Welcome to the forum   
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Share
Options
View
Go to last post in this topic Go to first unread post in this topic
Offline Mikael  
#1 Posted : 15 February 2005 10:00:03(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Hi.
After my great success in constructing my own digital signals, I have now got an even more crazy idea: A "reduced speed" module.
The idea is to make some sort of electronic device, that enforces loks in its control area to have a reduced max speed. That could be a single block, a station or a shunting area. My vision is to see a train at full speed, slowly decelerate and drive gently into a station, and then slowly come to a complete halt at the platform. The only catch is that I don't want any feedback from the loks, and I certainly don't want to see a PC connected to my layout Cool

One other practical use I see, is to have this speed limit in my shadow station, so trains won't blast full speed in there, and derail before I've had a chance to slow it down.

So how about it. Is this an ambitious but good idea, or am I just downright crazy?
Offline nico van zon  
#2 Posted : 15 February 2005 10:32:17(UTC)
nico van zon


Joined: 25/07/2004(UTC)
Posts: 202
Location: ,
This functionality already exists in DCC Cool
Offline Mikael  
#3 Posted : 15 February 2005 11:06:52(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Cool. So now we just need to convince M* to discard the Motorola and mfx protocols, and make the switch to DCC biggrin
Offline digilox1  
#4 Posted : 15 February 2005 11:28:33(UTC)
digilox1


Joined: 28/05/2003(UTC)
Posts: 719
Location: ,
Mikael,
An unqualified suggestion from a non techie.
What about designing modules that have to be hooked up downstream of
the control unit, transmitting speed commands to the rails, that can be programmed into the module.
The module could also be designed to respond to switch commands so, it can be integrated in routes.

Regards,
Manfred
Offline Mikael  
#5 Posted : 15 February 2005 11:55:49(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
That is a possibility. I haven't decided on the practical approach just yet. I have two basic ideas:

1. Make small, individual modules, that is more or less integrated with a new version of my signal module. This will be able to control a single block with a single train only.

2. Make a small adapter, that will be inserted between the controller and a booster. The output from the booster will then be speed restricted, and can be used to feed an entire station area (and more).

The first option is by far the most flexible one, and would allow for some very interesting operational aspects. Unfortunately, every module of this type will have to have some power electronics (in effect a mini-booster per block), and they won't be cheap to make.
The second option will cost a real booster, but can feed a large area where you may have had the need for a booster anyway. You will only need one adapter for the booster, and I think that can be built for around $20.

I haven't decided which way to go yet.
Offline Lars Westerlind  
#6 Posted : 15 February 2005 13:59:31(UTC)
Lars Westerlind


Joined: 19/10/2001(UTC)
Posts: 2,379
Location: Lindome, Sweden
I made some experiments some years ago, with a resistor (10 - 40 Ohm, actually, the Viessmann power resistor) intended for analog locos. Of course this is a really bad solution, as different locos have a significantly different current comsumtion. But nevertheless, I came to the opinion that braking by reduced voltage must include individual handling for different locos. And if so, it could be as good to try to modify digital signals.

As I have LocoNet, another idea has been to make a module that listens to LocoNet commands at one side, and transmit modified commands on another to a booster like Power2, with a signal that is like the original but with less speed. But this is far to much for me to build in real life.

You could use a second digital system of course; fixed set to a slow speed step, and swith in this signal into brake sections. Not a very elegant one, but would work, as long as it is assured that short through slider passing sections borders are avoided.

/lars
Offline Mikael  
#7 Posted : 15 February 2005 16:13:20(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
It was something along the lines of your LocoNet idea, I had. I was just thinking of using the original M* booster connection.

Are there any documents on the specifications of Märklins booster connection and/or LocoNet somewhere (in public that is)? It may be wise to study the connections/protocols before deciding to go on with this project biggrin
Offline digilox1  
#8 Posted : 15 February 2005 16:46:10(UTC)
digilox1


Joined: 28/05/2003(UTC)
Posts: 719
Location: ,
Mikael,
http://home.arcor.de/dr.koenig/digital/homepag.htm

...should contain a wealth of information on almost any issue regarding trinary model train control; scroll down to get to the English translations.

Regarding Loconet:

http://www.digitrax.com/...conetpersonaledition.pdf
http://www.digitrax.com/ftp/tx%20patent.pdf

Regards,
Manfred
Offline Lars Westerlind  
#9 Posted : 15 February 2005 19:28:36(UTC)
Lars Westerlind


Joined: 19/10/2001(UTC)
Posts: 2,379
Location: Lindome, Sweden
Yes,
you've got the links. With your knowledge I suspect LocoNet should be interesting; It's really designed to allow for hobbyists to interfer.

