RailMaster 1.70 REVISION 2 now available.

[96RAF]

The choice of a replacement DCC system has been the subject of much discussion on several forums with NCE coming out as a firm favourite. Again at which level of a preferred system you decide to buy is subject of more discussion.

 

If you want to run under PC control then you have f.o.c packages like Rocrail and JMRI or several commercial packages, all of which can be bought into at various levels of features and at much greater cost than RM.

 

Setting up any of these model railway software packages is unfortunately not a simple PnP arrangement, which is why I gave up on JMRI but I managed to get Rocrail going at a basic operational level  - no loco detection yet.

 

Rob

[Somerset Modeller]

Railmaster is a very good entry level system but with lots of limitations.

A good alternative is Digikeijs 5000 Controller with iTrain, the problem is that you are looking at £300-400. But you can do a massive amount with it.

The Digikeijs 5000 is a awsome bit of kit for the price.

[Alan K]

Railmaster is a very good entry level system but with lots of limitations.

A good alternative is Digikeijs 5000 Controller with iTrain, the problem is that you are looking at £300-400. But you can do a massive amount with it.

The Digikeijs 5000 is a awsome bit of kit for the price.

I think the more you try to do with RM the more the limitations become apparent. As I said I did try using a tablet PC as a slave with a second full copy of RM. Hopeless - among other issues if I hit a function key on the slave it turned on a different function on the host PC (and on the loco).

Just had a look at at Digikeis. Does seem to offer a lot for the money though I don't know why I might need the various supported protocols with a mix of decoders. Interestingly it will work with the Z21 app. Seems much cheaper than an actual Z21.

Several of the paid for software packages come in various versions with enormous differences in price.- anything from under £100 to £500+. I don't always find it very easy to work out which would actually be required. And you dont know how you would get on with any package until you try. Not all have evaluation facilities.

[96RAF]

If you trawl the forums for recommendations then you are likely to get as many choices as there are folk posting about it. In the end its down to - does the kit in hand suit my purpose and pocket.

 

Recently on such a discussion the Dynamis got absolutely slated, but previously it was the best thing since sliced bread.

 

Often the kit is not totally at fault and its more the user and their preferences. No matter how good the kit, someone will find fault with it.

PICNIC is a term I have seen on another forum - meaning - Problem In Chair, Not in Computer.

[Tomchatz]

Installed 1.70.2 today, everything seemed fine booting up RM but elink isn't working, the green light is on but locos won't move, sounds won't play and points don't work. Device manager says it is connected on com3 and it seems to communicate with RM, I have checked the .ini and it is all correct. The only thing i can think is that elink wasn't connected when I updated to 1.70.1 can anyone help me. Unfotunatly i don't have a track tester to see if its geting power to the track

[Chrissaf]

Device manager says it is connected on com3

.

Does it also show Com3 in RM 'System Settings' screen?

Does the RM 'Controller A' part of the 'System Setting' screen show the eLink in pink highlight?

.

I have checked the .ini and it is all correct.

.

With respect, how can we be sure it is correct. I suggest you copy and paste your .INI file content into a new post, so it can be reviewed.

.

I would also unplug the eLink from the PC with RM running and then plug it into a different USB port so that a new Com port number is assigned. RM should protest and display an error message, do not acknowledge this error message at this time. Instead check 'Device Manager' to see which new Com port has been assigned to the eLink. Then use the 'pull down' field in the RM 'System Settings' screen to select the new Com port number shown by 'Device Manager'. Then acknowledge the RM error message so that it can re-initialise the connection.

