See this is where I have been less than impressed with the DCC Specialties items, at least for the price. A Quad device has an MSRP of $85, for Tortoises with Loconet feedback. The DS64, which is not exactly one of the cheaper options, has an MSRP of $60 and includes the feedback, local control, and 4 additional sensor inputs PLUS works with both Kato and Atlas style solenoid motors AND Tortoises. I can add the auto-switch feature if I want to, for the extra $20, by adding some block sensors to the existing inputs on the DS64.
But then the cheap PA Dutchman in me comes out and says why even use Tortoises any more, servos are less than $2 each. Yes, they need more complex control devices, but commercial ones are available that give you, for the price of a Tortoise plus DPDT toggle and a couple of LEDs for position indicators, a servo, the control circuitry, panel buttons for manual operation, with LED indicators, AND a DCC stationary decoder.
The thing about loconet is you usually don;t need direct feedback, unless the device is only controlled via track commands - and even then that's not always the case. I wasn;t sure the DS64 worked that way until I hooked one up on a friend's layout and noticed the throttle indication for the specific turnout address changed even when I used the locl pushbuttons. There was no feedback wire from a Tortoise contact, just 2 wires to the Tortoise and a pair of pushbuttons hooked to the inputs for the corresponding DS64 output. Operate the turnout from one throttle, all others showed the new position. Operate it from the fascia buttons, all throttles showed the new position. This is, in many cases, good enough, since with models there are no lives at stake and a mechanical malfunction will only cause a derailment. You can carry it too far - how much feedback is enough? If you put a microswitch on each side of the throwbar to indicate position, what if the point rails break loose from the throwbar? To be absolutely certain you'd need to detect each point rail and know which one is touching a stock rail and which one is not. I can think of a way to do this but it requires making the turnout inherently unreliable as well as not very DCC friendly.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
I saw that on the Loconet Hackers list. I guess since you only need one, it may be fine, but otherwise I suspect this will be relatively expensive.
When it comes to any sort of device that users a microcontroller to endocde/decode a specific protocol like Loconet, making a thing that has just 1 or 2 lines end up being very expensive per line since for maybe $1 more in hardware costs ( a different micro of the same family) you can get 16 or 20 lines, and the code is not much different for 2 lines or 20 lines. Plus there is additional development cost since a commercial product will require at least the address to be programmable, whereas making a 1 off for yourself, you can just hardocde the address and leave out all that complex code. My Arduino-based system I am building is going to use CMRI protocol, and I have no plans to make my boards use a configurable address, I was going to hardcode the address each time I download the code. I suppose I could stick in yet another 8 bit shift register connected to a DIP switch and set it that way, reading the address setting each time the code starts.
This just in from Larry Maier of Tony's who make the Wabbit and who I also asked for interface recommendations... QUOTE Your timing is perfect. We are about to release an interface board specifically designed to connect the feedback outputs of the Wabbit or Hare to LocoNet. I don't know the exact timing yet, but it is on the order of a week or two. UNQUOTE
For this application, it should be pretty straighforward. You probably don't need any programming mechanism, rather you can just hardcode the Loconet address it uses and send one of two messages depending on the pin state of the pin you have connected to the Wabbit. I did see on Loconet Hackers that Alex said his library doesn't have switch position reporting messages but there's not really anything that would prevent you from using a block occupancy message to signal the turnout position.
