Giottist

Saturday afternoon, its raining nonstop outside and I'm a little bit bored. Perhaps the best situation to compare NiMH with Li-Ion technology. 1) Let me introduce my charger. It is designed for model operation at open sites and needs just a 12V power supply, i.e. a car battery. I connect it to a well dimensioned power supply. The charger can charge Pb, NiCd, NiMh ald Lithium cells. The cells can be also discharged with constant current. A microcontroller controls the charging and discharging process. Actual state, current, voltage and capacity is shown at any time on the display. I'm using the charger for several years and can verify absolut reliable function and measurement precision. 2) Let's have a look at the energy density of NiMH and Li-Ion cells. The Wikipedia gives an energy density of 180 Wh/kg for typical Li-Ion cells and about 60 Wh/kg for NiMH cells. Now lets have a look to our suitable cells / battery for the PU hub: The six AAA NiMH cells seems a little bit larger than the Li-Ion block. The six NiMH cells weight together 72.4g, the Li-Ion block only 25.6g. Please keep in mind that the weight contains the housing and the contacts, internal electrodes and so on. Since the NiMH battery comes as six individual cells the weight factor for them is inferior to the compact Li-Ion block. And now let's measure the capacity. To do this I programmed five cycles of discharging with 200 mA down to 6.6V and recharging with 100 mA. The result is for the Li-Ion block as mentioned above 470 mAh with 7.4V = 3,5 Wh. and for the six cell NiMH battery 350 mAh with 7.2V = 2.5 Wh. Let's check the charger results with the energy density values from literature. If all the mass of the Li-Ion block would be available for energy storage the capacity would be 4.6 Wh. with the real measured 3.5 Wh we get a mass factor of 76%, a realistic value since the block is really compact. The six NiMH cells should store 4.3 Wh if all the mass would be available but six times container, electrodes and contacts claim their tribute. 2.5 measured Wh means a weight ratio of 58%, a realistic value too. The used NiMH cells are rather new spare parts for my cordless phone. At the last three cycles there were no significant change in capacity of roundabout 350 mAh. The print claims a capacity of a whopping 1000 mAh which is completly impossible if the energy density of 60 Wh/kg for NiMH tecnology is realistic. To realize theese values the energy density must go up to Li-Ion levels -> impossible for the NiMH chemistry. There might be some improvements but not of a factor of three. I am sorry, but I have no reason to alter my claim that capacity values printed on the cells belong into the cloud-cuckoo-land. And now let's enjoy our running LEGO trains for some relaxed hours!

Yeeeeeeeessss...No, not at all. In the past I tried a lot of different NiMH cells and have had reliable lousy results. It seems unavoidable to use chargers with single cell supervision and LSD cells. Up to now I have my doubts, if the capacity value printed on the cell is realistic. My problem is the additional costs for a technology which I do not want anymore to check it out. Caparezza recommends a suitable combination of theTechnoline BC700 charger together with eneloop (pro) cells, wgich are definitely LSD cells. At ebay I've found several dealers who offers all the stuff for about 70,- € (charger and 12 cells). Uh oh, expensive just to check out a claim. But perhaps that's not neccessary if somebody lets its train motor idle together with the lights from full charge down to orange hub blinking. Any volunteer here? BTW: Li-Ion battery fires are possible but extremly rare. In most case such is reported just to encourage us a little. A leaking NiHM cell pouring auround its potassium hydroxide contents is not a matter interesting enough to be reported. Since there are billions (american notation 1 billion = 10^9) Li-Ion batteries out there, I see no safety issue.

Hey Capparezza, Thanks a lot for your info! You're using so called LSD cells (Low Self Discharging, means new technology). If we collect all available information we can do a lot for the LEGO train community PS.: Just found technical information about the Technoline BC700 charger. It charges and supervise each cell individually. And the eneloop cells are new generation LSD cells. This explains your sucess! Let everybody know this, if Li-Ion technology seems suspicious ...

Perhaps this video by Bananenbuurman can help.

