Hi Martin,

many thanks for posting your observation, I'll see what I can do. I am using a 100 MHz oscilloscope, so the signal plots should be reliable.

Perhaps you can solve your problem if you drastically decrease the resistor to 100 Ohms. A capacitor of 47 uF should be o.k.

As it seems some loks and decoders need more current in the brake section than others, so a smaller resistor may help. I'll check that when I'll back at home next week.

It's definitely a good idea to have a short connections to the power supply. It is not only the voltage drop from the ohmic resistance of long track sections, but also the inductivity of the center rail which may cause a severe signal deformation and confuse the decoder.

Regards

Hans Martin

Hi all together from a new forum member

Question: Why doesn't the switching presented by jte require a transition section? Isn't there any danger to get short circuit by the locos slider passing from the "normal" section into the braking section?

Thanks!

Rob

Hi friends,

Today I did some further work with the brake modules discussed here. Allthough I have it in use for quite some time now, and it works well with various FX, MFX, and ESU Lokpilot decoders, I have now encountered a problem. The problem occurs only for passenger trains with coach lighting, pulled by locos with MFX/ESU decoders.

The symptoms are as follows: when the MFX loco enters the brake section of a halt signal it first starts braking and reduces its speed. However, as soon as the ski contact of the first coach enters the brake section, too, something unexpected happens: the lok will no longer keep on braking, but go on with regular speed until it stops abruptly in the stop section.

After the previous discussion the reason for this problem is quite obvious: when the lighting of the passenger car takes its current from the brake section voltage breaks down in the time intervals where the diode is closed, as the RC bridge can't supply the 120 mA for the lighting. As a consequence the MFX decoder forgets brake modus and makes the lok proceed with regular speed.

I have tried to reduce the resistor and to increase the capacitor of the RC bridge. But I don't think that this can be recommended as a solution because the idling current of the brake module is drastically increased.

So if you encounter the same problem as me, here is a circuit proposal that should help (didn't try it yet, but have only little doubt about it):



The idea is simple: divide the brake section in a number of shorter ones. Each section has its own Schottky diode (actually you need 2 diodes per section now), and its own RC bridge. The length of each section should be shorter than the distance between the ski contacts of the lok and the first passenger coach, i.e. probably 1 or 2 tracks of type 2200 or 24172/24188.

As you see you can still handle the complete sequence of brake sections with a single 7039/7040/7041 signal.

Another possibility is to install current conductive couplers between the passenger cars and put the ski contact to the car at the end of the consist, so it will never get in conflict with the brake section. Not always possible, of course. Therefore, any other solutions for the problem are wellcome.

Hans Martin

Hi friends,

Here comes another circuit idea which makes it possible to use locos with MFX/ESU decoder, lighted coaches, and the brake module without transition section. I don't like my last circuit proposal myself any more. Imagine you have 50 signals on your layout: you'll get crazy with transition sections. [xx(]

This solution here sounds a little strange but is pretty simple: put the buffer capacitor on the train, not on the track.

For this purpose you have to install the following circuit in the first lighted coach behind the lok:



I plan to install this circuit in my 43603 streamliner baggage car which usually rides directly behind my mfx GG-1. The baggage car is the only car in the whole consist with a ski contact. It supplies five or six 436xx lighted streamliner coaches using conductive RTS couplers. This requires to brake module to produce a total current of 0.75 amps for coach lighting plus perhaps 0.3 to 0.6 amps for the lok. Under these conditions also a regular M* brake module may throw in the towel here without loosing face.

How it works: First assume the passenger train is outside the brake section, with full digital signal available at the center rail. Then the positive pulses of the digital signal will make diode #1 conductive, and the relay will pull its contact down, thereby separating the large 1000 microfarad buffer capacitor from the digital signal. Ignore diode #2 and the 1 k resistor here. The small capacitor parallel to the relay coil (few micro farads) smoothes the relay voltage a little, so the relay will not flicker.

