LokSound 4 and Function 0 headlights under RailMaster

[Loxbourne]

Can anyone help me with a weird issue with RailMaster and Function 0 on a loco with LokSound 4?

The loco's headlights do not operate most of the time - triggering F0 does nothing. I have tested them electrically and they work fine. They work normally when the decoder is removed and the blanking plate reinstalled. The model runs happily and the onboard sound works.

However! On a fresh reset of the decoder, when the loco is on ID 3 and ONLY when it has ID 3...the headlights work. As soon as I change it to any number other than 3, the lights go out.

LokSound 4's manual says that if F0 does not work, "make sure the controller is set to give loco #3 28 or 128 speed steps". Is there a way to check how many speed steps RailMaster is using with a loco? I have tried duplicating the loco I keep in ID slot 3 just in case there was something about its' particular settings, without success.

(I am going to test the loco with a different decoder shortly. That will at least provide more data).

[Fishmanoz]

I can see the Loksound logic here. For directional lighting to work, 128 speed steps must be set in CV29. Given a short address, this usually means a value of 2 in CV29 (128 speed steps only) or 6 (128 + DC running).

And how does this cause the problem? While it shouldn’t, there has been previous reporting of changing ID causing spontaneous changing of CV29 value.

A solution - set ID to 03, check lights are working and read value of CV29. Now change ID and re-read CV29. If changed, write the previous value back to it.

[96RAF]

Having set the address on the decoder, the loco needs to be listed in RM loco roster with that address and the functions need to be defined in the function list that opens up when listing a loco. F0 needs a description such as Lights on/off which is a standard in the pull down. If you do not have the on/off appended the lights will go on then off immediately.

[Chrissaf]

I have Loksound V4s and the reply by Fishmanoz is spot on. When you are changing the DCC ID address from 3 (after a reset) to something else, the value of CV29 is being changed (it shouldn't do this, but there have been previous reports that RM does sometimes do this). My directional lights on my Loksound V4s need a valid CV value for working directional lights. I use CV29 = 2 to disable 'DC Operation' along with enabling 'short addresses'. As Fishmanoz has stated, CV29 = 2 or 6 will enable directional lighting, whereas on my Loksound V4s CV29 = 0 disables them.

I predict, that when the lights are working you will find your CV29 equals 6, but gets changed to 0 when they don't after an address change attempt. The factory reset is reverting the DCC Address to 3 and also resetting CV29 back to a value of 6, which is why the 3 address 'reset' works the directional lights when other addresses don't.

Follow the procedural guidance given by Fishmanoz to fix this issue, also note 96RAF's comment about the on/off label on the F0 function button.

I may be wrong but I suspect you are using the green tick button to perform a Decoder CV read function prior to writing the new DCC Address. It is more than likely that by performing this green tick read CV function, RM is picking the incorrect CV29 = 0 value from the decoder database in RM. Then writing 0 to CV29 when you write the new address. When writing a new address, double check the CV29 value in the CV writing screen panel before clicking the 'pen' write CV icon and make sure that RM has not inadvertently set it to 0.

[Loxbourne]

Thanks everyone, RM was indeed writing to CV29. Once I knew what to look for it, it was easy to spot and correct. I will watch out for that on future LokSounds.

[Chrissaf]

Q1 - Was RM populating 0 in CV29 in the CV writing screen?

Q2 - Were you preceding a write address action with a green tick read CVs task?

I only ask as it helps identify a pattern for this issue that will assist others in the future who are reporting the same issue.

[Loxbourne]

Answers as follows:

Q1: No, it was entering a value of 16.

Q2: Yes.

[Chrissaf]

Thank you.

CV29 = 16 was unexpected, but interesting all the same.

[96RAF]

16 = complex speed curve and no other CV29 choice - odd.

[Fishmanoz]

If it’s going to write when it shouldn’t, it may as well make a proper job of it and write something nonsensical. Murphy wouldn’t allow it to write the correct value.