New Daisy Chain USB Connector?

Hi Rob - Let me expand on the idea I mentioned earlier.

Use Case: Connect multiple buildings in a show environment.

Background: Clubs (and some individuals like myself) increasingly assemble collaborative layouts at shows.  There are standards for interconnecting but not for lighting.   In KLUG we have three themed collaboratives that all use a proprietary lighting standard.  The idea is to power a collection of modules/buildings quickly and reliably.   Small lighting connectors are delicate, take time to connect, and can fail eventually with frequent use.  Horizontal power connectors require a regular grid, are expensive, and can be fiddly.  And a power hookup is still required.  And these can be overkill for displays with just a row of buildings.

Proposal: Daisy chain buildings along the back of a table using a new USB connector board. USB connectors are strong, robust, cheap and VERY QUICK to assemble and disassemble at shows. USB 3.0 is rated at 900ma which could support 3-4+ buildings on one power supply. Lightailing sells a similar board: https://www.lightailing.com/products/lightailing-converter-module but it has 4 USB and uses only pico connectors. Adding a Brickstuff regular connector would be superior in that it can carry more current for the trunk lines, but keep the pico connector for flexibility. USB of course can be connected to mains or battery.

Thoughts?

 

 

10replies Oldest first
  • Oldest first
  • Newest first
  • Active threads
  • Popular
  • Very interesting idea, thanks for the additional background.  I think it's a great idea.  My only question would be: why pick the USB connector as the "standard"?  Why not go with something like an Ethernet cable, which are generally available, cheap, available in multiple lengths, and have the ability to control multiple "channels" over a single wire?

    Believe it or not, 10 years ago (!) when we were just starting out, we actually designed a board like this.  Have a look:

    I absolutely cannot believe we've had this design for TEN YEARS now!  We never ended up moving forward with production, but I've always liked the idea.  You could use each channel to control a segment of lights if you wanted-- for example, putting all streetlights on "channel 3", all first-floor lights for buildings on "channel 2", etc.  Then connect it all up to one master controller, maybe with a remote, and you can command your entire street.  Setup and take-down is also easy because of the RJ45 connectors.

    Let me know what you think!  I'm also interested in what WRTyler thinks of the idea for his setup.  Jim Pirzyk as well....

    Like
    • WRTyler
    • -the National Mall guy-
    • wrTyler
    • 2 yrs ago
    • Reported - view

    So, except for potential power loss to the  USB hubs themselves, how would this be different than using some daisy-chained hubs?  Two four port USB 3.0 hubs, with the second hub connected into the first provides seven USB ports.  In fact, that's what I did the first year I had lights on the National Mall.

    Now, I don't know the electronics behind this to know what the limit might be for running power before encountering problems, but I'm not seeing a definitive benefit.  Sure, it might be simpler and easier to have a dedicated USB connector, but I look at it as reinventing the wheel, except the wheel might last slightly longer.

    Am I missing something here?

    -wrtyler-

    [the National Mall guy]

    Like
  • So my initial thought is also why the USB to just transport power, it does seem a waste of the electronics in the USB adapter.  As Brian Williams mentioned, KLUG has their own connectors, they use barrel connectors for power.  That seems a more cost effective solution while providing the speed of setup.

    I do like Rob Klingberg 's idea of having multiple power channels on a cable.  I want to move my streets to be able to have 3 separate power lines (shared common) and this would assist with that.  Doing multiple power channels will require a special device to put power on each line.  The downside of using an Ethernet connector is the size.  the USB could be contained w/ maybe a 1x2 brick, whereas the Ethernet cable would need something more like a 1x2x2 brick.  I used a RJ11 cable for wiring up my signal bridge.

    If we tried to implement that for the USB case, I think that could be dangerous if someone plugs in a device that expects standard USB connection, one could possible damage that device.

