|
Query from a Nimble Newbie,
How many commands can one send at a time from an "if-then" Logix statement? E.g. IF sensor 'x' goes 'active' THEN....switch Signal '1' to danger, signal '2' to caution, signal 3 to 'clear', turn on coffee pot, etc...
It seems to THEN part of the statement fires odd the commands so quickly that the NCE system misses some of the actions [somewhat randomly]. Not a wiring issue.
Is there a way of slowing down, or at least separating out [even by milliseconds] the list of ACTIONS to occur upon a given antecedent?
Nimble of Oz
|
Been there, done that. Light-it's can not keep up with the commands. I wired up a bunch of them and ended up ripping them all out and replacing it with RRCircuits Signalman product. RRCircuits performance has been 100% over 2 years now even with complex logix. I actually documented all this on my youtube channel linked below if youd like to check it out.?
?
Lou
?
|
And, for what it's worth, I tried slowing down the commands and even added a line of code to resend commands automatically. It seemed as if the light-it was just overwhelmed. It was an ugly few months for me tracing, retracing, and retracing logix over and over before I ripped them out and replaced with signalman.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Hi Lou,
Thank you for your honest reply, much appreciated. I feared
something was amiss as I too have now wasted many hours trying to
get the NCE Light-its to work reliably. Very frustrating, and
something which should be brought to the attention of good folk before
forking out $$$$'s.
I'll have a watch of your YouTube channel while trying to recover
my frazzled nerves.
Should you ever come to Sydney Australia, you now have free
accommodation and meals. Just email me on this channel.
Now I think I'll contact NCE...not they they will like it.
Regards,
Allan B [aka Nimble], St Clair, Sydney, Australia
On 1/05/2025 6:41 am, louis.dirosso via
groups.io wrote:
toggle quoted message
Show quoted text
And, for what it's worth, I tried slowing down the commands
and even added a line of code to resend commands automatically.
It seemed as if the light-it was just overwhelmed. It was an
ugly few months for me tracing, retracing, and retracing logix
over and over before I ripped them out and replaced with
signalman.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Thank you for the invitation!
before you go chewing out NCE, try to remember that the signal decoder are built without any sort of compatibility or NMRA standard to meet.?
?
The light-it decoder is intended to be used with a handheld NCE throttle and when it is, it works well. at no point was it designed to interface with JMRI and I would imagine not designed to accept the commands as JMRI. When we start interfacing third-party hardware with other third-party hardware and it doesn’t work, as frustrating as it is, we should not blame the manufacturer. We should blame the standard.?
I held onto my light, it decoders and will use them as accessory lighting for buildings streets, etc.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Am I looking at this too simply if I think that it should be possible to write some not-too-difficult spacing code that would delay commands to be sent to light-its too quickly one after the other? And on this point: is here information on what the minimum time between commands is in order for a decoder not to choke on it?
Wouter
On Fri, 2 May 2025 at 14:11, louis.dirosso via <louis.dirosso= [email protected]> wrote: Thank you for the invitation!
before you go chewing out NCE, try to remember that the signal decoder are built without any sort of compatibility or NMRA standard to meet.?
?
The light-it decoder is intended to be used with a handheld NCE throttle and when it is, it works well. at no point was it designed to interface with JMRI and I would imagine not designed to accept the commands as JMRI. When we start interfacing third-party hardware with other third-party hardware and it doesn’t work, as frustrating as it is, we should not blame the manufacturer. We should blame the standard.?
I held onto my light, it decoders and will use them as accessory lighting for buildings streets, etc.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Lou,
The Light-It does support the NMRA Signal Decoder standard. ?This is supported in JMRI by using the "DCC Signal Decoder" option for the signal heads. ?You do need to modify some of the appearance numbers if using appearances other than red, yellow and green.
I suspect the problem might occur when using the signal mast "Signal Head Controlled Mast" option to combine multiple signal heads. ?For example, changing a two head mast aspect from Clear to Diverging Clear requires two signal head appearance commands. ?
