Dave B, Does it make a difference the number of turns in the coil? I suspect that is what is meant by 'ratio'? Also this site http://home.cogeco.ca/~rpaisley4/DccBODvt5.html lists several coils and available suppliers. Dave M
Not sure about the turns & ratio - all of the units I tested (details on the web page) worked without their burden resistor so I don't think the number of turns varies much - give one of the ones I tested a try as they do work dave
David, Here is a a 4 channel block detector by Iowa Scaled Engineering. http://www.iascaled.com/store/MRB-BD42?filter_name=dcc see documentation tab. Schematic; http://www.iascaled.com/docs/mrb-bd42/mrb-bd42-schematic-hw1.1.pdf and a 4 channel expander: http://www.iascaled.com/store/MRB-BD4X The source code is GPL v3 https://github.com/IowaScaledEngineering Single Channel DCC Block Detector code https://github.com/IowaScaledEngineering/ckt-bd1 Four Channel DCC Block Detection code https://github.com/IowaScaledEngineering/mrb-bd42 Maybe it can provide some more insight. Kevin
I am using NCE BD20s for block detection. With three wraps on the coil they will detect with an 18K resistor. They can be directly connected to the Mega dio inputs. Works fine in all my testing so far.
Until you need 64 of them. The Digitrax BDL168 will detect 16 blocks for $120 or $7.50 a block. It just won't work with the Arduino. Dave B's coil and simple circuit is very appealing. Dave
Dave - Before you start building 64 circuits I would spend some time testing it in your operating environment. Please keep in mind that the coil and circuit that I have been using works well on my bench. It has not been tested with long leads or in an electrically "noisy" environment. It may work well there, too, but needs lots more testing to be sure! dave
What is the recommended resistor to use in making your own resistor wheelsets? I am working with HO Scale so would it be a different draw than say working with N Scale? The voltage is higher in HO Scale than N Scale? Suggestions? Thanks.
Thanks Dave, some onsite testing is coming up but I am grateful for the path you have blazed. We recently moved to a home where we intend to live out the rest of our lives and the 5x16 walk around layout doesn't fit in the new space, so a 31 ft along the wall layout has been begun. It will have staging on the lower level with a double track helix to the main level and is being built in 9 sections (modules). Excluding staging/helix the mainline run is about the same as the old layout. Staging and a working yard add 18 turnouts to the previous count. I have 2 Zephyrs, 2 SE8c, 2 DS64 and 3 BDL168 but would need another BDL168, 6 DS64 and 18 Tortoise to equip the layout. And that doesn't count the servos and HES sensors for the Faller car roadway. BTW automated running and remote operating by grandchildren are part of the "Givens". Enter DCC++. I read about it a few months ago but when TwinDad connected the Base Station with JMRI it got my attention. With DCC++ each section can utilize a Mega that does all the servos/signals/detection for that section, making each section a very tidy package that would include 3 power districts and two auto-reverse sections. I have done some experimenting with Geoff Bunza's Pro Mini DCC decoder and servos to control the Faller cars and expanding that technology to the rails is a logical step. Dilemma: Using Digitrax Zephyrs, signaling and detection with servo turnouts will cost about the same as starting from scratch with DCC++ but I loose the modular electrical advantage. Continuing with Tortoises and DS64 adds considerably to the cost. DCC++ and Arduinos are a natural fit with the Faller cars. And the question always exists where will DCC++ be in 10-15 years, the expected life of this layout? Where will a Digitrax system be in 10-15 years? Thoughts of other members of this group are welcome. Thanks again to Gregg, TwinDad, Dave B and all those who are helping make DCC++ a practical solution. Dave
HVT, you raise some good points....I've been using DCC++ on a demo module layout for christmas time during the mall shows....my own layout uses Digitrax Super Chief with boosters, wired and wireless throttles etc...I've been using Arduinos for quite sometime experimenting with signals, block detection etc. Mostly as an automated demo layout. As far as long term I plan to stick with my Digitrax system until the open source systems become stable in the operations.
Scott, thanks for your comments. Dave Bodnar's success with detection coils have provided the missing link. Getting an Arduino to feedback to LocoNet is problematic at best, but a DCC++ Arduino Base Station eliminates that problem. In my own case when testing proves successful (and I expect that it will) I will most likely sell the Digitrax components and move forward with the new technology using help from members here to overcome obstacles. I'm kinda a gadget guy anyway. Dave
I've ordered the components to make the block detectors and once completed the block detectors along with the Turnout controls using the relays for coiled switch machines such as atlas and bachmann brands, I will have a complete automated demo layout to use this upcoming train show season.
Not trying to muddy the waters. What about using an RFID database on an Arduino to do block detection? Each locomotive or EOD device carries a small RFID tag. and the reader is mounted in the track bed. Reads can get down to milliseconds, and up to 12cm away. If there were someway to pass the read info from the database back into the GUI then you could get active feedback of what train is where. http://modelrail.otenko.com/arduino/tracking-trains-with-an-arduino-and-rfid-implemented
Though RFID would be able to detect unique trains, and an RDIF tag is cheap, the sensors are expensive. The one in the video is $32. Would be fine for a single station, but maybe not so practical if you needed 10-20 on a layout.
Hello gentlemen. I have designed and tested this adjustable sensitivity circuit of a occupancy detector, based on an induction coil. By adjusting the value of R3 the sensetivity can be increased or decreased. With the 5.6kOhm resistor shown I was able to detect 10kOhm across the tracks. I used a sprog 3 with 15V PSU as my arduino motor shield has not arrived yet. I have only tested it in a lab setting so the sensitivity might need readjusting to cope with track inductance etc. R3 could be replaced with a 10k pot for easy adjustment. I used a PE-51719 current sense coil, http://www.ebay.com/itm/171786677437?_trksid=p2057872.m2749.l2649&ssPageName=STRK:MEBIDX:IT but any current sense coil (or maybe even a regular toroidal coil) with 100 windings or more should work. kim.
Kim - interesting circuit - let me make sure I understand your diagram. V1 represents the DCC power on the track. S1 represents a locomotive or truck with a 10K resistor between the wheels. T1 is the coil with several turns from the track wound through the coil. D1, D2, D3 and D4 make up a bridge rectifier who's output goes to the base of the NPN transistor, Q2. When a car is on the track (S1 is closed) the Arduino pin is low, when S1 is open it is high. Does that sound right? thanks dave
Yes you are correct, any load on the track exceeding 15v/10k=1.5mA will trigger the sensor, if this is too sensitive lovering the value of R3 will make it less sensitive. You only need one turn of feeder wire in the coil for it to be effective. The circuit's current draw, from the arduino, when inactive is 0.1mA and 0.5mA when activated. thanks, Kim