This morning I started the attempt to get a set panasonic eneloop AAA cells and a matching charger to reproduce the experiences of toastie and others. I has no sucess, the only offers I found were doubtful noname chargers and cells with pantasie brands - the same stuff I've destroyed multiple times in the past. I have no motivation to burn more money. Then I looked up original cells with original chargers in different online stores - an expensive pleasure. For the same money for one set for a LEGO hub I can buy another 10 (!) Li-Ion blocks. Ok, without charger, but I have a reliable Li-Ion charger. So please understand me that I cancel my test plans for NiMH solution for economic reasons and continue to recommend the usage of 9V block Li-Ion batteries if a suitable adpater is available. The advantage are: Li-Ion technology is reliable, reproducable, offers sufficient capacity and run time, there is no risk to buy a wrong combination of charger and cells and the cells are by far more economic. I for myself am satisfied with a real train runtime between 2 and 3 hours. The disadvantage is at the moment that not ready made adpter is available. Now I like to look around somebody with a 3D printer to manufacture a prototype for a foolproof adapter. On the other hand: If you are lucky to have a running and reliable NiMH combination of charger and cells, please use and enjoy it. There is no need to change anything. Using NiMH AAAcells is more expensive but needs no modification. That's true. 8 hours later: Addenum: I've found a very detailed website about NiMH cells in german language. I try to translate the most important statements into english. (Please forgive me my bad language ) Now it is explainable why Toastie and others have good experiences with NiMH cells and others including me not. 1) To have a reliable and functional power source you have to use NiMH cells with low self draining technology. The website mentions eneloop and ReCyco+. The manufacturers are Panasonic (Sanyo) and GP. 2) Please let any noname cell at the store. Ordinary NiMH cells have a self discharging rate up to 50% a month (!) and have to be recharged very often, best every week. If you do not so the risk of deep discharging is high. 3) Deep discharging below 0.85 V will destroy the common NiMH cell very fast. Low self discharging cells are less sensitive but still a little endagered. It is neccessary to use NiMH cells often. This is the best for them. Playing only in long intervals will destroy even low dicharging NiMH cells 4) The website recommends to use a dedicated charger with control for each individual cell, since the parameter variation for NiMH cells is rather high. This sounds somewhat expensive. Many thanks to Chr. Caspari from www.elektronikinfo.de

Thank you all for the discussion. It helps to find out the properties of different solutions for rechargeable batteries. @toastie: I did not plan to make real science out of this question but have an overlook since I have not the best experiences with (noname ?) NiMH stuff. I'm using the PUP train motor 88011 as the motor idling in bottom up position and the lights 88005. I know it's rather difficult to measure the average current, voltage and power for a PWM device. For a first crude look I've chosen this simple setup. At the moment I'm very busy by profession and have no opportunity to get original Panasonic Eneloop stuff, but you make me curious. It's a nice idea to make a comparison.

If I unstand it correct, we can write small c# code which is executed as a result of different sensor messages. Does the C# interpreter (?) allow loops like for oder do while and branches like if else or case and local variables?

Ouh, better not. I have no plans to modify the original hardware, so I did not try or even thought about it. The only exception was the experiment with the two AAA Li-Ion cells which I count as a flop. I can remember a pcb image in the tear down thread on the first page. It seems quite larger than a 9V block.

Another info update: I warned in another thread not to use the original LEGO app if you're using android 5.0.2 or older since the original app is not only bloated but also contains a lot of bugs. With Cosmik42s train software there is no need for this app anymore. With one exception: To update the firmware on the train hubs. Fortunately this works still with my old smartphone with android 5.0.2. Three of my four hubs were updated (to what version ?), the fourth has a recent firmware version. Immediatly I threw the LEGO bloatware out of my smartphone after the firmware were updated. Some train hubs are recognized now as "Hub No 4.", the others as "Smart Hub" but function is not affected. The BOOST hub is recognized allways as "Move Hub". All hubs cooperates with the software without any flaw.

Thank you all for the hint with original eneloop cells in combination with an original panasonic charger. It's worth to test them. May I ask anybody to let a train motor idling together with the lights from full charge to an orange blinking hub and tell us the time? Otherwise I spent another 30,- € to test them. I'm curious now. I can faint remember I've use eneloop batteries in the distant past but no original panasonic charger ... Hmm. BTW, at the moment I'm very content with my Li-Ion solution.

