector involves so much more than just exchanging a single part, especially if the customer - rightfully so - expects an actually electrically sound and standard-compliant device. Including changes is assembly cost (more parts, worse yield because of finer pitch connector and additional parts), accessories (type c cables cost roughtly two to three times as much) and the required engineering we'd have to increase the price of the device to a point very close to three figures. To me this change would be plain terrible value. Where did I state that I'll be selling prototypes? All units that will ultimately be shipped will be as final versions as a small volume product can be. That they will most probably come with 3D printed case (not your typical FDM print having that said) is the only difference to what we wanted to do with an injection mould initially. On the other hand about half of the backers on Kickstarter explicitely chose to pledge on a reward that would have had a 3D printed case in any outcome for example, which makes me think that this might not be that big of an issue for the niche we are in. If your comment was about me stating that we'll use the battery cells we already have on stock: these are not different from what we would have ordered if we needed thousands of them at once. If you're unfamilliar with how supplier parts are evaluated: you typically buy a sample big enough to be able to make statistically relevant testing on them, in this case we bought 60 cells from that supplier (not the only one we tested) and used about a third of them in our testing. The remaining 40-odd cells have been patiently waiting untouched since. These 40 packs are the ones that we would have shipped as "First Served" units in our Kickstarter - as we openly stated. This is simply us being honnest. The Kickstarter campaign was a market fit test first and foremost, even more than the chance to generate pre-orders. At the point in time I wrote the above sentence we were stuck at about 25% funding which was way below the point at which we would have considered it worth bothering with. Creating, maintaining and marketing a product requires considerable effort and as a small business we can't afford doing so for a product that seemed to gain little traction even with enthousiast. Now things have changed considerable in the last few days, with genuine interest and/or confidence sparked by the reviews that start showing up, as well as genuinely encouraging discussions with some resellers. We have to figure out a way to make the whole deal viable for them as well, which will be a challenge on its own, but we're definitely seeing things comming together (albeit in a slightly unexpected way). Thanks for posting, didn't manage to do so yesterday.

Yeah, kind-of what I thought would be possible. A female USB port for the output would require the end-format to be ~3 studs wide and probable 4 studs long. Height will probably be like 1.5 standard bricks (just because of the usb port basically). The "9V electric plate" you're speaking about is the "Power-Functions" connector right? Thank you. Any help is much appreciated.

Thanks for the warm word. That one was the starter for us :) good to see that we're not the only ones really annoyed by this. Not as-is. The charger IC we're using has an operating range of 4.5V to 6V (it does withstand up to 10V on its input terminal without getting damaged, but stops charging once it reaches its over voltage protection threshold for safety reasons). So this would require an additional step down conversion from the input voltage to ~5V in addtion to the required bridge rectification to account for the changing polarity of the supply. None of this is particularly difficult or expensive to implement, but it wouldn't really fit into the available board space currently. To be honnest, this is the sort of add-on we'd really like to offer if Keybrick comes to be at a sustainable scale! Noted it and will eventually look into the topic! Thanks

Hi TuffTuffTuff, we got a bit further on this issue I think even if we are a bit late to some of the "parties". The biggest drawback of being a tiny company: so much to do in a limited amount of time. On the good news side: there are a handful of review units that have reached their destination this week. So reviews should start popping-in soon! On another note: we've been asked by a few folks if we could provide some externally accessible power tap on Keybrick, so one could power an Arduino or ESP32 board from it for example. As Keybrick doesn't have a suitable internal voltage rail we decided that it would be easier, cheaper (and usaable for everyone even without a Keybrick) to tap into the existing "VCC" rail on the Powered Up connector: The "Keybrick Powered Up Breakout-board" uses an IDC connector to cut through the insulation of the original cable. All six lines get exposed for you to hack on a standard 0.1" pin header. I designed as small installation helper by making the actual board just a break-apart piece of the whole product, so lining up the cable with the IDC connector in the right orientation becomes less of a gamble :) the whole thing can be removed again without much damage to the cable. Thanks to the offset-cuts there's no risk that neighboring conductors short out, so a bit of tape to strengthen the cable again should be enough! Other than than we validated the "in operations charging" feature, so that's happening. And the latest revision of the case fixed the last issues we had on our laundry list on that side (better fit to the hub, an additional notch to suppress some side-wall flex on the battery cover and improved tolerances which makes Keybrick now as sturdy as we feel it should be). All the best, - Yannic

Thats not really the situation as is. In order to have the same load in a train (or other vehicle/ set) you would need to have a contiuous pretty steep incline to climb. At steady speed the load settles to under half that value with small increases of load in turns. This is pretty much what you see in the "real world oval test". And the boost mode is really kind of extreme with reproducible derailung at 7/10 speed settings on the smaller radius turns. It's cool for high speed straights and hill climbing because you need a lot less momentum to get going. Really the bench test values are just an extreme situation test which allows me to run a fairly quick comparison.

