Hello all,

During the recent weeks I have been finishing the automation on my layout, and installed 32 contact tracks. I have made the contact tracks with the well-known method by cutting the electrical connection between the rails underneath the C-track, isolating one side of the rail from adjacent track pieces, and connecting them all to s88's (or Viessmann 5217's actually).

While doing this I have noticed one potential problem. Since now only one side of the rail provides the ground contact when the loco is on a contact track, it may happen that the loco momentarily looses the ground and gets jerky or even stops abruptly. How sensitive it is, depends on many things like if it is straight or curve piece, if the track is banking on either side a little, etc. I often needed to try and isolate one side first and try if the loco gets through smoothly, and if not, then isolate the other side and try again. Usually one side worked well, but you never knew which one without trying.

Have any one of you experienced the same with contact tracks, and do you know any "tricks" to improve the contact that the loco makes to the remaining grounding rail. The cause of the problem seems to be that if the loco leans even a little bit to one side, it looses contact to the rail of the opposite side. And if this is now the grounding rail, it looses the ground contact and stops.

Those were the kind of things that I noticed, too. Isolated inside rail in curves is the best bet, but then again I also had one or two occasions where the inside rail connected better! So I had to change and isolate the outer rail instead. But now 31 out of 32 contact tracks seem to work without trouble. There is still one beast that gives me headache, and it is straight! No matter which side I isolate, the loco always momentarily looses contact at some point on that track, and goes like having a hiccup, when it looses the contact and gets it back one millimeter later. That track may be somehow bent, or twisted, so there are always some wheels which are in the air.

Hello Nuno,
the s88 don't have any problems att all having contacts for that time. They draw a minimum of extra current from the s88 bus that's true, and that's all.

/L

I have gone the route of using the modified C-track for feedback also and set it up to monitor sections of track. I have not had any problems yet and been running for about 2 years now and like the concept.

I chose to monitor all sections of track, except for turnouts in order to ensure that all tracks which are ment to be clear for a locomotive to pass through is really clear. It is quite possible that a railcar can detach and sit on the track waiting for a locomotive to crash into it. With entire track being monitored the computer will not release a train into a section of track that is still occupied. This method prooved its self a few times already.

Hi Kimmo, Lars, Nuno, Matthew, all,
Thanks for sharing your experiences. I´m going to use this kind of "massive" monitoring, cause I´m ready to start with PC control. I´ve been using a few contact C-tracks in my shadow station, but now I´ll need more for the new purposes, so my question is: Which length do you recommend the C-contact track must have?

Hi Kimmo,
I have the same problems on my layout which will have about 12 Viessmann 5217 modules (192 contacts). I have already prepared 144 contact tracks and I am in the testing phase.
The symptoms are the same with yours. The biggest problem is with the E40 engine which has two grounding axes, so only two wheels are in contact with the grounding rail. The engine is lighter than B18 which works smoother.

I cleaned the rails very well (using Philips degreaser)and also the wheels. The majority of the problems have gone away, however there are still track sections producing problems especially with E40 and at low speeds. By turning on the engine lights one can see easier where the problems are (flashing lights). I am conducting many tests with the following results (for the time beeing):

- The main problem is between the wheel and the rail
- Almost each time the engine stops, if I connect the two rails (with a screwdrive for example) it starts again showing that the problem is between the wheels and the rail.
- There seems to be a problem also between the axles and the engine ground (occasionally)
- The problems appear on straight lines (I did not test curves much)
- Most of the times the problems appear, is because the engine is moving sideways
- The behaviour of E40 is improving if one removes the rubber rings from one of the axles used for motion (the inner axle produce better results), but then the engine gets very noisy.

Some thoughts:
- Change the inner motion axle with a simple one (like that using for ground) without taking motion from the motor. I don't know what impact will it have on the traction capabilities of the engine
- Add some kind of brush in the middle of the engine, which will be in contact with the rails permanently (optically bad solution).
- Add some kind of horizontal springs to avoid sideways movement.
- Connect the ground through the coupler to the car behind the engine so that the car's wheels will also be used for grounding (but this will mean the engine will always have this car attached to it.
- Maybe my E40 is faulty concerning the ground contact (I hope so)

I am also surprised by the fact that there are not so many people having the same problems. Maybe its something I am doing wrong, but for the last two weeks I couldn't find it.

By the way Kimmo, you say that you solved your problems. How did you do it? Only by cleaning track and using the inside rail for contact tracks?

George

One question about the track sections. Can you create good isolations without contact rails (24995)by using the same isolators as for signals? I bought a set of contact rails and a S88 and it works okay, but the 24995 in combination with the s88 is not very cheap per contact.

The concept is the same. Use one of the two rails for connecting to the s88 module and leave the other one for grounding. See the Workshop11 file at http://www.windigipet.de/download4.html
for a detailed description of making contact tracks by cutting the ground bridge underneath the C track

George

George

With C-track, the only sensible way to create contact tracks is to convert normal track piece into one, unless you want to spend money on something you really don't need. Red Märklin plastic insulators are enough for the isolation. When you cut the bridge between the rails underneath the track, be careful that the loose ends don't touch each other or the center rail. If they touch each other you get constant feedback, and if they touch the center rail you get a short.

As George mentioned, to ensure correct detection in all cases, the section would still need to be longer than the longest distance between two axels. And then you could not use the studs underneath to connect the wire to the rail, but would need to solder it directly to the cut-off rail segment.

Not tried it, but looks a good idea if you want a momentary signal. And it's not visible like the "Dremel" method, so that's an advantage. Suggest a shorter section of 2 studs is probably better than 3 studs.

Again, however, not sure how/if this could work if you need full-length detection.

Clever, didn't think on that. Background is as follows:
the s88 actually indicates occupation if current may flow out from the s88 to the track. This happens if there is a connection, AND if the voltage potential is less than about 0.5 V. So, grounding does the job, and in fact also a digital signal and AC does, as they sometimes are negative. The positive part is neglected by the s88, just as an temporary bad connection.

So what happens with the diode is: when there is no train present, the diode doesn't allow any current to flow out from the s88, so it indicates: no.
When there is a train present it grounds the track causing the s88 to detect: yes. Should there be a demand to return current from the loco which isn't satisfied by the other rail, this current flows to ths s88 rail, through the diode to ground. When the digital signal is negative, the diode will not do anything, but it appears to be enough if it allows power to flow during the posive phases.

Regards,
Lars

Hello Lars,

That explains it very well. I reached to the same conclusion, but only after I saw the diode working. The key to the solution is that we have AC current feeding the tracks, so the diode does its job only in one of the phases.

George

Hello from Norge

It is one more positive ting abot the diode i dident told Georg.
You can have longer vires on the input signal to the S88. And you geth not so much bad/wrong signal in. You can se it, when you run on a computer.

Hilsen Nils

I have read about the sensitivity of s88's, and the tendency to get wrong signals. But for some reason I have not experienced any problems of that kind myself. I am reading the signals on a computer, and all readings have been very reliable all the time.

Hi Nils:

Welcome to the forum - you'll like it !

Regards
John

Hi Nils,

Welcome indeed, normally new members enter the Forum with questions rather than solutions !

I use current detection on the central rail, for track occupation (see figure), and isolated ground track for momentary pulse. I never encoutered any problems with that, but my lines are short. I will certainly try your diode.


Enjoy the Forum,
Gregor