I have a question regarding the block occupancy dedectot Twin T (or single T) used years ago by a lot of model railroaders.
Can this occupancy dedector be used with DCC for detecting trains and switching the signals and switch a asymetric(ABC brake system) signal to a part of the previous block so a following train will stop?
Conceptually it ought to work but I doubt any of the transistors specified for the Twin-T are nearly fast enough to handle the DCC frequency. Most of the design featuresof the Twin-T were there to overcome the biggest problem with detection on DC - the fact that when the train is stopped, there's no power to the rails. Sicne DCC always has power to the track, there are far better (anc actually simpler) ways to detect occupancy. Some use diode drops (similar to how the Twin-T worked), but I prefer the types that use current transformers, like Bruce CHubb's DCC Optimized Detector, the DIY on Rob Paisley's site, the one Dick Bronson at RR-CirKits has, one of the MER DIY ones, and others. The difference is the diode drop ones wire the diodes in series with the trakc power, reducing it by one or two diode drops. If you have undetected sections mixed in, there can be a noticeable speed difference unless you install the same diodes (sand the rest of the circuit) in the indetected areas. The current transformer detectors are compeltely isolated fromt he DCC power, the track power wire for the block to be detected passes through a current transformer, where it induces a current int he secodary which is amplified by the rest of the circuit. Some of these will have outputs that cna rigger a relay, whcih can switch in the diodes needed to generate asymmetrical DCC for the stopping blocks.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
To add to what Randy has said - You might have problems finding the transistors or replacements. A club had a Twin-T detection system and after about 10 years started seeing false indications(leaky transistors). The bias current feature in them to allow detection in a DC block that is turned off is not needed with DCC.
There are built-up good detection/signal offerings from a number of manufacturers. Unless you need lots of detectors, the commercial offering may be cost effective. Be aware - detection/signaling can be costly! I have 8 blocks of detection, the ABS logic, the signals, and detector wheel sets - About $800 involved so far!
Jim
Modeling BNSF and Milwaukee Road in SW Wisconsin
Thanks Randy and Jim for the info.
On my old layout(dismanteled 20 years ago) I made my block detection my self(not the Twin T but a detection with one 2N3055 and 1 diode) and building that was also part of the fun
I think if I am following my plan( I am now finishing the place{dividing my garage in 2, plaster walls, rised floor,etc...} for my layout and than I will start building) I will have for the double track mainline between 12 and 20 blocks(6 to10 in each direction, including 2 main lines in the staging yard).
My goal is to have a completed automatic{because there will be only 2 operators, me and a good friend},not computerized, main line(would be in dogbone configuration), manual stations and yards and depot, taking out trains of sequence(meanwile other trains running automaticly) switch with those trains and send them back in the sequence, etc....
Beacuse in the club, where I am a member, they have a commercial block system and the price is relativly high; al my cars and freight cars wheels are treated with conductive paint or have a resistor to been dedected so I was thinking on using a current sensing detection.
Because the output of it can switch on/off (via opto coupler) a relay or transitor(prefer a transistor) for switching the stop section of the block(on{previous block occupied}= power via diode to stop section: asymetrica, ABC, Red signal; off{previous block not occupied}= power direct to track, High Ball) and also the aspect of the signals.
Personal I am not for current sensing transformers(wires thru a coil) someone told me they are not so accurate or I am misinformed
You were misinformed about the current transformers. They can be set so sensitive that a change in humidity will set them off. Obviously you wouldn;t want to leave them like that. Depending ont he circuit they can be reliably set to trigger with as much as 20 to 30K across the rails, a pair of 10K axles on each car will be dead on reliable.
As for the no computer - but you want ot have sequence operations. This is WAY easier to automate with a compute in the mix instead of trying to have this detector trigger that stop section (then how do you start moving again? The block is occupied). As much as JMRI is free, RR&Co is WAY easier to accomplish this with. Don't even need DCC stop blocks, you just set up the layotu with the block detectors in RR&Co and you cna have trains move along following occupancy and not overrun each other. Or add in scheduled station stops and so forth so they aren't just chasing their tails. A friend of mien ahs autoamted a troley line through his town, I striggled with JMRI's automation scriting and never got it to work reliably - in fact there were cosntant false occupancy detections which really drove it nuts. One week he downloaded a demo of RR&Co and within a week had it going, perfectly fine. I've seen it keep chugging with 4 trolleys never getting stuck or crashing, for several hours straight while we worked on other things.
Same with signals. I've read the manual for signal controllers like Team Digitals that don;t need a computer. Figuring out the CVs for all that stuff - no thanks, I'd rather use the computer with pretty easy ot understand stuff like (Digitrax example here, but use a different system and the only thing that changes are the names of the devices) LH100 = Yellow IF LS201 OR LT203 = Reverse
Signal 100 is yellow if block 201 is occupied or turnout 203 is reversed. That sort of thing. And it's flexible, you can set up two files, one for ABS and one that overlays CTC and has a dispatcher panel, mostly just some logic changes to make it all work either way. No rewiring, no complex CV programming.
One of the advantage of the ABC braking system is that if you built it by yourself, it would not cost you “an arm and leg”, the diodes( 4 in series and 1 in the other direction) are cheap and a small relay will do. For the yellow aspect you make a short on 2 of the 4 diodes and the train will slow down in the stop section and continue and slowdown in the next block.