This 'personal edition' is the description intended for this purpose. There is also a possibility for firms to do it professionally, I get the impression that Digitrax are rather encouraging in this, even though, there is a license fee for this I think (not so for personal use).

I think you should look at it, but I'm not sure it's good for this specific problem.

Regards,
Lars
Offline Mikael  
#10 Posted : 15 February 2005 21:55:04(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Thanks. I will have a good look at those (and other) pages, and then we'll see what will be possible within reason.
Offline perz  
#11 Posted : 15 February 2005 22:21:20(UTC)
perz

Sweden   
Joined: 12/01/2002(UTC)
Posts: 2,578
Location: Sweden
Hi Mikael,

I think your original idea of having a "translation module" inserted between the control unit and the booster is a possible and not too difficult solution.

One problem with this solution is that you would need to have transition segments between the normal and the slow segment, since the commands will not be synchronized between the slow section and the normal section.
Offline digilox1  
#12 Posted : 15 February 2005 22:47:22(UTC)
digilox1


Joined: 28/05/2003(UTC)
Posts: 719
Location: ,
http://www.netwings.ch/umelec/hompa52.htm

A somewhat different approach...Do you think this principle could
be adapted to the trinary protocol with homebrew loco decoders?

In DCC, it`s gaining terrain. After Umelec, Lenz, ZIMO and ct- elektronik have announced the feature to become available.

Regards,
Manfred
Offline dx2krudop  
#13 Posted : 15 February 2005 23:17:29(UTC)
dx2krudop


Joined: 20/02/2002(UTC)
Posts: 2
Location: ,
Viessmann already makes one. I tried it and it works. The problem I had was with the lok lights dimming. Also, since I have C tracks with the 74460 decoders, there wasn't enough power for the turnouts to work. I had to reroute the wires for the turnouts in the area where I wanted the train to slow down.

http://www.reynaulds.com...smann/temp.asp?item=5216
Offline jorge_vilarrubi  
#14 Posted : 15 February 2005 23:50:59(UTC)
jorge_vilarrubi


Joined: 15/12/2003(UTC)
Posts: 655
Location: Buenos Aires,
This is what I've done in my layout in order to have trains reducing its speed when entering the station:
I feed my IB with 16V and control tha main layout with it.
For the station I have a booster fed with 15V. A Delta control can be used as well.
This 1V difference makes the trains to reduce its speed.
If you want less speed, you can use a 14V trafo.
Or you can increase the voltage to the IB up to its limit.
You have several alternatives.
It's simple, easy to install and it works.
Best regards,
Jorge Vilarrubí
Buenos Aires
ARGENTINA
Offline Mikael  
#15 Posted : 16 February 2005 00:13:36(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Are you sure this will work? Both the resistor and the lower voltage?
All of my loks are regulated, and I would guess this counteracts both of these solutions. If I ran unregulated delta or even analog loks, I can see the point.
Offline Mikael  
#16 Posted : 16 February 2005 00:23:18(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
I just had a quick look at LocoNet, and my initial impression is that it will be just as easy/difficult to make as tapping in to a M* booster plug. LocoNet has a big and complex protocol, but it isn't used by boosters. Instead, a low voltage copy of the track output is present on two separate wires in the plug.
I had started to wonder how they could achieve perfect synchronisation between master controller and boosters, when all sorts of units communicated on the same cable. Well, with a dedicated line for boosters, there is no problem.
Offline efel  
#17 Posted : 26 February 2005 17:57:28(UTC)
efel

France   
Joined: 23/02/2005(UTC)
Posts: 800
Hi,
I'm glad to see I'm not alone with this problem!
I'm looking for an inexpensive solution for reducing speed in the shadow station, to avoid derailment, and halter the train on a shorter distance (no much room....).
I started also with a series resistance in the traction power: the results were not good, especially as I have both delta and "digital" locos: it's not easy to find the balance for the resistance value, my delta loco being very sensitive to this parameter.
So I made some current measurement, and saw that, surprisingly, the peak current for the locos are larger when the programmed speed (mobile station) is lower. (Of course, the current is pulsed, and the duty cycle is much lower at low speed and, at the end, the mean current is lower at low speed.)
Here are some measured values:

S600 (delta): low speed 1.5A peak, during 1.5ms. Duty cycle:1/4; high speed 0.6A peak, during 5ms. Duty cycle:4/5
S16 ("digital"):low speed 1A peak, during 1.5ms. Duty cycle:1/4. high speed 0.4A peak, during 5ms. Duty cycle:4/5
(these values are rather approximate for the signal is very "noisy" (there is also low frequency variations) and I don't have a sophisticated scope). But that explains the result: using a series resistor, the voltage drop not only depends on the loco, but is larger at lower speed, leading in some cases to halt the loco, even programmed for medium speed!

