Hi Everyone,
I hope someone knows something about this. I have a fully automated line using contact track sets and semaphore signals all from Marklin using the CS3+ and events. The setup is pretty neat:
Two stations (one at each end with loops to turn around
Station A: 3 tracks with a 3-way c-track switch
Station B: 2 tracks with one of the basic c-track switches
One track between them and a passing track around the midpoint.
The idea is that 5 trains trade off between the two stations. Each station sends out one train and whichever one gets to the passing section first waits for the other one. Once the second one arrives they are both released to continue onto their destinations. When they arrive they park and the next train leaves. I have it programmed using events on my CS3+.
Here are the two problems (and how one has been resolved in a not so good way)

1: I first noticed that at Station A with 3 track the 3-way switch would fail to switch from right back to left like I wanted it to . Sometimes it would switch half way to straight and sometimes it would not switch at all. This caused train accidents. I tried using an old K83 instead of the modern 3-way decoder in the switch and that improved it greatly but it still had to be watched because it would sometimes fail. Next I redid the events to only have two trains at Station A (center track and left track from the 3-way). This solution while removing a lot of the fun of the line worked about 90% of the time until problem 2 arose.

2: I let the line sit for over a month while I was working on a new line and when I came back to it something new was wrong. Before I go further I will mention that I removed all connection/booster/etc to the additional new line completely and this new problem still persists. Everything will be working fine for maybe 5-10 minutes and then a semaphore (changes which one) will only raise half way. From that point on the switches cease to work and other semaphores are hit and miss. I am at a complete loss. I noticed one semaphore was having it happen more than others and it seemed it was having a harder time raising up so I unplugged it but that did not change the problem. In fact tonight when that happened I had the regular switch at Station B (also controlled by an old K83) stick on trying to switch for several seconds until I could hit the STOP button.

Have any of you had similar problems? Anything I should try? I am sorry for the long post but I really wanted to explain the problem as fully as I could.

Usually yes but sometimes it doesn't change either and quite often I loose control and can no longer manually change a switch, I loose the ability to click on it until I do a STOP and start.

What does the cs3 show as current draw on the system menu?

Your fault sounds to me like the cs3 is operating near the limit of its output capability causing the output stage to get warm and go into current limiting.

Also what power supply do you have, and what power supply is the cs3 configured to use?

Interesting thought. My events are all in response to sensors being activated. Do you think if two happen at the same time (or the second happens while the first one has its tasks being done) it could cause the problem? I tried adding a delay for the 1st problem I described with the switch going to straight, then delay and then going to right. It didn't work though.

Forgot to ask, how do I check what transformer the cs3 is set for?

I have not made any recent event changes. Thank you though Peter!

Email to theses guys. marklindudes@gmail.com. They are Marklin digital reps for the US.

Roger

I just assumed that a semaphore controlled by a CS3 was digital, how stupid of me

And by the way, upon reading the first post again, the semaphore are Märklin, not Viessmann

Per.

Yes?
And so what?
I did just wrote a type example if you use Viessmanns semaphore too.
The problem for the TS are possible wrong set in the CS3.
A good example is to deleted all turnout decoder and digital signals.
Start over again by set decoder into CS3.

Very simple!
A analog Viessmann semaphore draw about 0,7 Amp!
There is difference but the real problem are set decoder m83 in every output and so does for the digital signals too.
Märklins CS3 are complicated to set digital signals and decoders like m83 and m84.
You must keep in the mind to avoid problem with the CS3 when you set system.
I too do have sometimes problem if i want to use other competitors accessories to connect with the m83.
Märklin recommend you to use mfx register digital signals as standard set in the CS3.

It is a rich lesson to know about what needs to be learned about.
TS did possible set CS3 wrong and to fix it you deleted and start over again.
This time to set as standard in the system.
Viessmann or Märklin digital signals or analog...it does matter.

Have you checked the booster(s) for line giving you trouble? The problem may not be related to the data bus at all. I've had issues with boosters where something goes wonky and the booster will overheat. At a certain point loco control is lost - they keep running at the last digital setting and I wont be able to operate turnouts. Next time you have the problem check the booster temperature on SYSTEM>BOOSTER for abnormally high readings before you reset anything. My problem seems to be triggered by too many inductive loads on the rails such as smoke units and incandescent bulbs. My boosters usually run around 35-45ºC. When trouble starts I have seen the temp climb as high a 75ºC before I hit the stop which will at least briefly clear whatever was going on.

OK, So I have been thinking about what everyone has been saying and so I tried something new. I disconnected my other line completely. I also disconnected by bus box that is used for boosters and my L88. I plugged my L88 directly into the CS3+. No boosters were on the line (it isn't a very big line). Most of the time just two trains are running with a third running for maybe 3 seconds as they trade off. It doesn't seem to me that there is much power consumption there. Do you disagree? The same thing happened about 5 minutes into running. For a minute I thought the problem had actually been solved but nope:{. Any ideas?

Hi to all of you!
This problem about the delay with the sensors to activate an event its familiar with my layout. I run 4 trains (2 same direction inner circle, 2 outer circle), all of them trigger about 10 track contacts and magnet sensors. I notice that if they trigger 2 or 3 sensors at same time, one of them won't work and not activate the event.
I changed the settings on L88 to the highest milliseconds and i notice some improvement, but problem remains

Yes i know that. Problem is that s88 won't triggered to start the event.

No it's not. But only if to many trains running on layout and activating more than one s88 contact. And that happens very rare but create the problem.
All the s88 contacts works perfectly if one train alone run over them

And that makes any difference?

I'm very confused 😕😕
Thanks for the help although