About 1h30-40 which is the normal time for a full charge from fully depleted. As the hub draws roughly 320-350mA at that load wirh Keybrick in boost mode it's the expected performance. Keybrick delivers 3.3W of power for 1h17 which cumulates to ~4.1W or 85-90% of the cell capacity.

So I have some interesting news. While I wasn't able to validate my results with our external EMC testing partner just yet, it looks like we should be able to enable "in operations charging" with just one very tiny hardware change to get rid of a nasty conducted EM spur which doesn't even require us to modify the existing board because it just requires one additional small value capacitor in a previously unpopulated placement as a RF shunt. This means that we'll just have to do a minor change in the firmware to enable the feature! A massive thanks to everyone who asked for this! I did simply discard that possibility based on my measurements on just the hub alone, but apparently Keybrick did a good job of killing most emissions as is and the single remaining spur just went away with that additional capacitance! Awesome! In addition to that some other news of the week: Battery life testing re-done by request of a few folks on Facebook we re-did some of the battery life testing and here are the results for our two setups: On the "lab test-bench" (test setup described in a post above) in a high load scenario: Duracell Plus alkalines: 2h03min Eneloops (white): 1h16min Keybrick ECO: 2h07min Keybrick Normal: 1h37min Keybrick Boost: 1h17min On the second test setup (which is bascially the Oval out of the 60197 set with two wagons attached to the loco, running at speed 5/10): Alkalines: 3h27 Keybrick (in Normal mode): 2h34 Eneloops (white): 2h22 Till didn't manage to cobble together a loop counter but he did time the loop duration at 3 points: Test subject: Loop times at 5 min after start / after 1h15 / 5 minutes before stopping Alkalines: 5.31s / 5.83s / 8.13s Keybrick: 5.47s / 5.74s / 5.68s Eneloops: 5.77s / 6.29s / 7.08s Note: we measured 5 minutes after start to ignore the sharp drop in loop time which seems to be caused by the motors heating up. Inductive charging and "remote USB port" Great ideas, we would love to look into more deeply. The remote port is relatively simple to implement, the inductive charging much less-so (technically doable still). But at this point we need some more support for Keybrick One to be able do so. Review units on their way We are super pleased to announce that a few review units are already on their way to very valued members of the community. We're super eager to see what they think about Keybrick Getting the word out ...proved to be a bit more challenging than expected. At this point we really want to thank everybody here for being open for discussion and as helpful as you all were! We've got some media coverage in the pipeline luckily Have a nice weekend, - Yannic / Team Keybrick

Thanks Currently not enabled for EMC reasons. Basically the 88009 hub doesn't do a stellar job at keeping brush noise from the motors away from the battery pack (at higher speeds that is). As this question has been raised a few times I'll be looking into possibilities to make this happen nonetheless. Maybe it could be possible if we force the user into ECO mode while the charger is attached. That way the maximum power, and by extension motor speed, will be limited which might reduce the noise/EMC issue to a level we're both comfortable and still compliant with :) I have one sitting in my lab, bought it to evaluate if we could make something similar. We didn't start anything but its a very real possibility.

Sadly no, we're using micro-USB, for various reasons. USB type C connectors are considerably larger (at that scale anyway) and would have forced us to place the connector on the long side of the hub, which again decided against because it seemed less practical. Additionally the conectors are a good bit taller too, which would require us to use a mid-mount type because we would run out of height. Finally, implementing USB type C would require a good bit more circuitry inside of Keybrick (because the device is actually supposed to communicate with the charger in order for it to be properly implemented) which comes at significant expense and with the matching board space requirement.