When the coach enters the brake section, it will find a voltage at the ski contact which is always negative, or at least which is never positive. Thus, diode #1 is closed and the relay falls back. The buffer capacitor is now electrically connected to the center rail via the ski contact and helps to maintain the negative voltage, quite similar as in a usual M* brake module. As a consequence the mfx decoder of the lok keeps on braking in spite of the additional current which the coach lighting take sfrom the brake section. The 4.7 ohms resistor is for protection of the relay contact at a short circuit.

Finally, what is diode #2 and the 1 k resistor good for ? It provides a small negative current to the buffer capacitor such that it is already charged when the relay falls back. This avoids a voltage surge when the coach enters the brake section.

Only thing that might be critical with this circuit: if you use the car an analogic AC layout the buffer capacitor is in danger if the relay fails to pull the contact down, which may happen at low voltages.

Hans Martin

Lars, you are right, of course: If the relay falls back outside the brake section, e.g. due to a short voltage interrupt, the buffer capacitor will be connected and short circuit the whole section, oops[:I][:I][:I]. Also, because of lack of voltage the relay will not come up again.

For this reason I added this small 4.7 Ohms resistor in series with the buffer capacitor. So voltage across the tracks can build up even the relay has fallen back. So the relay should be able to interrupt the circuit again. I am not sure that 4.7 ohms is sufficient, but I'll test this. The circuit should be stable with respect to power surges on the tracks, naturally.

Passing a closed signal in backward direction is still a challenge. I know there is a solution working with an additional 7244 relay, also working under analogic operation.

With DCC I unfortunately have no experience. Isn't it the same as Motorola with respect to braking: constant (negative) voltage at the decoder makes it initialize brake modus ?

Hans Martin

Hi everybody!

I built this circuit yesterday and made some tests. It works fine with my mfx-locos, but the blinking lights are very annoying. Anybody has an idea how to fix this? I'm using mfx-Decoders 609xx and LEDs as lights.

Thanks
-Alex-

Hi Alex,

These flickering coach lights are annoying, indeed. Had the problem, too. If you coach lighting has LEDs, I would propose to take a capacitor of around 50 micro farad parallel to the LED circuit. Best results are if it is placed at the output of the rectifier diode which supplies the LEDs, such that the capacitor is charged to full track voltage.

I have LED lighting in my caboose car, with the very same problem in the brake sections. So I installed a 470 uF capacitor. This is perhaps a little too big, but no other capacitor available at that moment. Now the caboose has light for 2 or 3 seconds after power is switched off.

BTW, wellcome to the forum

Hans Martin

With the plethora of circuits that have been posted here, I'm surprised that no one has mentioned the braking module designed by Huib Maaskant (http://www.floodland.nl/trein/info/digitaal_en_8.htm), which is a cheap alternative for Marklin's 72441 signal module (about $NZ18). I have made several of these and can report they work well, however they do require a transition section. One thing I have found is that this module seems to need a solenoid module load attached to make the signal module work. I am waiting for my Viessmann colour light signals to arrive, so thought I would just run the signal module without any attached signals. Well, I found the module would not work until I attached a C Track turnout point motor in series with the signal module. Hopefully, the colour light signals will provide a similar load (when I finally get them!!).

Other than that the modules work well. Locos with 60901 and C-Sine motors work OK, and locos and coaches remain with lights on when stopped on the braking and stop sections. If you need locos to stop when approaching from behind the signal (against the normal direction) then I guess you would just replicate the connections for the transition and braking sections to behind the signal.

Hi,

Bigdaddynz, many thanks for posting Huib Maskats break module site. I have tested this circuit some time ago, too, and I can canfirm it works fine. BTW I had no problems with it, no additional load needed. Might it be that your relay has some additional internal electromagnetic interference protection capacitors or so. This could explain your observation.

Roland, I use the following combined semaphore and brake circuit on my layout for train control.

It has the restrictions concerning passenger trains with coach lighting as indicated above. It does not require brake or stop section, and it works with analogic and digital operation (fx, mfx, ESU decoders, but not with delta, however).
You will find a more complete description of this circuit in topic https://www.marklin-user...ault.aspx?g=posts&t=4759
(see the 2nd page)

Regards

Hans Martin