Dave Sand
----- Original message ----- Subject: Re: [jmriusers] NCE ProCab- Light-it as signal decoder - Simultaneous commands #nce Date: Friday, May 02, 2025 8:10 AM
Thank you for the invitation! before you go chewing out NCE, try to remember that the signal decoder are built without any sort of compatibility or NMRA standard to meet.? ? The light-it decoder is intended to be used with a handheld NCE throttle and when it is, it works well. at no point was it designed to interface with JMRI and I would imagine not designed to accept the commands as JMRI. When we start interfacing third-party hardware with other third-party hardware and it doesn’t work, as frustrating as it is, we should not blame the manufacturer. We should blame the standard.?
I held onto my light, it decoders and will use them as accessory lighting for buildings streets, etc. ? Lou -- Modeling Conrail’s Boston Line in HO
|
Buffering commands to handle fast incoming traffic while dealing with an operation that is currently underway is basic programming 101. I am surprised that this hasn't been done in these decoders. It isn't as though it is a hard problem to solve, just a little extra programming of some basic operations.
On Fri, 2 May 2025 at 14:11, louis.dirosso via <louis.dirosso= [email protected]> wrote: Thank you for the invitation!
before you go chewing out NCE, try to remember that the signal decoder are built without any sort of compatibility or NMRA standard to meet.?
?
The light-it decoder is intended to be used with a handheld NCE throttle and when it is, it works well. at no point was it designed to interface with JMRI and I would imagine not designed to accept the commands as JMRI. When we start interfacing third-party hardware with other third-party hardware and it doesn’t work, as frustrating as it is, we should not blame the manufacturer. We should blame the standard.?
I held onto my light, it decoders and will use them as accessory lighting for buildings streets, etc.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
My question to users of the Light-It, which method did you configure the decoder to use? ? If the choice was Signal Address Decoder, I would not expect any single Light-It to get more than one command for quite a while. Each should be able to deal with that without concern what other addresses might be getting commands. ? But if the choice was as loco decoder or accessory decoder, I could see some interference between the multiple commands for a combination of features. ? Which have people been using when finding issues with the Light-It? ?
|
Hi Lou,
The idea of using the NCE decoders to light buildings, even some
street lamps etc... is something I am thinking about. I won't
actually bin them...yet! And? you're right in saying they work
okay via NCE throttles; and yes they do work from JMRI , but not
without the issue[s] we've discussed. It would be nice if'n I
known about these limitations before purchasing the hardware. NCE
don't tell you, nor their supplier here in Sydney.
I guess my next questions is, "How much NCE hardware now becomes
redundant [as far as signalling goes], if I use another product
decoder?" For example, will any of the RRCirKits signal drivers
simply replace the NCE ones and just hook-up onto the NCE track
bus? If so, then I would be able to keep my NCE AIUs and the
myriad of sensors [NCE BD20s] I have installed. OR...do I need to
set up an entirely new and separate BUS for to carry the signal
logic, etc... [An LLC control bus, for example] and would any of
the NCE items still work with that?
Perhaps the answer are in you YouTube posting, which I will view
shortly.
Thanks again for responding and helping me regain my composure.?
Nimble.
-------------------------------
On 2/05/2025 11:10 pm, louis.dirosso
via groups.io wrote:
toggle quoted message
Show quoted text
Thank you for the invitation!
before you go chewing out NCE, try to remember that the
signal decoder are built without any sort of compatibility or
NMRA standard to meet.?
?
The light-it decoder is intended to be used with a handheld
NCE throttle and when it is, it works well. at no point was it
designed to interface with JMRI and I would imagine not designed
to accept the commands as JMRI. When we start interfacing
third-party hardware with other third-party hardware and it
doesn’t work, as
frustrating as it is, we should not blame the manufacturer. We
should blame the standard.?
I held onto
my light, it decoders and will use them as accessory lighting
for buildings streets, etc.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Hi Dave, Lou,
Dave, I understand what you are saying, but the problem is still
evident even though I am using the "DCC Signal Decoder" option for
the signal heads and just trying to employ some simple logix
commands. I did try using the signal mast "Signal Head Controlled
Mast" option, amongst others and always got to the same point,
that is, signals would miss updates - somewhat randomly. Could be
any of three aspects.
