Loksound 4 start up variations

[zeroOneman]

I've recently fitted a loksound 4.0 to a Graham Farish 411 CEP EMU. I'm driving the railway using Railmaster & an elink module. Everything appears to be working well, except I have a program I created that shuttles the EMU from one end of a branchline to the other. When it returns to the 'home' station I then make it creep up to the buffers, so that next time the program starts the loco starts from exactly the same position. THis worked fine until I fitted the Loksound decoder. From observation, it looks like that when I start the loco off, the initial sounds,  such as the brake release, are random & as such the time it takes for the loco to actually start moving vary. This makes using any program based on timestamps useless.

Am I imagining the variation in the startup time? Has anyone else come across this?

Dare I mention waiting for loco detection to come to my rescue?

Thanks in advance

Andy

 

 

[Chrissaf]

All my locos are Loksound. Variable start delays are just something I got used to. As I don't use RM programs it is not something I have bothered to look into with regard to CV tinkering. One observation I will offer though and its not necessarily ideal but could be a workaround for you, but if you start the loco off first without F1 sound, the loco starts to move straight away. Once it is moving (even if only slightly), if you then issue an F1 command, the loco engine sound kicks in straight away with the loco on the move, without all the loco start up sounds playing to add a delay.

.

From memory, I think there is a start delay CV in the Loksound decoder manual. But as I said above, not something I've investigated in any depth.

[zeroOneman]

Thanks Chris, I'll look further into the start delay CV - if I make it long enough, then the start should always take the same time. I'll post back with my findings...

[2e0dtoeric]

I have noticed something similar on Bachmann and Hornby 'full' sound decoders (not TTS) at the other end of the driving spectrum - sometimes the brakes squeal, and sometimes they don't, on various loco's, as they are stopping!

[St1ngr4y]

@zeroOneMan

 

Hi Andy,

Do you have notch up and notch down functions on your Loksound decoder, and if so, do you use them during the running of your program? Maybe if you issue a notch down command or two when stationary, it will return the sound to the low end of the sound cycle. Then maybe you can restart at the same place in the cycle every time to give consistent startup sounds. I don't have a EMU with Loksound. Do they make a sound when stationary?

Ray

[Chrissaf]

Yes Ray they do. My Loksound EMU Class 411 sits there stationary with a number of random sounds playing such as compressors kicking in and out and other similar non descript sounds that randomly occur (well at least they seem to be random, but probably have a timed sequence to them).

[Chrissaf]

Strange thing is HFM, five of my Loksound V4 locos have CV124 with value 20 as default. A value of 20 sets Bit 3 as 0 (zero) not 1, yet these locos still have the startup delay enabled. I agree, the manual on page 71 states deleting Bit 3 (I assume this means making Bit 3 = zero) should disable the startup delay, but it doesn't seem to in my decoders. I will be honest and admit that I have not read fully all the way through the manual as there is a lot in it, plus I am quite happy with the default decoder configurations as they are. But this is a conundrum all the same.

.

/media/tinymce_upload/e25cfa7a8b4c2bf78b0d72398bdee655.jpg

.

EDIT: Found the answer on Page 83 of the Loksound manual. The reference to CV124 Bit 3 on page 71 of the Loksound manual is a typo misprint. The startup delay is Bit 2 of CV124 or you could call it the 'third bit' but not Bit 3 as written in the manual (page 71).

.

The evidence for this statement is below, pasted from page 83:

.

/media/tinymce_upload/9377a5f1df950b60ee7bd0b8c55f68dd.jpg

.

CV124 Bit 2 is used to enable / disable 'Startup Delay'.

CV124 Bit 3 is used to enable / disable SUSI (Serial Standard User Interface). CV124 Bit 3 being for SUSI is also confirmed on Page 50 of the manual.

.

Thus, assuming that CV124 has a default value of 20 as per my first table above. Then writing decimal value 16 to CV124 will disable the 'Startup Delay'.

[Chrissaf]

Typo edit to last reply:

"(Serial Standard User Interface)" should have read....

"(Serial User Standard Interface)".

