Block detection/wheel set resistors
|
Brent Johnson
I am building a large layout that I am blocking for future CTC signaling. I know that it’s recommended to place resistors on wheels of freight cars to help detect the entire train and not just the engine or lighted cars. Thanks, |
|
|
|
Bill Wilken
You've given me an idea. Before testing again, I will remove every car/loco off the track. On Wed, 2020-09-02 at 11:21 -0700, Brent Johnson via groups.io wrote:
|
|
|
|
There are a few possible solutions. (I didn't think of any of these.)
1- Put gaps in the spur tracks and connect the feeders to the power section bus, not the signal detection bus. This is what I'm doing. 2- As you suggest, detect the locomotives and cabooses only. The problem with this is that short sections (i.e. shorter than a typical train) would show clear when it's not. If you're doing CTC, most of your OS sections (like the turnout at the end of a siding) would fall into this category. 3- Put optical detectors at the block boundaries (I got this idea from Dick Bronson). Set it up so that if the optical detector is blocked it makes both track sections show occupied. This eliminates the problem described in #2, but is more work and cost, and you might not like the appearance of the optical sensors, which sort of have to be visible. You can combine 2 and 3, and just use the optical detectors around the OS sections and other short sections. In either case it won't protect from a car that wanders out onto the main track by itself. I prefer number 1 and have all cars detectable. Tim Rumph Lancaster, SC |
|
|
|
emrldsky
If you do do this, you will have to make sure that EVERY train you run has am engine and a caboose, and nothing else with resistors, but ALWAYS those two. Otherwise whatever computer program or logic that is used will easily get confused. There are software approaches to help sort out any confusion, but it gets really complicated. Plus Mr Murphy guarantees there will always arise one more situation you have not thought of.
Peace, Mike G.
On 9/2/2020 11:21 AM, Brent Johnson via
groups.io wrote:
just placing resistors on the caboose so the front and rear of train has detection |
|
|
|
Arthur Hammeke
Do leds and lighted coaches confuse block detection
Sent from my T-Mobile 4G LTE Device
From: w4dccqa@groups.io <w4dccqa@groups.io> on behalf of emrldsky <azMikeG@...>
Sent: Wednesday, September 2, 2020 4:36:17 PM To: w4dccqa@groups.io <w4dccqa@groups.io> Subject: Re: [w4dccqa] Block detection/wheel set resistors If you do do this, you will have to make sure that EVERY train you run has am engine and a caboose, and nothing else with resistors, but ALWAYS those two. Otherwise whatever computer program or logic that is used will easily get confused. There are software approaches to help sort out any confusion, but it gets really complicated. Plus Mr Murphy guarantees there will always arise one more situation you have not thought of.
Peace, Mike G.
On 9/2/2020 11:21 AM, Brent Johnson via groups.io wrote:
just placing resistors on the caboose so the front and rear of train has detection |
|
|
|
Allan AE2V
Your system may only get confused if you have a big gap between your locos and your cars that have LEDs or lighted cars. Ideally, block detection is best done with every car drawing some power to activate the block detection.
Allan Gartner Wiring for DCC |
|
|
|
Brent Johnson
So to summarize the best recommended practices for reliable block detection suggested in this thread:
toggle quoted message
Show quoted text
1. Every car should be equipped with a resistor or something to draw power. 2. Isolate every spur track from the signal detection bus to avoid spotted cars from being detected. Does this pretty much sum it up? Brent Johnson On Sep 2, 2020, at 6:41 PM, wirefordcc <bigboy@...> wrote:
|
|
|
|
AD
Why nothing else with resistors? What if the car in middle of train also has resistor wheels wont that just help detect the train?