On a different topic: theres a discussion about our testing going on in one of the Facebook groups. The main concerns are about the contenders we chose and our testing methodology. We tested: Brand alkalines (those with the "copper top", which we found to be only bettery by about 3% as when compared to "no-name" cells from our local discounter market in this application) LIDL NiMH batteries (which many consider a good cheap option) agains Keybrick in Normal mode The setup we used for the number on the website was basically the City Passenger Train 60197: loco plus two wagons on the supplied oval. Setting it to maximum speed and leaving it until it stops. One criticism we are addressing in the next few days is the choice of rechargeable batteries where we are going to test against Eneloops today. Another test I have been running in the lab is using two PUP motors directly coupled and connecting the second one to a four quadrant power supply (basically a power supply capable of sourcing and sinking energy) which allows me to simulate many load situations, while measuring power. This way I am able to simulate a variety of load conditions from freewheeling to brack-driving the "generator" to simulate stalling. We didn't publish any numbers from that testing because we felt that it was too synthetic to actually mean anything to most people out there. The other aspect we find hard to emphasis properly is the one about the discharge curves: with alkalines (and to a lesser extent with the NiMHs) you can clearly see that the train gets significantly slower as the batteries deplete, to a point it's really crawling to a stop (which is actually a significant percentage of the time in the oval loop benchmark, and a significant part of charge where our Kids would complain that the train is getting too slow). Comparing that to the constant power output of Keybrick where you basically have the same power at your disposal until it finally stops is somewhat difficult, because even if the absolute time is a fair bit shorter (alkalines: 3h36 vs 2h34 with Keybrick) it is keeping its pace the full duration. What we weres asked whas: provide some numbers that show the evolution of speed (first ten rounds vs. the last ten on the different power sources) show the difference between the power modes of Keybrick how the different power modes impact speed / power / battery life Is there anything you would like to see over that? If you have any suggestion how we could make the battery life comparable I'd be super thankful :)

Pretty much all there is to say, the best: you don't have to take my word for it either: https://www.wirelesspowerconsortium.com/blog/qi-wireless-charging-and-cardiovascular-electronic-devices The FCC found that even within two centimeters of such a device would be safe as when compared to the norms the ICDs and pacemakers must comply with regarding electromagnetic immunity. On the other hand this is really a very theoretical discussion at this point, because we don't have any inbuilt wireless charging feature currently.

I'll look into options. The main challenges will probably be getting the coil alignment approximately right, and the distance between the emitter and receiver sides as small as possible. I guess the hardest thing will be to get it to look pretty in the end ? Yes. Plus making it usable for other 3rd party solutions would be the best solution for the community too.

Wireless charging may be something we could look into indeed. It's a bit tricky to implement because of geometrical reasons at least if the charge pad should be placed beneath the train / "in" the rails. At least the rail spacing are a problem if one wanted to use off-the shelf standards as Qi (which requires ~55mm coil diameter to transmit a reasonable amount of energy). If something on the side or atop the train would be an acceptable solution then this might actually be both simple and reasonably cheap to implement. What exactly are you interested in? The cells are high quality pouch cells designed specifically for high-drain applications (its a 1S2P pack, and the cells are rated at 20C repeated discharge rate 30C peak, meanining the pack would sustain 1200mA * 20 = 24A discharge rate. We tested up to 14A - we don't have lab equipment that would allow us to do more and the cells were just fine after over 30 cycles of rapid discharge and charging at 1A. As they are not going to see more than 7.5A peak load inside Keybrick the battery is going to have a pretty relaxed life, which will be great for both safety and longevity) We are using a switching mode charger IC which we set to 800mA constant current rate and 4.2V charge end voltage (which is below the specified normal charge rate of 1200mA for this pack and the manufacturer technically allows for up to 4.25V charge voltage) The microcontroller in Keybrick continually monitors the output current and cell voltage. It starts indicating "empty" batteries at 3.1V (which is 30% state of charge) and cuts the output at 2.75V (which is 17% SoC). Even if this cutout was to fail for some reason there is an in-built pack protection IC which will deal with error conditions such as short-circuit, over-current, over-charge, over-discharge. The output is over-load protected by a PTC fuse (self resetting) which will cut output supply after a few seconds in case of serious overload (~2.2A for more than 15s) as when someone was to short the output terminals The switch mode converters have built-in over temperature protection, but that should never come to action unless something shorts out internally (see PTC) We tested the design a LOT. Even sustained continuous maximum load (1.6A@9V in boost mode) doesn't harm the electronics taking them at ~77°C in the enclosure with ambient at 30°C. And that's is a pretty unlikely scenario as the Hub will cut supply if it detects such load condition from the motor side. The maximum sustained load we were able to create was about 1.1A which was with two motors attached to the hub running short of their stalling point. I think I said it all, but you're welcome to ask more details :)

That's something we can definitely investigate, great idea! The point with the flat flex usb cable was mainly about having a very small space taken up by the actual connector. The 90° connector in the picture above would actually just need a mere 1x3 studs. We could probably even build it a bit differently (in a custom design) to have it flush with the side of the hub... I'll make a few drawings tomorrow of what should be possible ?

