How could I provide power and/or data coupling between train cars?

[ALCO]

Now if only there was a solution for the old magnet couplers... If one were using them for American trains without buffers....

[legoman666]

9 minutes ago, Lowa said:

Thanks!

Yes, putting a second one on top of the coupling seems to be possible.  It does start looking quite bulky though...

For the power + data connection I was thinking about getting some magnets that are slightly smaller (2mm diameter instead of 3mm) and arranging them like this:

That would be a lot more elegant than using 2 connectors with 2 magnets.  My only worry is the alignment of the magnets.  2 magnets will always make contact, with 4 they will have to be flush enough to make the electrical connection...  I'll give it a try!

3 would always make contact also as long as they're in a triangle shape

[TK2241]

I like the 3D printed idea - I think smaller diameter magnets would probably work better.  I really want to try getting 4 or 5 conductors into each car.  I'll try my hand at something as well, and hopefully I'll have a 3D printer in a couple months ;-)

[Rob Klingberg]

This is a topic that has occupied hundreds of hours of my time over the past three years, and many more hundreds of $$ in test part ordering.  There are some really good ideas here.  I won't claim to have solved for the scenario needing more than two wires, but I am working now to bring a 2-wire solution to market to be sold through my company, Brickstuff.  The plan is to officially release at Brickworld Chicago coming up in a few weeks.  We will have a solution both for train couplings as well as for connecting anything LEGO (floors within buildings as well as buildings themselves, sections of LUG layouts, or just about anything that needs a non-protruding connection that is magnetic).  As with the development of any product, the devil has been in the details: need to find a strong magnet, but not too strong, that can adhere to standard LEGO geometry, that is flexible to allow for connections at uneven points, but that is not so strong that it pulls the wires out when you disconnect.  I've also tested just about every glue, epoxy, and putty available.  Looking at an earlier post in this thread by @igordost, it looks we're using similar connectors (I sourced mine in bulk last year).  I'm using 3D printed ABS collets to hold the magnets inside the coupler shell.

Attached is a pre-release photo of the beta product aimed at the LEGO train market.  All parts are original LEGO.  Our solution for buildings and other structures will use 3rd-party injection molded coupler shells and pivots (in LBG color), only because the original LEGO parts are increasingly scarce and very expensive.  I have more photos (showing mounted configuration in bogies) here:

https://1drv.ms/f/s!AgcctLgYvJZfhrwsxUiQQJfmTTMvAg

We use very thin (but strong) wire, so the wire will seamlessly fit between bricks, plates, etc.  We've worked hard to make the couplers practically invisible once mounted.

Down the road, we're also researching 3- and 4-wire solutions, though there have proven to be special challenges there (namely, that the magnets are too strong and don't disconnect easily), but I'm confident we will find something workable soon.  We believe magnetic PF couplings for Technic vehicles are a must for the community to have.

Feedback is appreciated!  You'll be seeing more photos, video, etc. of this product line in the coming weeks.

--Rob Klingberg

Edited June 2, 2017 by Rob Klingberg

[skaako]

Awesome work Rob!! Can't wait for the 3 wire version to be available 

What would the price point (rough estimate) be for these once in production?

Would you consider selling the parts as a set? I'm guessing the labour involved for putting these together must take a while.

Edited June 2, 2017 by skaako

[Lowa]

2 hours ago, legoman666 said:

3 would always make contact also as long as they're in a triangle shape

True, and 3 wires would already allow one directional communication with, for example, a SoftwareSerial transmission line.  The only thing that would be a bit annoying is the polarity of the central magnet, we would need two versions... but that's not a fundamental issue.

I'm first going to give it a try with 4 magnets because that would allow bidirectional serial or I2C communication.  I'll first try with the 3mm magnets that I have.  If that works fine, I can try to shrink it using 2mm magnets...

 

[legoman666]

2 hours ago, Rob Klingberg said:

Our solution for buildings and other structures will use 3rd-party injection molded coupler shells and pivots (in LBG color), only because the original LEGO parts are increasingly scarce and very expensive. 

Steve Barile has injection molded lbg magnet holders in the past and sold them via his company, Altbricks, you should give him a buzz.

 

PS, I still haven't found a slip ring small enough to work.