Please do not believe what's printed on the cell. This has nothing to do with reality. I'm using a charger with a lot of rather precise measurements i.e. the real capacity. I've found the AAA Li-Ion cells has less than 200 mAh in reality than 350 mAh printed on. I have severe doubts that your 950 mAh are real. The half at maximum is more realistic. More important is the fact that NiMH cells are difficult to charge. I have tried a couple of chargers (to be honest more than five ...), but each destroyed the cells after 3 to 5 cycles by overcharging because charge end detection is very difficult and not solved by most chargers. Perhaps you are the lucky owner of a working one. Can you tell us which kind of charger you use? I for myself banned all NiMh stuff: My Li-Ion solution offers 460 to 480 REAL mAh (measured and confirmed multiple, the 650 mAh printed on the battery is pure marketing) and runs a train with realistic load for more than two hours. And charging is reliable, the life span will be hundreds of cycles. How long your trains runs with full batteries?

Hmm, Cosmik42s software makes it possible to use standard LEGO hardware with similar functionality. For one SBrick I can get three train hubs (at the moment at BrickLink).

I think at the people from TrixBrix ... I hope both read this discussion.

Hi LEGO train folks, I've found a preliminary soution for replacing environment polluting AAA cells with a Li-Ion rechargable batterie. Using one way AAA cells is not a pleasant game since you spend a lot of money and produce a heap of problematic waste just to run your train for a couple of hours. What's about rechargeable batteries? Original hub with six expendable AAA cells Yes, there are two solutions. First is to replace the AAA cells with Li-Ion cells of the same form factor, and yes, they exist. I've found a supplier on ebay who sold me the cells for little money. Two needed cells and how they are put into the battery holder These AAA Li-Ion cells have the same form factor than common AAA batteries but nearly triple voltage. You must not put six cells into the holder, otherwise you destroy the hub immediately. Just put two cells in it and solder a small wire between the firs and the last contact in the holder. The VCC contact for the first cell is connected to the GND connector of the last cell by a soldered wire. It is easy to solder it and needs so little time that the holder plastic has no opportunity to melt. Some experience with soldering is a good idea ;) I've tested it with my train motor and the lights: It runs for just an hour, recharging told me the capacity is abot 200 mAh and not the 350 mAh printed on the cells. Do not believe whats printed on the batteries, this is pure marketing. Hmm, not the best result but no surprise since the hub is filled mostly with air instead with battery cells. There is a much better solution. There are Li-Ion batteries with the form factor of a common 9V block available. I've found them on ebay too. Look for "6F22", "9V block" and "Li-Ion". Average price is abot 6 € if you buy one, 5 € if you buy two or four and down to 4 € if you buy 10 batteries. The nominal capacity is 650 mAh but this is overoptimistic, real capacity ist abot 460 mAh measured by my own. Not bad at all. The big advantage is it fits nealy perfect into the train hub. Filling the hub with batteries instead of air is promising. Another advantage of the form factor is the connector clip which avoids fatal polarity errors! The 9V block form needs only an adapter to the hub contacts to make the whole nearly fool proof. I've found here on Eurobricks other "solutions" with an adapter made of cardboard and some selfadhesive copper band. This is the best way to destroy the hub by wrong polarity. Using a common connector for 9V block batteries avoid this. The common connecto clip for 9V blocks The other side ... Please look for the postion of the contact strips ... and the mounted adapter .. The adapter it self is very simple and consists just of a matching piece from an expired "what-so-ever"-card (in this case an expired bank card, which has to be destroyed to prevent misuse) On one side two strips of self adhesive copper band are put on the plastic sheet just at the positions for good contact with the leads of the LEGO hub (see the middle image above). The clip is soldered in this way that it find its final position on the other side of the plastic sheet. The red wire connects to VCC (the contact in the center of the hub), the black to GND. Soldering is easy and quick since the copper strips have nearly no heat capacity. Some soldering experience is still usefull. Several extensive tests are very promising: I've run repeatedly a train motor idling together with the lights for 3 hours 45 minutes without a break. Recharging told me the capacity of theese Li-Ion blocks is about 460 mAh and the average Voltage is 7.4 V as expected for two cells in the block. The measured values let expect at least two hours of operation with a mechanical load, not less than expendable AAA cells. Not bad, isn't it? The measured capacity of the 9V blocks do not match to the run time in comparison to the AAA Li-Ion cells, but I do not trust the AAA cells. 3h45' could be peproduced several times and this is the only important value. Next step will be a 3D printed form piece to replace the cutted card. This 3D adapter should contain the connector for the 9V block, the metal strips to connectz the hub and a notch to fit into the hub to prevent exchanging the polarity. Who is able to design this device? I'm sure there is a large demand ... Hope I can help you with this solution ... There is potential.