toggle quoted message
Show quoted text
Tony On Sep 2, 2020, at 5:36 PM, emrldsky <azMikeG@...> wrote:
|
|
|
|
emrldsky
Depends what you are doing... If you are using
them for start and end of train detection, then what does the
one in the middle say? On 9/2/2020 3:38 PM, AD wrote:
Why nothing else with resistors? What if the car in middle of train also has resistor wheels wont that just help detect the train? |
|
|
|
AD
It says, when you are running on automatic and one train is following a first and somehow that wheel on the caboose loses detection you have a backup contact to delay the restart of the second train and lower the probability of a crash
toggle quoted message
Show quoted text
Tony On Sep 2, 2020, at 7:48 PM, emrldsky <azMikeG@...> wrote:
|
|
|
|
whmvd
Brent, Yes for number 1. But I would not call what you describe for number 2 best practice. By a long way. Ideally, there should not be any undetected trackwork at all. A spur should be at least one block. A turnout also. Yes, that is a lot of blocks. And a lot of detectors. And a lot of electronics to get the status for each block back into the command station. If you have even the vaguest idea of wanting to use signals at any time in the future, do not skimp on blocks now. At the very least, ensure the block boundaries exist, each with a separate feeder. That's a good first stage whichever path you choose later. After all, you can simply connect them under the layout until you want to progress to signalling. But if you don't do it now, the job will be a lot bigger later on. Good luck Wouter
|
|
|
|
On Thu, Sep 3, 2020 at 05:08 AM, whmvd wrote:
My layout will include ABS signals (already has some on the parts I've built). For that there is no need for detection on anything but main tracks. If you want to monitor every track, which you might do if you're planning for remote ops, I can see doing what Wouter suggests. I use the LCC stuff made by RR-CirKits. If you want to be ready to implement 100% detection, go ahead and add the current transformers to the leads to the spurs and solder the twisted pair wires to the transformers with enough lead to reach wherever you might put the detector. Better yet, make up a bunch of CT coils with 10 feet of twisted pair wire on the workbench The CT coils from RR-CirKits are $1.70 each (I'm sure you can find them cheaper), and the twisted pair wire (stripped out of CAT5 cable) is next to nothing. Add the twisted pair to the coils now so you don't need to try to do it later, when it will be much more difficult to get at. Finally, label everything. Your layout does not look the same on the bottom as it does on the top. :) |
|
|
|
Brent Johnson
Wouter, Brent |
|
|
|
dsabourne
For my layout planning, I am planning on using optic detectors from Azatrax ... linked back into my Digitrax LocoNet. My issue with detector is the lack of commercial available resistor wheelsets for my non-electric rolling stock.
Regards, David Bourne |
|
|
|
whmvd
Hi Brent, Only with complete detection do you stand a chance of JMRI (or any automation) to keep track of where which train is, and to have working signalling. The thing to think about is not whether you want that today (clearly not), but whether you might want automatic running/signalling to grow more complex and complete later. If so, it's best to prepare. Second thoughts later will mean a lot of extra work which would in the early stages be very little. Your railway - your rules, your choices. As long as the choices are made consciously, all's good. Wouter
|
|
|
|
emrldsky
No, it says when you are running two trains, instead of the expected 4 loads, you now have 3 to deal with. So how does that happen? How do you know if the missing load is from the first or second train? Peace, Mike G.
On 9/2/2020 6:18 PM, AD wrote:
It says, when you are running on automatic and one train is following a first and somehow that wheel on the caboose loses detection you have a backup contact to delay the restart of the second train and lower the probability of a crash |
|
|
|
Puckdropper
You can get away with just detecting the start and end of the train provided the blocks are longer than the train. I really cheated on a layout by only detecting the locomotive, but I was doing grade crossings and the typical train length meant the last car was past the crossing by the time the locomotive exited the detection area.
If you don't want spurs and other than main tracks detected, you'll have to isolate them. You can run them off of the same circuit breaker as long as you branch off before you get to the detection point. (The detector or coil.). You can even run a 3 wire buss if you'd like, one common, one detected, and one not detected. Puckdropper |
|
|
|
Hi Gang: If you are building a layout, I might suggest dividing into more "blocks" than you can possibly imagine. Electrically you can connect several to make a large block. And if you wish to add a block later, well the gaps are already there and you just need to move a few wires. This is from my experience, I rewired my layout three times before the blocks worked right. So if I had put the gaps in during construction it would have been much easier. Carl. On 9/3/2020 6:01 PM, Puckdropper via
groups.io wrote:
You can get away with just detecting the start and end of the train provided the blocks are longer than the train. I really cheated on a layout by only detecting the locomotive, but I was doing grade crossings and the typical train length meant the last car was past the crossing by the time the locomotive exited the detection area. If you don't want spurs and other than main tracks detected, you'll have to isolate them. You can run them off of the same circuit breaker as long as you branch off before you get to the detection point. (The detector or coil.). You can even run a 3 wire buss if you'd like, one common, one detected, and one not detected. Puckdropper |
|
|
|
Don Vollrath
Bill Wilkin did you find the root cause of the detection problem with consisted locos?
DonV |
|
|
|
Jerry Michels
Brent, you don't have to have any particular block report occupancy. The resistive wheelsets are just a means to an end. If there is no detector connected to the block, it won't be detected. Jerry Michels |
|
|