[Lowa]

3 hours ago, TK2241 said:

I really want to try getting 4 or 5 conductors into each car.

I gave the 4 wire option a try, see below.  

As expected it's trickier as the magnets really need to be flush in order to make contract, but it works!  Below you see a quick video that I made.  The car on the left has an Arduino that 'blinks' on outputs D2, D4 and D6.  The car on the right has 3 LEDs: wire-1 = GND, wire-2 = LED1, wire-3 = LED2 and wire-4 = LED3.  As you can see the 3 LED blink one by one, so the 4 wire are connected.  This was done with the 3mm diameter magnets that I had.  For a 4 wire coupling, the 2mm would certainly be a better option as the 4 magnets make the connection between the cars very strong... I'm not sure the locomotive would have power enough to pull the cars apart on a decoupler.  

 

 

[dr_spock]

Nice. Would lightly spring loaded contacts also work?  Although extra magnets could help keep very long heavy trains together better.

 

[Lowa]

25 minutes ago, dr_spock said:

Nice. Would lightly spring loaded contacts also work?  Although extra magnets could help keep very long heavy trains together better.

 

I'm actually starting to think about using a mix of magnets and spring loaded contacts: 2 magnets on the outside to hold the connectors together and two spring loaded contacts in the middle.  That should eliminate all alignment issues and always provide a good connection.  The two magnets could be used for power lines, the 2 spring loaded contacts for data lines.

I should to do some tests with the decoupler to see how strong the additional magnets can be before having decoupling issue.  But you have a good point that for heavy trains the additional coupling force could actually be helpful...

Edited June 2, 2017 by Lowa

[TK2241]

That's great!  What about doing two magnets, maybe power and ground, and then pogo pins or something for the other connections?

I'm planning on using a fixed decoupler on the track rather than building one into the engine. That way I can manage any number of cars. 

Another thought that would be cool would be some sort of clamp mechanism (similar to a real world train coupler) that could be powered to open and close. I'd have to do some sketches. 

But I really like the magnetic coupler here - great start!

[Lowa]

1 minute ago, TK2241 said:

That's great!  What about doing two magnets, maybe power and ground, and then pogo pins or something for the other connections?

Indeed, I was thinking the same thing!  Two magnets on the outside for the power lines and two pogo pins in the middle for the data lines.  I also think we have to add a notch somewhere to make sure you can only connect them in the correct way.  Because if you turn the car around, you would connect it with reverse polarity...  

 

 

 

 

[Lowa]

7 hours ago, ALCO said:

Now if only there was a solution for the old magnet couplers... If one were using them for American trains without buffers....

Are you talking about those?

You would need a 'donut' shaped connector that you can slip over the coupling but that would look strange it guess...

Now, If you don't want buffers would a fully 3D printed coupling work for you ?  It should be possible to design and print a coupling with integrated connectors.  

[ALCO]

12 hours ago, Rob Klingberg said:

 

These are the style couplers I had in mind. This solution looks promising, but how do you account for the polarity of the magnets and the orientation of the car? I would think the magnets have to be able to figure out each other's polarity especially if you design it to work regardless of the orientation of the car.

[pirzyk]

3 minutes ago, ALCO said:

These are the style couplers I had in mind. This solution looks promising, but how do you account for the polarity of the magnets and the orientation of the car? I would think the magnets have to be able to figure out each other's polarity especially if you design it to work regardless of the orientation of the car.

Since they effectively become single sided and a matched set as long as Rob manufactures them the same, i.e. all innies have South Pole to the top and outies have North Pole to the top the polarity will be consistent.  It does force all your cars to be the same orientation and determinant on the power source car.

[Rob Klingberg]

12 hours ago, legoman666 said:

Steve Barile has injection molded lbg magnet holders in the past and sold them via his company, Altbricks, you should give him a buzz.

 

PS, I still haven't found a slip ring small enough to work.

Yes, Steve and I have spoken about this.  Unfortunately his manufacturer destroyed the molds they used to make his last batch of couplers several years ago.  That would have been a great option.  Thankfully I've found another supplier and already have a bunch of housings and clips in-house.