Hey Cosmik42, please have a break and a coffee and enjoy your sunday afternoon At the moment I have to wait until my long run test of a solution for rechargable Li-Ion batteries for the train hub will come to an end (It runs since 3.75 hours and the voltage display shows still 64%...) And wow: i discovered you're using the WeDo 2.0 distance sensor which is available yet in contrast to the color/distance sensor which can be obtained only by a complete BOOST set or for an incredible charge on ebay or bricklink. That can be very useful for signals ... Minutes later: Ah, the battery is empty. It took 3h 47' to drain it by a idling train motor and the lights. I will publish the results in an extra thread ... A quarter hour later and I can confirm: Yes now everything works as planned. Wow, what a program!

Thank you a lot but take up time. For the color/distance sensor I have found the following behaviour on the BOOST hub: If the sensor is connected to port D nothing happens. If connected to port C it delivers a lot of correct messages in the debug frame. Absolutely sufficient for development purposes. If a motor is connected to port D the sensor message in the hub box is updated to an erratic value when the state of the neighbor motor is changed. This can be confusing, but the debug display is correct. In the new version 0.3 the color/distance sensor is detected by the train hub on any port. But it works only with port B. Like the behaviour with the boost hub the display in the hub box is still erratic but the debug messages are correct. Unfortunately I cannot reproduce the heap of red error messages after connecting more than one hub to the software. It may be an effect of BT radio transmission (?). Still it does not affect the function this is only a neglectable note ... PS.: The software needs up to 30 seconds to detect a hub (both BOOST and tain) if the color/Distance sensor is connected, but in every case and combination the detection is reliable: Perfect!

Perhaps I need help: I've activated the debug message but there is still no display ... Did I unstand something wrong? The bug is gone (ok, there is still a heap of red error messages), thank you a lot for your IMMEDIATE reaction. I've found a missing program part: If I try to disconnect one hub by pressing the button until the led at the hub goes off the software does not recognize this but crashes after a while. For me it seem this part is not still programmed ;) Ok, this function is not so important since all hub switches off themselves after terminating the software.

This is not a double post. I downloades the new version beta 0.3 and tested it immediately a I am still thrilled! It works very well with all connected hub either train hubs or BOOST hubs. All are recognized immediately and configure themselves in a correct way: Imressive and very very comfortable. Any plugged device is recognized and connected correct despite it was connected after or befor starting the BT communication. All three avalable motors a treated in the same way: Like the train motor. You can set the speed from stop (0) to forward (+100) or backward (-100). This is not full speed, I've found in my last posting that full speed backward is -128 and full speed forward is +126. For the next version ... Only the sensor shows no reaction (?). It is recognized immediately but there is no display of distance of any object (white paper, my hand and so on) or color (different bricks and plates) In version 0.1 the result was displayed, ideal for test purposes. I used it to discover that the sensor is supported even by the train hub. Perhaps the new version can be enhanced by this function? Is it correct that the color sensor influences the programs? To test this I have to begin to programm own test sequences, somewhat complicated just for a try out. BTW, Creating own programs is a large advantage of the software. There is still a minor bug: If the first hub is switched on the software finds at first 29 non existing hubs, the the real hub. Further hubs are added to the list without flaws. Perhaps the empty error messages can be surpressed, they are a little bit distracting. But if you put the scrollbar to the far right you find the real existing hubs and can work with them . At the moment I know nothing better than Cosmiks software, much better than the original LEGO bloatware.

Hello everybody, I have two more results to report: First the speed value is not limited to 100. I can varied from -128 to +126. The value -128 means full speed backwards, 126 full speed forward. 0 stops the motor (and switch the lights off), but 127 do so too, by what reason whatsoever. Speed reglation is very smooth. What do we want more? Second yesterday a hand full of WeDo 2.0 medium motors arived and I could tested them this night: They are identical to the motors from the stubby batmobile (better chunk or brickball?). They are recognized immediately by the software and handled equal to the train motor in difference to the original LEGO app. This are good news since we brick railroaders need a lot of such motors for our mocs.The bad news are that theese motors are still incredible expensive. 20,- € or $ for a simple M motor is to much, LEGO. The old M motor from PF costs 9,- at the LEGO store. Now I download the new version an will play with it. I'm still thrilled