I could always update any given signal manually from the JMRI
interface on my PC screen, no problem. But getting the signals to
update absolutely reliably on the layout via the sensors/AIUs as
the train moved around...nope! The JMRI screen would update and
tell me "Signal 'x' is green' but on the layout it might have
stayed at 'caution' or 'danger'.
I also noted that the more signal I added, the more likely I was
to get update failures [not always the same signal either]. So
random.
Nimble
--------------------
On 3/05/2025 1:13 am, Dave Sand wrote:
toggle quoted message
Show quoted text
Lou,
The Light-It does support the NMRA Signal Decoder standard.
?This is supported in JMRI by using the "DCC Signal Decoder"
option for the signal heads. ?You do need to modify some of the
appearance numbers if using appearances other than red, yellow
and green.
I suspect the problem might occur when using the signal mast
"Signal Head Controlled Mast" option to combine multiple signal
heads. ?For example, changing a two head mast aspect from Clear
to Diverging Clear requires two signal head appearance commands.
?
Dave Sand
----- Original message -----
Subject: Re: [jmriusers] NCE ProCab- Light-it as signal
decoder - Simultaneous commands #nce
Date: Friday, May 02, 2025 8:10 AM
Thank you for the invitation!
before you go chewing out NCE, try to remember that the
signal decoder are built without any sort of compatibility or
NMRA standard to meet.?
?
The light-it decoder is intended to be used with a handheld
NCE throttle and when it is, it works well. at no point was it
designed to interface with JMRI and I would imagine not
designed to accept the commands as JMRI. When we start
interfacing third-party hardware with other third-party
hardware and it doesn’t work, as frustrating
as it is, we should not blame the manufacturer. We should
blame the standard.?
I held onto my light, it decoders and will
use them as accessory lighting for buildings streets, etc.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Nimbles reply...I tried a few JMRI signal profiles, none proved
any better than any others. The Light-its are using the Signal
Decoder subset of addressing, as outlined in the NCE documentation
and what JMRI lists as "check the offset box"? for each signal.
---------------------------
On 3/05/2025 4:06 am, Ken Cameron
wrote:
toggle quoted message
Show quoted text
My question to users of the Light-It, which
method did you configure the decoder to use?
?
If the choice was Signal Address Decoder, I
would not expect any single Light-It to get more than one
command for quite a while. Each should be able to deal with
that without concern what other addresses might be getting
commands.
?
But if the choice was as loco decoder or
accessory decoder, I could see some interference between the
multiple commands for a combination of features.
?
Which have people been using when finding
issues with the Light-It?
?
|
Greetings Lou,
I have just watched several of your YouTube videos. Thanks for
taking the time to produce these helpful vids.
Just a couple of things I'd like to clarify regarding the
signalling system you have settles on:
1. Am I to understand that you found BOTH the NCE Light-it &
the RRCirKits HD2 to suffer the same/similar issue of "not being
able to keep up?"
2. Thus you went to the RRCirKits "Signalman" decoder because it
does the job properly, even though it be percieved by the NCE
system to be a "loco" and addressed as such? [Thus abandoning the
use of the separate "signal addressing" system.]
3. Can you confirm if it is the NCE AIU channels that "trip" the
signals and inform JMRI of the status of things [including turnout
feedback, which I am using]?
Any clarifications by you [or others] would be most appreciated.
Allan B [Nimble],
Land of Oz
--------------------------------------------
On 1/05/2025 6:41 am, louis.dirosso via
groups.io wrote:
toggle quoted message
Show quoted text
And, for what it's worth, I tried slowing down the commands
and even added a line of code to resend commands automatically.
It seemed as if the light-it was just overwhelmed. It was an
ugly few months for me tracing, retracing, and retracing logix
over and over before I ripped them out and replaced with
signalman.
?