Are there some modelers that use the ABC braking system(either commercially{ex: Lenz} or either build by themselves) on their layout?
Someone told me that it is better to feed the stop section with regular DC, when a locomotive enters the stop section and you have green aspect the locomotive will continue and on red aspect the stop section is switched from feeding with DCC to feeding with DC and if the locomotive enters the stop section the decoder will sense the DC and will slow down(only for yellow aspect I do not know how you can make it happen).
But I would know how to build, by myself, a simple current sensing circuit for DCC(like the TWIN T, it was simple and effective) to detect trains in the blocks.
And it may look silly but maybe I love to see trains “chasing other trains their tails”, I enjoy to take out trains of the circuit and switch in the station or hold them there and take another in and send the first back out, and in the staging yard{6 tracks, the main line not included} the first train comes in and wait there till the 6th train is on the 6th track and then the first goes out( at the start of the session the main line and the first track of the staging yard must be free); but everything starts with a good and reliable detection system, but I do not want to spend all my wages on it.
The Chubb DCCOD or Rob Paisley's detector are only a coupel of bucks per block, if you wire them yourself instead of buying circuit boards. They both are just open collector outputs you can use for pretty much anything.
If you look on the JLC web site, or in the MR article where the DCCOD was presented, it looks like you have to buy the transformer from them, but they are available from Mouser, I looked up the part number. Rob's will require some research, the transformer he specifies is no longer available. I did look it up inthe past and got the rating, but I forget what it was, there are plenty of options though. Or just ask Rob, he readily answers questions about his circuits.
You will also find other useful elements among his circuit pages. http://home.cogeco.ca/~rpaisley4/CircuitIndex.html
Also, if you use DCC, another option, even if you don;t use Digitrax, is the BDL-168. It's a diode drop board, but 16 detection sections for under $150 (way under street price, that's MSRP), with the circuitry needed to feed detection information to a computer or other boards is pretty good. The thing about using RR&Co like I mentioned, there is no need for any DC or ADC stoppign blocks - when one train gets too close to another it is slowed down by actually tellign the decoder to slow down or stop as needed. And you cna do rerouting liek you mention, send oen train into a siding so another cna pass it, then come back out behind, or hold it for a few cycles before sendign it back out.
rrinker The thing about using RR&Co like I mentioned, there is no need for any DC or ADC stoppign blocks - when one train gets too close to another it is slowed down by actually tellign the decoder to slow down or stop as needed.
The thing about using RR&Co like I mentioned, there is no need for any DC or ADC stoppign blocks - when one train gets too close to another it is slowed down by actually tellign the decoder to slow down or stop as needed.
The thing about using RR&Co like I mentioned, there is no need for any DC or ADC stoppign blocks - when one train gets too close to another it is slowed down by actually tellign the decoder to slow down or stop as needed. And you cna do rerouting liek you mention, send oen train into a siding so another cna pass it, then come back out behind, or hold it for a few cycles before sendign it back out.
I am very new to DCC for building my home layout, years ago I have got an analog layout.
So if I understand it well(I visited the site of RR&Co, but I did not have time to view it in depth) in that system there are no stopping blocks and are the trains regulated directly via their decoder and via the computer?
It looks simpel but where is the catch?
How do a train stops than in front of a signal?
The catch is the RR&Co isn't cheap. To get exact stopping positions, the easiest way is to add additional sensores that are point detection - like IR beams, so you know exactly where a train is, not just that it is somewhere in Block A. It can be done without thsi though, but you need your locos calibrated and using good quality decoders so the distance covered in X amount of time at speed step Y is known. A strong case for quality BEMF. The other way is to have each block divided into 3 sub blocks, entrance, middle, and exit. More accurate location than just a single big block but not as precise as a point detector. However, if you watch real trains, they don;t always inch up right ot the signal mast, sometimes they stop well back. Right by my house, sometimes WAY back - about 300 yards from the signal, the track is level with a parking lot next to a coffee shop.
I read the manual of the RR&Co and yes it is not cheap.
I looked up on a link provided here and some circuits are looking easy to construct.
Has someone good experience with home made those circuits?
Please let me know your experience.
Are there some model railroaders who are using the detection systems of Lenz BM1, BM2, BM3?
And what is the experience of the users with the Lenz BM1, BM2 and BM3?
Someone is using the BD20 of NCE or the Powerhouse BD20? What is the differance between the NCE and the Powerhouse?
And what is the experience of the users with the BD20?
Can the output of the BD20 switch or steer 1 or 2 opto-couplers( one opto coupler to swicht the red/green signal and the other opto coupler switching the stop section, providing asymetrical DCC to one track)?
Never saw it referred to as a Powerhouse BD20, NCE uses the Powerhouse name for their top DCC system. Same thing I guess.
Can it drive other stuff? Yes. You might have to devise appropriate interface circuitry, same as with pretty much any other detector. They pretty much all work the same, either source a 5V or so signal when the block is occupied, or act as an open collector output and pull the output to 0V when the block is occupied.
Not all decoders support asymmetric DCC for stopping, so using that method may be a consideration. The other consideration is, are you more comfortable with circuit design or computer logic, which would be the determining factor if you go for a hardware solution or use a computer to trigger events based on sensor inputs.