I tried then to put a constant drop voltage in the traction power, independant of the loco current: a zener diode. (in fact 2 zener diodes in series, inversed, to get a voltage drop both for the positive and the negative half-wave).
Using 6.8V zeners (warning:>2W), leads to a result acceptable for me:
-For "digital" locos: maximum speed reduced such that the maximum halting distance at the signal is divided by 2.
If the loco is at low speed, the speed reduction seems less, due to load regulation, but the halting distance is of course still lower!
-For delta loco: same as for digital for max speed, but for the extreme lower speed, it stops: I must use the "second" speed step.

Conclusion: far from being perfect, but also quite better than the resistance, and rather good if only "digital" locos are used.
And only 1$ (2 zeners) per district!

Any idea for improving ? [:p]

Fred
Offline Mikael  
#18 Posted : 26 February 2005 22:29:43(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Very interesting. I think I'll try this before I get carried away and make a microcontroller-based solution.
Offline Mikael  
#19 Posted : 12 March 2006 18:04:00(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Finally I have some news about this idea I've had for a long time:
I have just completed the first prototype of my reduced speed module (only the hardware for now), to see if it was at all possible to make a system powered by the track voltage only, and send a modified signal to a lok. Well.... It was possible biggrin
Now I "only" need to write some software for the microcontroller, that will slow down (or even stop) the trains. Unlike Tams similar module, I will try to retain the full control over the loks (except for the top speed). And I don't need a heatsink either wink

Here is the prototype as it is now. I expect it to be much smaller when I am finished with it (IF I can make it work reliably, that is).

UserPostedImage
Offline Mikael  
#20 Posted : 13 April 2006 18:50:58(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
I ran into some problems which had to be solved in hardware, but now it seems to work fine. I can drive a train from a normally powered track to this module, without any transition tracks, and without my Intellibox or Mobile Station complaining about a short circuit. The hardware became a bit more complex:

UserPostedImage


But all tests seem to suggest that there are no more problems. Now I just have to write a bit of software, and a slow train section should be a reality. My guess is that I have it up and running in a day or two.
Here is my testbench:

UserPostedImage
Offline MHauge  
#21 Posted : 13 April 2006 19:31:22(UTC)
MHauge


Joined: 19/10/2005(UTC)
Posts: 393
Location: Aarhus C, DK
Looks good, wish i had that kind of equipment available..

Michael
Märklin C-tracks, Mobile Station, Danish Ep 4
Offline foumaro  
#22 Posted : 14 April 2006 10:19:39(UTC)
foumaro

Greece   
Joined: 08/12/2004(UTC)
Posts: 4,430
Location: Attiki Athens Greece
So,now we have the Mikael protocol.biggrinbiggrinbiggrin
Offline Mikael  
#23 Posted : 14 April 2006 20:13:54(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
HOLY SMOKE. I do believe, I've succeeded.
It required quite a bit of programming to keep F1-F4 in the datastream, and I know now why Tams don't do it. It is far from easy to do so. Anyway, I got it to work, and right now I'm just riding a gigantic rush I got from watching it do exactly what it should.

Here is a video I made of the test. The module is set for limiting the speed to step 4. (10 MB Mpeg1 format):
http://www.ejberg.dk/trains/sig..._speed_demonstration.mpg

This module doesn't need any transition tracks. Just an isolated section with the module inserted between the normal track power and the isolated one.

Edit: Video file changed from Windows Media Video to MPEG1 due to some people having problems viewing it.
Offline MHauge  
#24 Posted : 14 April 2006 20:27:58(UTC)
MHauge


Joined: 19/10/2005(UTC)
Posts: 393
Location: Aarhus C, DK
It looks great, but seems tat engine sound is gone on the slow section ?