[Peter Ruifrok]

 Hi Andy, I am doing the same as you on my layout. All locs show a minimal variability regarding stopping and I alllow them to go gently against a buffer for the same reason as you do. I have one loc with a Zimo sound decoder and have not noticed any greater variability in stopping with this loc in comparison to other locs without such sound decoder. What I have noticed however is that the loc runs faster without F1 (which enables sound) on. On the other hand "mute" does not make a difference (other than that one does not hear the sound anymore). If you "play" manually with F1 while your loc is running, then that could be the reason for your trouble.

 

Also, if F1 is enabled, it will take longer before the loc starts running. But there is no additional time variability in this.

[2e0dtoeric]

Thanks, HFM, I hadn't bothered reading up or fiddling, as I'm happy with the way they behave out of the box. I was just commenting, not complaining! 😆

[zeroOneman]

OK, I have it working how  I  want it (for now), as mentioned earlier, clearing bit 3 to zero basically decouples the startup sounds from the motor start. The default for this byte is 20 , setting it to 16 clears bit 3.

Now I just need to wait patiently for LD & then I can set it back again.

I was using a Lais decoder before & it's noticeable that now I've cleared bit 3 the loco is stopping in exactly the same spot every time - with the Lais decoder there was more variation (2 or 3 mm) every time the loco moved along the branchline, which has an overall length of about 5 metres.

 

Thanks everyone for your help

[Chrissaf]

Just a correctional point made for clarity just in case other readers are reading your post above at some future date for the same issue.

.

Amending CV124 from 20 to 16 is actually clearing Bit 2 (the decimal 4 value bit) not Bit 3 (which is the decimal 8 value bit). The reference to Bit 3 is a typo error in the Loksound manual. See my post on preceding page for full details.

[zeroOneman]

you're quite right Chris & as a firmware engineer I should know better !

 

I've been wondering about why the loco is stopping in what looks like precisely the same place every time now, which is of course exactly what I was after - I reckon it's either the superior motor control of the Loksound, or because is freezing cold out in my shack & so the motor temp is staying so low that there's no variation due to it warming up. My money is on the superior motor control.

[zeroOneman]

Well the above solution has been working well for nearly a year, but with no sign that LD will be with us any time soon I've been hunting for a better, but cost effective solution. I've come to the conclusion that the Lenz BM1 module will do just what I want - one module at each end of my branchline will stop the train in exactly the right place every time won't they? But then I started thinking about it & it gets better - I have two main loops as well, I've experimented with using railmaster timers to stop & start my trains in the stations & decided if I keep going down that road I'll go mad, But If I fit signals & combine them with a relay to the green light & a BM1 then I can, again stop the trains in exactly the right spot - then I can usse timers to sequence everything without having to worry about variations in the stopping positions. This, to me, looks like the cheapest way of achieving a simple form of automation. I've found RAF96's website & the wiring diagram for a home made bm1 - so tomorrow I'll be hunting through all my spares to see if I can find enough diodes to knock a couple up tomorrow for the branchline.

[96RAF]

I dont know if the Lok decoder supports ADCC but as you are going that route it probably does. If so it will offer options via CV settings for uni or bi directional travel i.e. stop one way or both ways. Also if operating into a bay platform you can reverse out without switching the BM1 out of circuit e.g. no need for signal by pass switch in reverse.

 

Also look on the Lenz ADCC website for intermediate stop also decoder dependant.

 

I have used BM1 with Hornby Sapphire and it works a treat.

Rob

[zeroOneman]

the loksound 4 chips all support 'lenz ABC' so far I've spotted cv's to change the symmetry sensitivity, braking distance for both directions & which side the assymetry should be on. I didn't find any diodes so it'll be after xmas before I get this going - 50 diodes on order from ebay @ 1p each. I can't wait to get this going - it's going to be awesome.

[jane1707819582]

I have noticed something similar on Bachmann and Hornby 'full' sound decoders (not TTS) at the other end of the driving spectrum - sometimes the brakes squeal, and sometimes they don't, on various loco's, as they are stopping!

hi 

the braking is triggered by  cv4:the speed of the loco and then receiving a step 0 command : cv64 and 65 requires tuning to get your desired effect ,