Lou
--
Modeling Conrail’s Boston Line in HO
|
Nimble, ? If you plan on keeping JMRI present in your layout operations, you can use it to gateway any other device to work with other devices. If your signal logic stays in JMRI it is very seamless. ? Now anything that works with an AIU could also be moved onto the LCC or other RR-CirKits nodes. Just another choice and you could defer doing that to later. ? -Ken Cameron, Member JMRI Dev Team
? ?
|
Nimble, ? I recall there is a command station option that will cause a signal decoder message to be repeated on the track bus a couple of times. I could see if that was set to 1 as a default, changing it up a bit might be the cure. I see the default for repeating as this in the PH5: ? NUMBER OF ACCESSORY PACKETS This adjusts the number of accessory control commands sent when an accessory or signal is controlled. Factory default is 4. Press ENTER to skip setting this parameter. ? That should be reasonable. This is under the Packet Management Setup part of the command station configuration. But maybe other values are worth trying. ? -Ken Cameron, Member JMRI Dev Team
? ?
|
1. Am I to understand that you found BOTH the NCE Light-it & the RRCirKits HD2 to suffer the same/similar issue of "not being able to keep up?"
?
Correct. The SD2 was better, but still has flaws. I imagine this is why RR circuits dropped it and moved forward with the signalman.
2. Thus you went to the RRCirKits "Signalman" decoder because it does the job properly, even though it be percieved by the NCE system to be a "loco" and addressed as such? [Thus abandoning the use of the separate "signal addressing" system.]
?
Correct
3. Can you confirm if it is the NCE AIU channels that "trip" the signals and inform JMRI of the status of things [including turnout feedback, which I am using]?
?
No, the AIU is used for sensor feedback and switch feedback, if wired to do so. Each block sensor is fed to the AIU to trigger one of the AIU’s 16 channels (actually 14 because 2 channels are too sensitive unless it’s turnout feedback).
?
Turnout feedback can be handled internally on the NCE command station. When you program each switch on JMRI, select “monitoring” as feedback and the NCE command station tells JMRI the switch status.?
However, if you need local push button control of a turnout, then that turnout needs feedback to the AIU. If you use a push button off say, a Switch-it, the push button is “local” to the switchit and does not send a command to the command station. So the switch is thrown, but your command station doesn’t know it. Hence why you need feedback to the AIU.?
Personally, I ditched feedback and push buttons when I went to CTC and eliminated the extra bit of wiring. It’s much easier to have say, a Tablet mounted as a panel and control switches on it.
?
Dave helped me with the packet stuff last time. Nothing we did helped the lightit. Learn from my mistakes.
--
Modeling Conrail’s Boston Line in HO
|
Louis,?
On May 3, 2025, at 9:04?AM, louis.dirosso via groups.io <louis.dirosso@...> wrote: [snippet]
No, the AIU is used for sensor feedback and switch feedback, if wired to do so. Each block sensor is fed to the AIU to trigger one of the AIU’s 16 channels (actually 14 because 2 channels are too sensitive unless it’s turnout feedback).
Not to pick but a clarification. ?12 of the 14 AIU channels have good sensitivity except for #6 & #7, for which I created a modification to improve their sensitivity. I posted the fix on the NCE group.
I know this is outside the scope of your message to Allen (Nimble) but thought the info could be useful for someone.
John ?Bauchiero -?NCE PH-Pro & PowerCab, Pi4, DCC-EX - JMRI 5.11.5 - Java 17.0.12
|
Thanks John. I wasn’t aware of the fix and I hadn’t had my morning coffee yet on the channel count LOL
--
Modeling Conrail’s Boston Line in HO
|
I agree,? I did the same all light-its are gone. Signalman in their place.
?
Light-its good if you know the command sequence will be simple single command based on occupation or turnout
|
Nimble......
?
Signalman uses signal commands to set aspects.? ?It uses a loco address ONLY for programming purposes.
?
You MUST set the address on an isolated track such as an off layout program track.
?
EVERY signalman is ALWAYS listening for address 9999.? If you try to program the address of a new signalman on your layout using 9999 you will reprogram EVERY signalman.? But once the new address is programmed it will work like a charm using POM.
?
John Crellin
Bradenton, FL
|
Ken Cameron
John Bauchiero