Of course I will do! I can add some information: The old, but once updated train hub is recognized as a BOOST hub, not as a train hub. The other train hubs are found as train hubs. Unfortunately at the moment I have no method to request the firmware version. Perhaps the just published docu from LEGO, JopieK or Treczoks know more ... I have to confess, I have no medium motors as in the batman chunk. It seems they are identical to the WeDo 2.0 medium motors, but it is rather expensive to check it ... On the other hand I need some since a lot of my MOCs run with medium motors (or even L motors which aren't available yet for PUP). The encoder motor from BOOST runs fine with the BOOST hub (as expected and tested) but shows no reaction anyway with the train hub. It is known that the train stuff tells the hub what is connected only by voltage levels on wire 5 and 6 but I do not know how the BOOST angle encoder tells his data to the BOOST hub.

@Cosmik42 First thank you a lot for your effort you share with the community. I am really happy, you make a real gift for us! Your software is really promising: Installation was easy, just unpacking the zip file to a directory and starting the programm. Both BOOST hub and train hub connect to the software immediately, but before this it is neccesary at the moment to restart the program. Switching off a hub by pressing the button for a while and restarting it does not work. At the moment I am not so familiar with your user interface but it seem easy to use after little training. The BOOST hub works fine with it. The internal motors and the external motor from the BOOSt set work flawless. On the first try he lights are recognized correctly but they do not react to any command. Instead there is a message like "Message unparsed 0x0a recieved". Ok, the train motor is not recognized by the first try but it is on the second together with the lights which are recognized as external motor on the second attempt. Then the train motor works together with the BOOST hub too but the software produced the unpased message in parallel. For me only a sign that there is still a lot of development work to be done. And I tested the distance/color sensor. Yeah, this device works perfect with the BOOST hub. My eldest train hub does not cooperate with this version. It is found by the software nealy immediately, but there are three instant error messages: Port 50, 59 and 60 not found. The result is no reaction neither for the train motor nor for the lights. Ok, explainable. One reason can be a buggy firmware on my train hub. It was updated once in november but afterwards LEGO introduced a new app which does not cooperate with my elder smartphone any more. At the moment I have no opportunity to update any firmware on my hubs. I have to buy a new one :( Surprisingly the color/distance sensor works with the first train hub. Short before christmas I bought three more hubs with probably another firmware version (I do not know which since my smartphone does not cooperate with the official LEGO app any more): At first attept the software and the hub firmware crashed but the second attempt was successfull: Both the train motor and the lights are recognized correct and do was they should despite a stream of violett messages ;-) The color/distance sensor works too. Summary: I am deeply impressed by your intention to share your strenious work with our train community! I hope it is ok for you reporting the experieces? I am really happy with your software even at this early state: The user interface is wonderful straight forward and really easy to learn (Congratulations!) and it works together with the BOOST hub and my three newer train hubs. What do we want more? Best wishes for the new year!

Hey folks, the documentation is a xmas gift to the community And it tells us a lot about the potential of the new system. LEGO itself had not enough time (and perhaps not enough resources) to offer the final solution in one step, since LEGO is at first a toy manufacturer and not an IT company. But I'm absolutely confident there will come more .. Programming the PU hub in the new app is a promising step in the right direction. And please, let's stop discussing about the component prices at the LEGO shop, They disapperared very short after beeing published (by mistake?). It seems, there is a lot of internal discussion at LEGO. I decided for myself to be a little patient ... Merry christmas to all!

PAB is often cheaper. It is worth to compare ...

WARNING: Do not install this version 2.0.0 if you run Android 5.0.2 or lower. The "updated" app crashes immediately you paired the hub (this works in most cases) and choose a mode. In this state the app is useless. We have to wait for a functional one which has been tested before publishing. Addendum: Ok, Android 5.0.2 is several years old. It is explainable the LEGO developers don't care of such old stuff. Nobody will develop software for Windows XP anymore. After a little look up LEGO gives us informations about the minimal requirements, and they tell clearly "Please use at least Android 5.1" Uh oh - just missed ;-) Time to get a new mobile phone ...

Wow, thank you for the information! Would it be a good idea to take directly PUP train motors to get the PUP plugs? The motors will be used anyway and there is no hardware waste.