If you want a challange to keep it small and use use a 'big' Arduino which is really overkill when there's only one pin to monitor, you could adapt Alex's library to the ATTiny85. It's a little 8 pin chip that can be programmed with the Arduino IDE (though you do need an Arduino to do it - or an Atmel programmer, but an Arduino is cheaper). You would need to change what pins are used for the I/O in Alex's code since the ATTiny85 only has 6, but that's plenty for Loconet IO plus monitoring your Wabbit. I picked up 6 for $3.95 (and now that I am confident I am going forward with them, I'll order a larger quantity next time, at an even lower cost per chip). My whole control susyem is going to be DIY, signals, detection, turnout control, etx. The Tiny85's will be the servo drivers for turnouts to offload the timing critical parts (servos need a set frequency pulse to drive them) from the main Arduino that handles the block detectors and signals around the turnout(s). Plus I will also use the Tiny85's with buttons as controllers for servos on turnouts that won't be under computer control, like in the yards. At 85 cents or less for the Tiny85 plus a couple of resistors and capacitors, they are hard to beat. That's to control TWO servos! Though if I want indicator lights, it may end up being 1 servo per Tiny85. There are ones with more pins but thay are realtively much more expensive than the Tiny85. The Tiny84 has 12 I/O pins but in over $3 even in quantity. Neat thing is many of the ATTiny chips have capacitive touch sense inputs - which means I could replicate those Barrett Hill Touch Toggles, although I would then have to do that everywhere as I am a bif promponent of ALL controls on the layout being the same, not one type for staging, another for the main, and a third for the operating yard. That just confuses people. All pushbuttons, all toggles, oor all touch sensors, never a mix.
Thanks Randy. Your explanation cleared up a lot of confusion for me, most particularly how LocoIO, PIC and Arduino are related.
I agree with you that Tony's description of Wabbit FB having "...feedback turnout position reporting to Digitrax LocoNet, Lenz ExpressNet and NCE CabBus." is a bit misleading, particularly in that their Quad Wabbit has the same description and one version does indeed connect directly to loconet.
I've decided to go with a DIY arduino-based interface http://www.spcoast.com/wiki/index.php/LocoShield It's probably overkill in not only functionality but also time and effort but I am thinking I'll learn something useful about the intricacies of both arduino and loconet. It is a hobby after all so doesn't have to make sense.
Jan
LocoIO uses a PIC, not nan Arduino. Arduino uses Atmel microcontrollers, which are now owned by Microchip who makes PIC but none of that matters as you buy the chip pre-programmed for the LocoIO. LocoIO is a general purpose input/output board for Loconet - the outputs can be divided however you like as far as input or output, in your case they'd probbaly all be inputs connected to the feedback line of the Wabbit, When the Wabbit changes the position of the Tortoise, the state of the feedback line changes, which cases the LocoIO to generate a Loconet message which can be read by JMRI or another Loconet decoder and acted upon. The RR-CirKits Tower Controller is a 64 line version of the LocoIO, they happen to have additional interface boards to better make use of those 64 lines but it can also be used like 4x LocoIOs.
LocoIO outputs can drive LEDs for signals, but using 3 outputs for one signal head is quite wasteful when with some simple logic you can use 2 lines to indicate up to 4 states - that's what the signal driver boards for the Tower Controller do.
You see why I am slightly annoyed that Tony's calls those Wabbits with feedback (same with the PSX breakers) when all it means is they have an output line that changes state, they do not directly generate feedback for ANY DCC system, you need an extra piece. Til it's all said and done, some of the other options are cheaper per turnout - the DS64 for example does 4 turnouts plus local pushbuttons and updates Loconet even when the pushbuttons are used. And if you want cheap - the NCE Switch-8 does 8 Tortoises, and then you would use a LocoIO connected to one of the Tortoise contacts to get feedback - I think one Switch-8 costs less than a Wabbit. Of if you don;t have many cases of 8 turnouts near each other, the Switch-It is a 2 turnout version, I've used those in the past, but I've since given up on Tortoises and use model airplane servos, far less expensive even given that you need a more complex controller to drive a servo.
Several options: inputs on the DS64, inputs on the SE8C, the RR-CirKits Tower Controller (although 64 inputs may be overkill), the semi-DIY LocoIO from Hans Deloof, a Team Digital SRC162 if yuou can find one (they are discontinued, the CSC is the replacement that also does more), CML DTM30 (now sold by Sig-Nal Trak, or DIY with an Arduino and the Loconet library.
Does anybody have any suggestions on how best to get a contact closure onto Loconet? I have a Wabbit FB turnout deconder but, unlike the Quad Wabbit FB which connects directly to Loconet, it provides only an opto-isolated open collector transistor for feedback. The Wabbit documention suggests interface devices that are full function turnout decoders themselves which would seem to make the Wabbit redundant.