 

On the topic of slip rings, we will soon be selling the ones we use on our site-- not mounted into a kit, but individually.  Just right for any DIY LEGO project! :-)

10 hours ago, Lowa said:

I'm actually starting to think about using a mix of magnets and spring loaded contacts: 2 magnets on the outside to hold the connectors together and two spring loaded contacts in the middle.  That should eliminate all alignment issues and always provide a good connection.  The two magnets could be used for power lines, the 2 spring loaded contacts for data lines.

I should to do some tests with the decoupler to see how strong the additional magnets can be before having decoupling issue.  But you have a good point that for heavy trains the additional coupling force could actually be helpful...

You will need to be careful-- that design is already patented by LittleBits.  They have an ideal connector that they use in their products, but sadly they have been unwilling in my discussions with their leadership team to offer the connectors in bulk at a reasonable price.

As it turns out, Apple owns most of the patents for magnetic connectors that pass power and signal (Oh the patents I've read!), because of their MagSafe connectors from earlier MacBook models.  Any company seriously considering making their own custom connector and selling it at scale would be wise to keep the IP issues in mind.

2 hours ago, ALCO said:

These are the style couplers I had in mind. This solution looks promising, but how do you account for the polarity of the magnets and the orientation of the car? I would think the magnets have to be able to figure out each other's polarity especially if you design it to work regardless of the orientation of the car.

The great thing about this design is that you don't need to think about polarity at all.  As @pirzykmentioned, "innies" all have the same orientation.  The magnets in the male connectors are polarized to attract only female connectors, so as long as you use a standard in your consist (males at the front, females at the rear, etc.) the couplers will naturally attract/repel.  It's almost impossible to get a bad connection.  That was another key design consideration for us-- it had to be 99% automatic.

[Lowa]

50 minutes ago, Rob Klingberg said:

You will need to be careful-- that design is already patented by LittleBits.  They have an ideal connector that they use in their products, but sadly they have been unwilling in my discussions with their leadership team to offer the connectors in bulk at a reasonable price.

As it turns out, Apple owns most of the patents for magnetic connectors that pass power and signal (Oh the patents I've read!), because of their MagSafe connectors from earlier MacBook models.  Any company seriously considering making their own custom connector and selling it at scale would be wise to keep the IP issues in mind.

Good point!  I sort of overlooked that in my enthusiasm yesterday...  

I did take a quick look at the little bits patent (US D752,519 S) and it claims: "the ornamental design for a connector for a modular electronic building system, as shown and described".  It's a design patent not a utility patent; they don't make any generic claims on magnetic connector technology.  To me it seems like the goal is to make sure nobody makes connectors that are compatible with their products.  

The original patent on magnetic connector technology for electrical devices: Electromagnetic detachable connector, US 6030229A, Sumitomo Electric Industries, Ltd, https://www.google.com/patents/US6030229#legal-events) expired April 17, 2012 'due to failure to pay maintenance fee'.  So there shouldn't be any issue making magnetic connectors for LEGO trains, unless I missed something...

[Rob Klingberg]

37 minutes ago, Lowa said:

Good point!  I sort of overlooked that in my enthusiasm yesterday...  

I did take a quick look at the little bits patent (US D752,519 S) and it claims: "the ornamental design for a connector for a modular electronic building system, as shown and described".  It's a design patent not a utility patent; they don't make any generic claims on magnetic connector technology.  To me it seems like the goal is to make sure nobody makes connectors that are compatible with their products.  

The original patent on magnetic connector technology for electrical devices: Electromagnetic detachable connector, US 6030229A, Sumitomo Electric Industries, Ltd, https://www.google.com/patents/US6030229#legal-events) expired April 17, 2012 'due to failure to pay maintenance fee'.  So there shouldn't be any issue making magnetic connectors for LEGO trains, unless I missed something...

Good that you're doing your research.  LittleBits has many patents, most of which seem to deal with the connector.  If I'm recalling correctly, not all of them are design patents.  They're really broad, which means they might not be defensible, but LittleBits has tens of millions in VC funding, and they have been litigating aggressively.  Don't get me wrong-- I'd love to see advancement on this front.  Or maybe I've just been watching too much "Silicon Valley" on HBO... 

Edited June 2, 2017 by Rob Klingberg

[igordost]

22 hours ago, Lowa said:

Thanks!

