Trying To Fix Fake Red 06 Signals

boleyd

Well-known member
I have always been plagued with fake type-06 red signals. Delays over a minute setting at a 06-red are common. No other trains in operation. The only "guess" seems to be that route adjustment computation is taking the delay time. The relevance of 1.5 minutes to computation of a route is not reasonable. Now some stops on last 20-30 seconds which offers a pretense of reality - temporary occupancy somewhere on the line (but no other trains on the system).

However, since that is the only offered hypothesis I deleted the two, concurrently running, virus checkers programs an installed a copy of Norton (free from Comcast). No real change. Next I decided to make all 06 signals Permissive. That works. The train runs right through the red since the next block is ALWAYS clear - 1 operating train. Unreal, but "the trains run on time".
 
Last edited:
There has to be something else going on here to cause that. Can you show a screen capture or schematic of the route?

The message says exactly what it means. There's got to be something on the route that's occupying the block and preventing the signal from changing to another aspect until the block clears. A permissive signal is just that and will allow trains to pass the red because it's indicating that the next signal will be at stop. These are also referred to as advance signals or distant signals, thus, the "d" in the name if you are using Jointed Rail's SL(6d) series signals.

When all is setup correctly, things generally work as intended. There is an issue where the signaling logic loses steam after some time with some routes being better than others in this regard. This I was told was due to an inordinate amount of messages clogging the message queues. I have had routes work for about an hour before the signals forget their aspects and don't change states. The result is AI drivers sitting there not moving anywhere for long periods of time before they start doing what they're supposed to. A symptom of this being the issue is a simple test. Press pause, wait a minute, then unpause. If the AI resume driving normally, then it's the message queue problem. This is a known issue which will require a rewrite of the AI logic and reworking of the messaging queues I was told. Saying that, I don't expect anything any time soon unless N3V comes out with a new version.
 
SOLVED

Hopefully solved. I just needed more signals. Long unsignaled sections cause the 06 signal to set red and hold the train. A section of track that has a somewhat shorter, but unacceptable unsignaled section, stops the train for 10 to 20 seconds only. This seems to indicate that the unsignaled section can be checked within the maximum AI limit for computation. Fixed by adding 1 set of 04 signals on the two parallel tracks I was testing. In this case I was using 06 and 04 searchlight signals. The efficacy (Fauchi term) of this "fix" in other routes needs checked.

This resolution could be route specific and not a solution. Time will tell.

Is there a published max amount of unsignaled track that is permitted?
 
Last edited:
. . .

Is there a published max amount of unsignaled track that is permitted?

The DHR runs 54 miles of unsignaled track. There are no two-track segments except for a couple very short segments at stations. AI runs just fine on it.

Bill
 
This could be some combination of track, "brand of signals", train makeup, something else. Tomorrow it may fail again. All the usual caveats have to apply. N3V is a stew of change. Ideally N3v would get the funding for a 6 month audit to clean things that have been injected, or deleted, over the years. But, I will be happy if my issue remains in the Fixed category.
 
This could be some combination of track, "brand of signals", train makeup, something else. Tomorrow it may fail again. All the usual caveats have to apply. N3V is a stew of change. Ideally N3v would get the funding for a 6 month audit to clean things that have been injected, or deleted, over the years. But, I will be happy if my issue remains in the Fixed category.

The operation is pretty consistent for me. I would chock your fix up as resolved.

For extra long tracks, I insert permissive signals in between to allow consists to follow each other and keep traffic flowing. I couple 2 of the 2000T US Coal consists together and use those as my ruler for spacing out the Type 05 signals.
 
Back
Top