    Like
  • r/e Multi-Channel - Rob Klingberg Actually I had wondered about this and really like the idea.  Multiple channels permit zones to be setup and powered independently.  You could have tables 1,2,3, or buildings, streetlights, and vehicles/accessories.  This is not just a logistical advantage, but can also be used to load balance the current draw.  I'm using many motors now which draw a lot more current so having the option of multiple buses / channels would be great.  And you can put little LEDs on the board indicating 

    r/e Connector Type - I'm not partial to any particular connector but my thoughts:  My personal experience is  that RJ11 and RJ45 connectors are unreliable.  Sometimes you have to jiggle them to make contact, they corrode, and often the locking clips break off.  Maybe the companies I worked for used cheap ethernet cables.  For what its worth, I have several old Lifelites power distribution boards with barrel connectors that I still use after 10 years.   Lastly, a point in favor of USB is that the last cable plugs directly into a battery or power adapter.

    r/e Orientation - I like the board diagram that Rob Klingberg  shared.  One thought: the connectors need to be rotated 90deg so they both face the same direction.  If this board is taped to the back of a building, the cables should hang down.  If they are oriented horizontally then the cables will hang and put pressure on the connectors.  Also the cables may likely be visible on the sides of the building.  If used in a horizontal configuration, then the user will likely want both cables coming out the back side.

    Here is the SSOD that Jim Pirzyk  was mentioning.  Notice the power cables are spliced to provide multiple barrel plug outputs.  I think the daisy chain idea would be easier logistically and a less cluttered.

    Like
  • Jim Pirzyk Yes, the biggest concerns I  have with USB would be:

    1. USB-Micro connectors are incredibly fragile.  Would not recommend them for heavy/conference show use.
    2. The other option for USB would be USB-B, which is used by many companies like Arduino to provide a solid connection.  Problem there is that these plugs are rare.
    3. USB specifies a max current of 500mA (pre-USB-C) so you know people would plug the cables into their laptops.  I know because we have customers who do this every day.  I don't want to be on the hook for damaging someone's laptop.
    4. USB-C would be the choice if we did something with USB cables.  USB-C has a much higher current spec, and there is also a spec for "power only" that doesn't require use of the USB-C power negotiation protocol.  Problem here could be that USB-C is still not common in the USA-- we have our European customers telling us all the time that it's been years since they saw a MicroUSB plug or anything other than USB-C.
    Like
  • Also, we don't have to be limited by RJ45 or USB connectors of course.  Here are some other options we have lying around the shop.  Thoughts on either of these?  FYI, the 6-pin connector is not directional, meaning you can connect it in either orientation.  So there are really 3 usable pins here, not 6.

    Like
    • WRTyler
    • -the National Mall guy-
    • wrTyler
    • 2 yrs ago
    • Reported - view

    I have been reading this thread with interest.  And even though this has gone pretty far afield from Brian's original idea, it has brought up an interesting problem and looked at some interesting issues.  So I'm going to react to all this, even though I'm obviously out of my depth in discussing these ideas.  

    Brian first posited an idea for connecting several lit MOCs together so that the power connections would be uniform, universal and rugged even though the lighting system within each MOC might be different.  Since I have never worked on a collab project, I don't really know the scope of the difficulties, I can only imagine them.

    What I have observed and experienced is that currently most LEGO lighting systems/kits start with a USB-A connector.  I don't know if that is in the midst of changing, but something will probably fall out of the EU proposal that the standard port for electronic devices be USB-C.  But that's beside the point.  What I want to highlight here is that the majority of LEGO lighting ecosystems have a USB-A connector in common.

    So, in thinking about a power system to connect a collection of MOCs, it would seem reasonable (at the current time) to have it start with a USB-A connector, and then extending that, all the limitations that come with using that as the basis of power.

    Then, as I mentioned previously, I don't see a major benefit for having a custom designed USB-A power delivery system over using existing and widely available USB hubs.

