SC1 and Digitrax Signals

[markf]

Hi there,

I have an SC1, which I REALLY like the idea of.  If my tests prove-out, I expect to buy a number of other SC1s for my modular layout.  SC1 is ideal for modules as you don't need a whole load of signals on each module.

I'm using a Digitrax system (DCS100), with JMRI in support.

I had issues configuring the SC1, but caught a really useful thread here explaining the best way to configure the address was to use the program jumpers, and that has worked a treat.

I intend to use a mixture of bespoke signals (wired to the terminals on the SC1), and some of the Digitrax Signal Masts (using the 10-way ribbon and connectors).
In my "bench test", I'm using the Digitrax SH2AB kit.  I have an "A" and a "B" signal attached to a piece of 10-way ribbon with the connectors on it.  Both the signals are single head with R/Y/G LEDs.  When I configure the SC1 fully manually, I can get the signals to turn on and off each of the 6 LEDs (3 on each Digitrax signal mast), through both the Digitrax hand-held and the JMRI software.  All good so-far.

But as soon as I try to configure the signals for CV38=7 I just get really odd results, with various colours flashing and unable to select "dark" at-all. Same with CV38-11.
I have no idea why the results are seemingly inconsistent or spurious.  Even manually setting the signals as "closed" or "throw" according to table 6 in the SC1 manual seems to fail to get the expected results on the signals themselves. 

Ideally I'd really like to get one of the extended protocol configurations going, but I can't go there if the basic CV38=7 doesn't get the desired results.
I also am skeptical that my DCS100 will support extended protocol as it's a very early Digitrax unit (bought in the US before they were even available here in the UK). I can see what looks like extended protocol code when I monitor Loconet using the JMRI, but have elected to try to solve one issue at a time.

Is there any spacial configuraiton needed for these Digitrax signals please?

Thanks in advance


 

[markf]

OK. Managed to work through and will self-comment here to help others.

The set up is Digitrax Dcs100 with Digitrax signals attached to a ribbon cable. I have two three-led Digitrax signals (type A and B), on the same ribbon.

First I set the SC1 base address. Per another thread here, the best way to do this is by using the jumper. I used 005.

Then, I set cv38=8 (4x 3aspect searchlight). I found on Digitrax that the CV programming works great when you use Direct Programming mode (Pd on your throttle). Programs and reads back reliably.

Then I used JMRI to add signal heads.
These heads are "triple output" type in JMRI.
Per the SC1 manual, your first address is Red for the first (A) signal. So my first triple output config looks like this:
Green = LT6
Yellow = LT7
Red = LT5. (base address of my SC1).

My second signal (Digitrax type B), is also a Triple Ouput config in JMRI:
Green = LT10
Yellow = LT11
Red = LT9

Now, at this stage things will look a mess, and the colours WILL NOT correspond to what you set in JMRI!
To address the mess, you need to go into "Turnouts" in JMRI and set LT5 - LT11 to "inverted".
Once you do this, the signals will correctly display the aspects you select in the Head table in JMRI.

You can then go on to create the masts and logic you need in JMRI.

I have not managed to have any success with setting the SC1 to work with extended accessory protocol yet.....

[Paul Harman]

Markf

Try setting CV35=255 to get round having to set all the turnouts to invert. Command stations differ in how they interpret the state codes from JMRI for turnouts.

CV38=128 should be the correct setting for using extended accessory packets.

Paul