Cheap 3A DCC booster kit: Would you buy it?

[CaptOblivious]

This is a bit of a marketing survey. As some here know, I've been working on a design for a DCC booster. The primary design criteria were to be: Suitable for N-gauge (i.e. not an 8A booster!), relatively bullet-proof, easy to assemble, and dirt cheap.

 

Short version: I think I may have got it figured out. (Although the design hasn't yet been finalized, pending some last testing.)

 

The nitty-gritty technological details can be had on my blog, Railstars:

http://devices.artificial-science.org/tag/lolbooster/

 

The LOLbooster is a 3A booster, with lots of failsafes built-in. No need for a separate circuit breaker or power-management board. Compatible with most if not all manufacturer's command stations.

 

Although I am designing and building this booster for myself (why shell out $$$ for DCC equipment, when I can just build my own for ¢¢¢?), I wonder if others might be interested in purchasing one or more for themselves

 

The price point would most likely be somewhere around 75USD (maybe less; pending the final design), and would include all the parts, a nice PCB, instructions, full support, and a nice warranty (something like TCS's). You'd have to solder it together yourself (all the parts would be through-hole, easy to solder things), and find a wall-wart to power it.

 

So, the question is: Would you be interested in purchasing such a beast? Why or why not?

[KenS]

Regrettably, I just bought a new 5A DCC system last year (command station and a pile of circuit breakers), which is all the power I need right now (and I still have my Zephyr for bench work).

 

However, I've been following the LOLBooster work, and I think you have a good system there.  If I were building a new layout today I'd think seriously about using a bunch of these with my Zephyr as the command station instead of a more expensive 5A command station (and I could probably eliminate a few of the circuit breakers; I'm not sure it would cost less money in the end, but it might).

 

The big problem is that my Zephyr is limited to 10 trains by its software, and that's one of the things I fixed with a new system that a bunch of boosters wouldn't cure.  What you need to design now is a boosterless command station that speaks LocoNet. 

[CaptOblivious]

Regrettably, I just bought a new 5A DCC system last year (command station and a pile of circuit breakers), which is all the power I need right now (and I still have my Zephyr for bench work).

 

However, I've been following the LOLBooster work, and I think you have a good system there.  If I were building a new layout today I'd think seriously about using a bunch of these with my Zephyr as the command station instead of a more expensive 5A command station (and I could probably eliminate a few of the circuit breakers; I'm not sure it would cost less money in the end, but it might).

 

The big problem is that my Zephyr is limited to 10 trains by its software, and that's one of the things I fixed with a new system that a bunch of boosters wouldn't cure.  What you need to design now is a boosterless command station that speaks LocoNet. 

 

That's next on the table, except I'm leaning very, very heavily towards CBUS instead of LocoNet (for one, no royalties! for two, much better design…)

[Martijn Meerts]

I'm still not sure what to go for with regards to digital system once I start my N-scale modules. I'm pretty certain I'll use the Lenz system I have for the 0-scale stuff, even though they don't (yet) have servo decoders for the turnouts.

 

For the N-scale I'm leaning towards the ESU stuff. I already have the ECoS, and am quite happy with it. I haven't nearly explored all the features of it yet, but it's just a lot easier to use than anything else. ESU of course has their own boosters, a 4A and an 8A version. The 4A version costs some 200 euro which is definitely not cheap. Obviously it has a bunch of features that you hardly ever use but which drive up the cost. The one feature I do like is that when hook it up to the ECoS, you can see how much power is in use on the booster, so you know exactly how much you're using. It also has railcom, but not sure that's such a good thing :)

 

 

I wouldn't be surprised if quite a few people would be interested though.

[CaptOblivious]

Martijn, that's a good point. The design I'm finalizing now will have headers so that various data can be read off the booster (current consumption, fault conditions) by an auxiliary device like an Arduino. I want to eventually evolve the design into a "smart" booster that has all of this on-board, and with RailCom cutout support, but these features add a lot of cost to the design. In the name of keep-it-simple-and-cheap, I've left them off for the time being, though.

[David]

I've been following your project with interest, but unfortunately a booster is not on my list - it will be a long time before I need more amps then a command station will provide (once it hits retail I think I'll be settling for the new Digitrax as my first DCC system)

[CaptOblivious]

Well, for those who might actually be interested, I've got some PCBs on their way, and so I'm taking pre-orders on complete kits and bare PCBs on my blog, Railstars:

 

http://devices.artificial-science.org/lolbooster/

 

Next on the list is a command station and throttle to use with LOLbooster…

[Darklighter]

Is your booster compatible with DDX? I was thinking of building a ORD-3, but LOLbooster might be an alternative.

[CaptOblivious]

Wow, how is it I'd never seen the ORD-3 before? Interesting design; although LOLbooster does have a voltage regulator built in, unlike the ORD-3, it does not have a power supply on board. So you will need to use a wall-wart (more like a laptop power supply) with LOLbooster.

 

That said, so long as DDX outputs an NMRA conformant DCC signal (either common-mode or differential-mode will work), LOLbooster will be totally compatible with it.

[stevenh]

Woah... there was so much going on when I was tinkering around that I didn't know about.

If anyone can't find the ORD-3, then look here: http://wiki.rocrail.net/doku.php?id=ord3-cs-en (Google wasn't immensely helpful.)

 

Capt, nice work! I might just dispose of my junk yard and buy some shiny products from you when I get a sizable railway :)

[David]

That said, so long as DDX outputs an NMRA conformant DCC signal (either common-mode or differential-mode will work), LOLbooster will be totally compatible with it.

 

I'm not up on all the details of what an "NMRA DCC signal" is or how command stations even coordinate boosters - is it some kind of timing signal that tells the booster when to swap polarity? Do any commercial DCC systems actually support this - maybe a pin on the LocoNet cable carrying this signal?