It would be possible to manufacture the ends ourselves, but I'd rather try sourcing something good "off the shelf". Regarding the location of the USB port, the issue is not only having the variants in the case (3d printed or not) but the power/charger section of PCB would need to be redesigned too, as the USB port has to be physically close to it. A verstical mount option is pretty much impossible (if we want stay with a standard plug, because of the limited verdical space (3.2mm from pcb to case plus 1.1mm wall thickness, and the shortest vertical mount port I know of is about 4.9mm long). So that would stick out and make the mating surface unusable in that spot... I guess we could resort to building the cables ourselves in the worst case ? Thanks for the input ?

Thank you for the kind word zephy1934 the flex cable is definitely on the list of things we *are* looking into, it's just difficult to find a reliable supplier for them. The three we tested up now had all their very own set of quirks... Regarding the international shipping: it's generally not a problem to ship the batteries as long as they are built into equipment (which the cells technically are in Keybrick), it's just relatively expensive because we can't use classic international mail but are forced to express services. For the US we will probably resort to Fedex and/or UPS depending on the location. I've been shipping batteries etc. for the last three years so this isn't anything new for me (I'm running an hobby electronics shop [hhtronik] and am selling another consumer electronics device [called Luccia Light]).

Thank you for the trust and support That design had some pretty big downsides concerning manufacturability, which was really one of the top priorities for the current version. But we actually may have a solution on hands that could fit your needs (it's on our laundry list of possible stretch goals / addons): we have some very low profile 90° flat flex cable USB connector options such as this one:

Yeah, we went into this adventure because we were really sick of buying new batteries (actually Till, I didn't have any Powered Up stuff back then). Plus they managed to make changing the cell really annoying IMO. The screws might be a good idea, but the clips of the base are away too weak, at least on my two hubs the contact springs always hold the battery carriage so tight that I have to take it out in a second step... Any other performance figure of interest? Playtime is close to alkalines as I said. We have a loop test video in prep where we put the Keybrick up agains high quality (and thus expensive) alkalines as well as NiMH rechargeables and we got following results: - Keybrick in normal mode: 2h34 - Alkalines: 3h:36 (with the last 40minutes being a crawl to death where Till's son would start complaining) - NiMH: 2h02 (with no power from the start and the last 20minutes being the same annoying crawl to a stop...) We tested this on a relatively small oval at equal start speeds. As Keybrick doesn't slow down the pure time based comparison is a bit against up, because the hub will use available power. Running the same test using the "eco" mode (6.3V) we managed over 4h in the same test. Got no footage of that though.

Thanks for the info about hotlinking. Will edit my post accordingly! The output voltage is selectable between ~6.3V (lower would make the hub think the batteries are depleted) and ~9.25V (which is an arbitrary upper limit we determined will not cause the hub to heat up too much under load, the motors are pretty happy with more). The reason Keybrick generates a lot more torque is that we're using an regulated DC/DC (step up) converter in there which regulates the voltage to whatever was set for pretty much every load situation the hub will present (we have got a comfortable head room). This means the voltage isn't dropping under load as with the AAA cells. The battery pack is a 1S2P pack with 1200mAh nominal. As these are designed to deliver very high current (specced at over 30A) they do deliver quite a bit more in our case, with our testing showing consistently arround 1400mAh when discharged at 5A constant current to the discharge end voltage of Keybrick (2.9V, again with a comfortable margin the the specified 2.5V that would be acceptable from the cell manufacturers side). We have three cells undergoing invervall cycle test since august now, with each cumulating over 50 cycles now showing little to no wear in terms of net capacity (nominally down 20mAh average, still well above the specced 1200mAh). We did the DC/DC conversion to avoid having a voltage drop over the discharge curve as the regulator keeps that "invisible" to the hub. Step up was more practical for a multitude of reasons. About...?

Hello everyone, first of all, I'd like to thank @Jim for giving his blessing for me to write about our new project / Kickstarter campaign. So, when speaking about "us" I really mean Till and myself (Yannic) who spent the better part of the last 10 months working on Keybrick One. It's a rechargeable battery pack for the Powered Up Hub 88009 we started working on as the battery waste (and fiddly replacement process for said batteries) really bothered us a lot.After a first round of prototypes which looked like that: Over the course of a few months we evolved the design quite a bit for keybrick to become a replacement of the full lower shell of the Hub. It's become quite good in our opinion with being both more powerful than using alkaline cells (like drift-mode more powerful :)) and having a way longer play time than when using rechargeables in the hub (actually comparable to alkalines). Now we're launching Keybrick as a Kickstarter project because we think its a great addition to what TLC did with the Powered Up series, which really lack that rechargeable option in our opinion. The Campaign is here: Keybrick One on Kickstarter We'd like to use this to open the discussion with everyone intereste in the Powered Up series and are really eager to see what questions you all will come up with. All the best, - Yannic