    But this conversation has grown beyond just providing power to MOCs.  We are now talking about power AND control, which means any two wire system will not work...other than everything on or everything off.  That increases the complexity of the power system significantly AND changes what connectors can be used and how they are used.  This is a step away from a uniform, universal power system.

    With a control system, there needs to be a controller, a controllee and some way to communicate between them.  Something has to send a signal to turn on/off, something has to receive that signal and act on it, and (ideally) send a message back, "Okay I'm on/off."

    And this is where I'm opining on something I know nothing about.  Once you start integrating a control system with power, you are creating your own power ecosystem.  At that point, the system will need to include a way to get back to controlling power to elements that are from outside the ecosystem...back to turn on the power, turn off the power.

    There are lots of ideas floating around out there, and the three of you have all proposed interesting ideas worth pursuing.  But it now becomes, company A has system #1, company B has system #2, how do you get system #1 working with system #2 in a reliable, uniform and rugged way.  And what happens when company C comes in with system #3 a year later?

    I don't know the answer.  I don't even know if there IS an answer.  So my thinking is...design something that is so simple, easy to implement, economical and reliable that other companies will gravitate toward that rather than try to design their own.  Oh yeah, there will be outliers that will think, "Hey, I can do better than that."  But the challenge here is not to capture 100% of the market, but enough of it to be sustainable.

    Okay...ENOUGH!!  This is just me free-thinking and trying to make some sort of cohesive statement.  I'm just putting it out there so I feel like I've contributed *something* to this conversation

    -wrtyler-
    [the National Mall guy]

    Like 1
    • Josh
    • BCFR, BCEMS, CEFR, NMCSO
    • Josh
    • 2 yrs ago
    • Reported - view

    Tyler brought up a good point. Because the EU is on the brink of making USB-C the standard (thank goodness), that could make a USB-A design obselete.

    Modular cities, and collaboration efforts (like LUGs) is probably the biggest market for a system like this. MILs is the most accepted modular design and I think Rob, that your 6-pin magnetic connectors would be the best choice here...at least from my point of view. One connector on each side of a MILs building would make moving around cities ridiculously simple and would contain all electronics into the building, self containing all wires and connections. In fact, I was just designing my own magnetic connector lighting system for MILs just the other day. Oh and like Tyler said, simple is good, and I couldn't think of something more simple than just a single magnetic connector on each side of a building.

    Like
  • r/e value - WRTyler  - The value of this proposal is in the simplicity and interchangeability of daisy chaining.   Multiple channels only increases this and can be simply controlled by what you plug into each channel.  Consider the alternatives to lighting a dozen buildings across 3 tables.  Most likely you would run extension cords and surge protectors under each table and plug a combination of USB wall chargers with multiple cords running to each building, etc.  Or you might have a few high capacity batteries plugged into BS dist boards with spaghetti wires to all the buildings, etc (where I am at personally).  Being able to daisy chain simplifies and standardizes the connections.  This speeds up setup, take down, and means any of the buildings can be added, removed, or relocated with ease.  Thats a big help when you are setting up a show collab where you don't know who is going to bring what. 

    r/e USB-C - Again, I'm not partial to any particular connector.  It does seem that USB-C may poised to replace USB-A.  I like that its physical size is smaller.  But how reliable is it long term?  Who knows.

    r/e magnetic - I do like the mag connector idea.  I could see this being used horizontally between all  modules (buildings, streets, etc) and vertically too.  Create a module set with 4 connectors + 3 wires to connect and put them on all sides of each module.  Then plug a mag power connector onto the back of any of them and the whole layout lights up.  Would need to be cheap though.

    Like
    • WRTyler
    • -the National Mall guy-
    • wrTyler
    • 2 yrs ago
    • Reported - view

    Brian Williams  Thanks for explaining it one more time.  Finally got it through my thick skull what you are trying to achieve...and it makes great sense.

    Like
Like1 Follow
  • 1 Likes
  • 2 yrs agoLast active
  • 10Replies
  • 117Views
  • 5 Following