santa fe cajon-ish 1947 in a Kansas basement

[kc9jts]

Looking good.  Any plans to add a working signal system? 

[draskouasshat]

Being a signal construction foreman for BNSF, that will most definitely be happening! I'm still trying to decide on what occupancy detectors. I've actually been experimenting with my own track relay system for block detection. I do have the relay schematics drawn up for the proper aspects on cajon  pass but i need the block detection. I could, and might have to,  go worry really available options but I'd like to make everything myself.

Adam Draskovich

[jereising]

If you are DCC I can recommend the Digitrax BDL168 - reason being it interfaces with LocoNet, and once a signal is on LocoNet you can do some interesting things...

[draskouasshat]

All I'll need is the block detection. After that,  my whole signal system will be designed and wired by me with proper types of locking and all the goodies the real deal had in that era.
I plan to keep the dcc pretty minimalistic. No fancy computers or anything. Just enough to control locomotives. All switches that were hand throw,  will be hand throws on the layout.

Adam Draskovich

[kc9jts]

All I'll need is the block detection. After that,  my whole signal system will be designed and wired by me with proper types of locking and all the goodies the real deal had in that era.
I plan to keep the dcc pretty minimalistic. No fancy computers or anything. Just enough to control locomotives. All switches that were hand throw,  will be hand throws on the layout.

Adam Draskovich

When you previously mentioned that you had the relay schematics drawn up were you able to get a hold of old plans or just how it "should have" worked based on the aspects?  Just curious as I am signals myself.

[draskouasshat]

Mainly just experience with regular abs. I've been through enough prints to draw everything up with how it should work.
San Bernardino will be a tad tougher but the rest of the pass is just typical abs.

Adam Draskovich

[Cajonpassfan]

Mainly just experience with regular abs. I've been through enough prints to draw everything up with how it should work.
San Bernardino will be a tad tougher but the rest of the pass is just typical abs.

Adam Draskovich

Adam, if you're modeling 1947, it's before System 2 came along, so on the uphill track, there's no red; just green, yellow over yellow, and yellow, and the second head could only display yellow (or be dark).
Not quite the typical ABS as I understand it, but I can always be corrected...
Otto K.

[draskouasshat]

You are correct otto. The way i look at typical abs is in relation to the relay logic and how it's wired. Not 2 block 3 aspect,  3 block 4 aspect, typical indication clear, and so on line most do.

 Basically, how everything works is the same in regards to the relays , they just didn't wire for a red uphill and used that "ouput" if you will to light the bottom yellow.

[Cajonpassfan]

Well, using proto-based relay logic certainly is an interesting approach, but I get why you want to do it that way.
I'm having enough of a challenge with components made for a model railroad

For whatever it's worth, I'm using NCE BD20 coils for detection. I like the fact there is no electrical interconnect between the detection circuits and track power; the coils read presence or absence of current. They can be wired to operate relays with external power supply of 5-20V, something that might be useful to you.

These are getting installed before scenery while I still can get to them. For actual signaling, I'm leaning toward MCS Master Signal Controller, but that's still a while out.
Have fun!
Otto

[kc9jts]

You are correct otto. The way i look at typical abs is in relation to the relay logic and how it's wired. Not 2 block 3 aspect,  3 block 4 aspect, typical indication clear, and so on line most do.

 Basically, how everything works is the same in regards to the relays , they just didn't wire for a red uphill and used that "ouput" if you will to light the bottom yellow.

Not being too familiar with old ATSF standards (I work for a yellow railroad) was the Y/Y as described above essentially the same as a red signal with a "G" marker for grade?

[Cajonpassfan]

No, according the the 1930 rulebook in effect prior to 1948:

Y/Y = Medium speed, approach next signal prepared to proceed at restricted speed.
Single Y = Restricting, proceed at restricted speed *(prepared to stop).

(The ATSF did implement Red with a G plate as a restricting grade indication later, in 1953).
Otto K.

[draskouasshat]

Thank you Otto! I might just have to go with those! I think they're exactly what i need! Plus i like what nce puts out. I also like how simple they look over some of the other fancy offerings. Those will be my choice due to simplicity. Have they been accurate using just a current loop?

Kc, essentially, the yellow was treated as a red in non-ctc abs territory by today's standard while the y/y was just a normal yellow.
They didn't want to display red on the climb is what I've read in my books although that did change as otto mentioned. As per abs, red was a stop and proceed at restricted speed. They just didn't want any stopping in fear of stalls on the hill. There are a couple good pictures I've seen where the rear of a freight is within 3 tenths of a UP train coming up behind it between pine lodge and alray.

Adam Draskovich

[Cajonpassfan]

Yes, and to add to the excitement, between 1948 and 53, the time frame I'm modeling, there was a "restricting" R/Y
As to the sensitivity of the BD20 Adam, I can't say, don't have it wired to anything yet, but a friend does and it works well. The number of times you wrap the feed through the coil has an effect on its sensitivity to low current.
Otto

[GaryHinshaw]

My experience with the reliability of current detection is mixed.  (I am using a different brand than Otto -- the "Watchman" by RR-CirKits* -- but the basic detection coil is the same as the BD20.)   I find that locos are detected with essentially 100% reliability, but freight cars equipped with one 10K resistor per car are detected about 50% of the time, mainly dependent on how good the wheel/rail contact is.  For mainline trains with many equipped cars, the reliability is very high, even if the locos are in a different block than the tail of the train.  But single cars in a block can be problematic.  FWIW, the Watchman has many settable parameters in its DCC decoder: e.g. current threshold and timeout duration (how long the occupancy is held after the detector is triggered).  The latter adds reliability at the (slight) cost of inertia.  I would guess the BD20 has similar options.

Nice progress on your layout!

*P.S. If I understand the pricing correctly, the Watchman is quite a bit cheaper than the BD20: the board can manage 8 blocks from 8 remote coils at a cost of ~$59, or $7.40 per block.  The down-side is that the logic is centralized, so you need to run CAT5 wire to each coil, and the runs can get long.  I have one coil that is 20' from its board.  It works just fine, but it is extra wiring compared to the BD20.

[draskouasshat]

Thank you Gary. Ill look into those also. Ill be happy add long as i have a relay output. Id prefer something like the BD20 that would allow separate detectors, one at each signal location,  versus 8 at a centralized location.

Adam Draskovich