[CaptOblivious]

David:

 

The DCC signal I'm referring to is just an un-amplified version of what goes to the tracks. Digitrax calls it "Railsync", and it's available on Loconet cables connected to a command station (although I forget just which twisted pair carries the Railsync signal just off of the top of my head). The NCE Cab Bus also carries the DCC signal on one of its twisted pairs.

 

I imagine Lenz's XpressNet carries the signal too (since they did, after all, invent this whole system), but I don't know enough about XpressNet to comment any further just now.

 

I'm working up a manual now; I will include directions for interfacing LOLbooster with Digitrax and NCE systems.

[David]

David:

 

The DCC signal I'm referring to is just an un-amplified version of what goes to the tracks. Digitrax calls it "Railsync", and it's available on Loconet cables connected to a command station (although I forget just which twisted pair carries the Railsync signal just off of the top of my head). The NCE Cab Bus also carries the DCC signal on one of its twisted pairs.

 

I imagine Lenz's XpressNet carries the signal too (since they did, after all, invent this whole system), but I don't know enough about XpressNet to comment any further just now.

 

I'm working up a manual now; I will include directions for interfacing LOLbooster with Digitrax and NCE systems.

 

Thanks, that helped explain it quiet well. A quick search for "digitrax railsync" shows it's the 2 outer pins on the 6 pin connector. The first link also gave me some useful insights into LocoNet, like how the basic boosters are "dumb" and throttles and other devices actually take power off the rail pins. Lets me start thinking about how (relatively speaking) easy it would be to create a "compatible" wireless DCC command station.

[CaptOblivious]

Capt, nice work! I might just dispose of my junk yard and buy some shiny products from you when I get a sizable railway :)

 

Thanks, steven!  Although I don't know what you mean by "junk yard":you have quite a nice diy system already,. I think .

[KenS]

I'm not up on all the details of what an "NMRA DCC signal" is or how command stations even coordinate boosters - is it some kind of timing signal that tells the booster when to swap polarity? Do any commercial DCC systems actually support this - maybe a pin on the LocoNet cable carrying this signal?

 

As Don noted, what Digitrax uses as Railsync is really just a low-power (full voltage but current-limited) copy of the signal put on the rails. This is really, really similar to the NMRA recommended practice (RP-9.1.2) for communication between the command station and booster.  And it's also compatible with what Lenz uses between their command station and boosters, and apparently also with what Marklin uses (at least on one model).  Digitrax has some application notes about mixing their equipment with that of other manufacturers down on the bottom of their Manuals page. According to its manual (PDF), the NCE PB-105 is similarly compatible with both Lenz and Digitrax.  This suggests they're all doing the same thing.

 

However, as far as I know, none of them actually claim to support RP-9.1.2, and there may be reasons they can't claim full conformance with the standard, even though they work together.

 

Boosters and command stations need to be more compatible than just this communication link, however, as when a train moves from one electrical section to another across the insulated joint between them, current produced by one may be returned to the other (this is why the ground wire linking them needs to be as heavy a gauge as the track bus wires, presuming you aren't using a common rail design). If you read them, you'll note that all of the references above mention connecting ground lines or connecting lines to the metal chassis of the equipment between command station and booster. That's the wire used to balance the power out between the two.

 

As long as someone hasn't used an incompatible approach to generating the track power, and is using a copy of the signal between their command station and booster, at least in theory you should be able to mix-and-match command stations and boosters.  I wouldn't recommend it if you haven't seen instructions from at least one of the manufacturers though, as it's likely a really good way to fry the electronics if somebody didn't do things the right way, and likely one not covered under the warranty.

 

That said, I'd be very surprised if Don's booster caused a problem for someone else's command station, at least from the above set of known-compatible ones.  And I'll probably be testing one with my Zephyr and, assuming both survive the experience  , my DCS100.

 

BTW, one thing to watch out for with Railsync is that it's not present on all LocoNet wires.  Specifically, the UP5 panel does not copy it out the "front" or "side" jacks, only through the two rear jacks.  What goes out the other on the rail-sync pins is the voltage from the attached power supply (if any) or nothing.  Since the only thing that needs true Railsync is the booster, this doesn't matter in most cases, but it can be a real headache to hunt down the cause if you wire up a booster the wrong way and something in the middle is failing to pass the DCC signal. The reason for this is that the actual Railsync signal is current-limited, so you can't connect too many throttles to it. By reusing the pins as a simple DC supply on the front-facing jacks, they avoid placing a limit on the number of throttles that can be powered.

[CaptOblivious]

Having just lookled at it today: Railsync is a +5V peak DCC waveform, and LOLbooster can perate directly from it.

[KenS]

Hmm.  I'm not sure why you're seeing 5V. Is that with the LOLBooster on the line as a load? Are you drawing too much power? Digitrax says (Loconet Personal Use, pg 3):

 

"Loconet devices may draw up to 15 mA from the RAIL_SYNC+ /- lines whenever the voltage is greater than 7V. The unloaded voltage is between 12V and 26V max. It is general practice to provide a LOCAL current limited copy of the closest track voltages, to pins 1&6 of Throttl e jacks around the layout. In this case the master "backbone" copy of RAILSYNC +/- is not on the Throttle jack."

 

I've seen other online references to it being a +7V "differential" signal (which may mean one line is +7 and the other -7, which would match the 14V output of the Zephyr).

[CaptOblivious]

A differential signal is just one where the signal is carried on both lines, but inverted. DCC is a differential signal. It may have been +7Vp, but I simply misread the display as +5Vp. The measurements were taken with the Railsync signal unloaded. The documentation you cited is frustratingly ambiguous as to which voltage measurements they are referring to…no great surprise.