Michael
Märklin C-tracks, Mobile Station, Danish Ep 4
Offline Mikael  
#25 Posted : 14 April 2006 22:08:54(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
No, that is normal for this lok. It does that whenever the speed is reduced. In this test the reduction in speed is quite large, so it takes a long time for the sound to return. However, you should be able to hear the sound come back before the lok exits the section.
Offline Mikael  
#26 Posted : 15 April 2006 15:54:13(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
And then it didn't work after all.
It worked fine on the Mobile Station, but when I connected my Intellibox instead, there is nothing but trouble [:(]

Does anyone know if it is possible (via a SO) to alter the lok refresh cycle in the IB? It seems that the IB doesn't send out more than one double packet before it moves on to the next lok. The good old 6021 (and the Mobile Station) sends out two double packets for each lok, and I have taken advantage of this when making my reduced speed module.
Offline MHauge  
#27 Posted : 15 April 2006 16:42:29(UTC)
MHauge


Joined: 19/10/2005(UTC)
Posts: 393
Location: Aarhus C, DK
Ok, my trains with sound just have more sound when driving slow.
How about MFX, will that work too ?
I'm thinking about many sounds/functions.
Or is it "only" the first 4 that will work ?

Michael
Märklin C-tracks, Mobile Station, Danish Ep 4
Offline Mikael  
#28 Posted : 15 April 2006 18:57:40(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
No, the mfx protocol isn't readily available anywhere, so I don't have the slightest chance of making that work. Mfx loks operate fine through my module, including all sounds. They just don't slow down (and no autodetection either).
I don't mind that myself, as I normally use an Intellibox.
Offline MHauge  
#29 Posted : 15 April 2006 19:30:50(UTC)
MHauge


Joined: 19/10/2005(UTC)
Posts: 393
Location: Aarhus C, DK
Too bad, just assumed that they would slow down too.

Michael
Märklin C-tracks, Mobile Station, Danish Ep 4
Offline Mikael  
#30 Posted : 15 April 2006 20:39:52(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Yes. I would love to get mfx loks to slow down too. That would make this project much more useful for the generel public. But the mfx protocol isn't available for us mortals, and as far as I know, nobody has succeeded in reverse engineering it.

But the good news is that I got it to work with my Intellibox. I just had to reconfigure the IB a bit, so it does not generate DCC idle packets.
Offline mmervine  
#31 Posted : 16 April 2006 01:26:47(UTC)
mmervine

United States   
Joined: 30/01/2006(UTC)
Posts: 1,893
Location: Keene, NH
Very impressive, I hope that you share your design with the rest of us. This is great for station areas!
Märklin C-track, Marklin Digital & ECoS, multi-era French & Swiss
http://www.ete-ene.org/m...mervines-layout-gallery/
Offline MHauge  
#32 Posted : 16 April 2006 03:42:23(UTC)
MHauge


Joined: 19/10/2005(UTC)
Posts: 393
Location: Aarhus C, DK
I'm driving with a Märklin MS, at the moment i don't have any mfx stuff, but eventually it will come.
I'm using the DIY brakebodule found at :https://www.marklin-users.net/ht...gital/DIbreakmodule.html
Will that not work with mfx either ?
If that doesn't work i think i'll pass on the mfx stuff...
I made two of these for my station, coupled with 4 (24994) circuit tracks, two for each, so when one train enters track one, it stops and starts the one in track two.
Or is this still compatible with mfx ?

Michael
Märklin C-tracks, Mobile Station, Danish Ep 4
Offline Mikael  
#33 Posted : 16 April 2006 12:31:11(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
Don't worry. That brake module works fine with mfx.

And yes, I do plan to make the slow module available for everyone. I just want to make sure it works before I do. It still needs a few features, like how to control the module, so trains can stop completely, and start again when needed.
Would anybody be interested in trying it, before it is finished? If so, I could make the design available, before it is finished.
Offline Mikael  
#34 Posted : 25 April 2006 17:42:05(UTC)
Mikael

Denmark   
Joined: 10/09/2004(UTC)
Posts: 959
Location: Copenhagen, Denmark
For those of you that are curious, here is a preliminary schematic. It isn't complete (yet), and a lot of the parts do not have values assigned. But I got a hint that someone wanted to know the operating principle of my module, and I just wanted you to know too (in case you happened to be interested as well).
http://www.ejberg.dk/trains/signal2/slowschematic.png
Users browsing this topic
Guest
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

| Powered by YAF.NET | YAF.NET © 2003-2025, Yet Another Forum.NET
This page was generated in 1.548 seconds.