Yes, putting a second one on top of the coupling seems to be possible.  It does start looking quite bulky though...

For the power + data connection I was thinking about getting some magnets that are slightly smaller (2mm diameter instead of 3mm) and arranging them like this:

That would be a lot more elegant than using 2 connectors with 2 magnets.  My only worry is the alignment of the magnets.  2 magnets will always make contact, with 4 they will have to be flush enough to make the electrical connection...  I'll give it a try!

Looks promising... How it works on curves? Does it let lego magnet coupler to move without limits? I tried to use 3L thin liftarm and there were problems when cars negotiated switches. 

[legoboy3998]

19 hours ago, TK2241 said:

That's great!  What about doing two magnets, maybe power and ground, and then pogo pins or something for the other connections?

I'm planning on using a fixed decoupler on the track rather than building one into the engine. That way I can manage any number of cars. 

Another thought that would be cool would be some sort of clamp mechanism (similar to a real world train coupler) that could be powered to open and close. I'd have to do some sketches. 

But I really like the magnetic coupler here - great start!

I use O Scale Kadee couplers for my trains.  The look and function like real train couplers.  The standard way to uncouple them is either manually with a "pick" or with a magnet under the track.  They now offer a way to remotely operate their couplers.  Info can be found here.

Sal

WFB, WI

[davidzq]

23 hours ago, Lowa said:

I'm actually starting to think about using a mix of magnets and spring loaded contacts: 2 magnets on the outside to hold the connectors together and two spring loaded contacts in the middle.  That should eliminate all alignment issues and always provide a good connection.  The two magnets could be used for power lines, the 2 spring loaded contacts for data lines.

I should to do some tests with the decoupler to see how strong the additional magnets can be before having decoupling issue.  But you have a good point that for heavy trains the additional coupling force could actually be helpful...

Would it be feasible to have magnets on one coupler, and magnets mounted on springs (or maybe just the ability to slide in/out a bit)? This would allow the two left magnets to be fixed, while the two right ones move, ensuring all 4 connections have conductivity via magnets, yet not have alignment issues.

The other thing to consider is having magnets aligned vertically, thus eliminating polarity issues. One magnet above the coupler, one below. You could then have additional magnets on the sides for things that aren't polarity sensitive (Such as power via a bridge rectifier). You could easily make a variation of Tom's design for this; a single magnet, and a three magnet version. Gets you one through six conductors.

[Trekkie99]

Any chance these designs could be used for the PF system? I was curious if I could use them for my monorails in order to allow for a battery box in one car and a motor in another which would allow easy separation of the cars.

[Lowa]

On 6/2/2017 at 8:17 PM, legoboy3998 said:

I use O Scale Kadee couplers for my trains.  The look and function like real train couplers.  The standard way to uncouple them is either manually with a "pick" or with a magnet under the track.  They now offer a way to remotely operate their couplers.  Info can be found here.

Sal

WFB, WI

That's quite impressive!  But if I got it right you'll need a servo motor in each car that you want to be able to decouple ?

[Lowa]

On 6/3/2017 at 0:11 AM, davidzq said:

Would it be feasible to have magnets on one coupler, and magnets mounted on springs (or maybe just the ability to slide in/out a bit)? This would allow the two left magnets to be fixed, while the two right ones move, ensuring all 4 connections have conductivity via magnets, yet not have alignment issues.

That would indeed solve the alignment issues, but I don't see how to make such a connector in practice.  I haven't seen any off the shelf 'magnets on springs' (like pogo pins).  But when I did the test with the 4 magnet connector, there were always at least 3 magnets that made contact.  So most likely 3 plain magnets (no springs) would work fine.

On 6/3/2017 at 0:11 AM, davidzq said:

The other thing to consider is having magnets aligned vertically, thus eliminating polarity issues. One magnet above the coupler, one below. You could then have additional magnets on the sides for things that aren't polarity sensitive (Such as power via a bridge rectifier). You could easily make a variation of Tom's design for this; a single magnet, and a three magnet version. Gets you one through six conductors.

Aligning the pins vertically indeed eliminates the polarity issues.  The main disadvantage that I see for using a connector above and below the coupling is that it starts looking bulky.  But if that's not an issue for you, it should work fine...

[igordost]

Custom lights in action :)