[Tomchatz]

@Chrissaf I have tried it on com 4 and it still isn't working, here is mu .ini file

Tipper speed=80Tipper timer=86.4Turntable speed=40Turntable timer=23.56Points timer=0.75Conveyor speed=65TCP start port=30Check serial ports=1Show upgrade button=1Allow deactivate=1Use default curves=1Polling time=5Confirm delete=1Classic buttons=0Show point indicators=1Uncoupler time=5Detection Timeout=5Button bar vertical=0Throttle timer=5Handheld plan area=1,1,135,135Program tick sound=1PING time=60Reset eLink on start=0Enable mouse=0Warn static IP=1Load Hornby Locos=1Load Jouef Locos=1Load Rivarossi Locos=0Load Electrotren Locos=0Load Arnold Locos=0Load Bassett-Lowke Locos=0Point button arrows=1Spoken confirmation=0Controllers on top=1Double pulse=1Alternative Comms=1Check controller=1Elite feedback=1Arnold RailMaster=0

I have just got a class 66 tts decoder and want to try it i guess i can for now 😢

[Chrissaf]

I have just reproduced the last few lines from your file to save listing every entry. It is not the optimum order of entries for an eLink controller. It has been proven more than once that the order of entries in the file has an impact on RM operation.

.

Double pulse=1

Alternative Comms=1

Check controller=1

Elite feedback=1

Arnold RailMaster=0

.

Delete the two lines shown in bold italics above. Then retype them as the last two lines in the file as shown in bold below. If you are using the eLink, then I would also change the value of Elite feedback to 0 (zero) as shown in bold below:

.

Double pulse=1

Elite feedback=0

Arnold RailMaster=0  [ I would delete this line as it serves no purpose, I don't even recognise it as being a valid RM .INI file line....it is not listed in the manual. ]

Alternative comms=1 (note that the c in comms should be lower case)

Check controller=1

.

I'm also going to edit my previous method suggestion to a new revised method shown below:

.

I would also unplug the eLink from the PC without RM running. When RM is started, it should protest and display an error message, do not acknowledge this error message at this time. Instead now plug the eLink into a different USB port so that a new Com port number is assigned.check 'Device Manager' to see which new Com port has been assigned to the eLink. Then use the 'pull down' field in the RM 'System Settings' screen to select the new Com port number shown by 'Device Manager'. Then acknowledge the RM error message so that it can re-initialise the connection.

.

If none of the above improves matters, then change the value of Alternative comms from 1 to 0 (zero) and test again.

.

Although Alternative comms=1 should be the correct value for an eLink. Some users have reported that for them a zero value works better when using the 1.70.2 RM release.

[Tomchatz]

I will try and get back to you

[Tomchatz]

i tried the com port again and it has now lost elink drivers and i cant get it to reinstall them, i also changed the .ini file 

[Tomchatz]

just cleaned the contacts on the elink and tried a new cable and it seems to be working, 

[Chrissaf]

Glad you are making progress. The changes to the .INI file I recommended will not be wasted.....they should help with longer term reliability, even though it would appear that your issue was one of physical electrical connectivity rather than software related on this occasion.

[merciaman]

Chrissaf, noting your comment about removing Arnold Railmaster from the ini file, out of interest, on checking my ini file tonight, I have Arnold Railmaster listed 13 times!!!! I will now as you suggest delete these entries from the ini file, to reduce the clutter.

[pd91223]

I've bitten the bullet and upgraded today from RM v1.69.2. Although I haven't had time to do exhaustive testing yet (made more infuriating by an on-going issue that RM often doesn't close down properly, leaving a window/process open "Please wait ... backing up important files", which can only be killed by Task Manager!), let me report on progress so far. Note that I've reverted to 1.69.2 a couple of times today just to re-test a few settings.

 

First of all, I'm running on a Lenovo Yoga 3 laptop with Windows 10 update 1803 using an Elink (only controller) connected to a USB3.0 port, with which I've had no problems to-date other than those mentioned later re. Train-Tech signals. So, with:

AC=0; CC=0 - seems to work okAC=0; CC=1 - loses connection with Controller within a few secondsAC=1; CC=0 - seems to work okAC=1; CC=1 - seems to work ok

 

Points & signals seem to set correctly; points change correctly; initial trails with locos seem fine. However, all of my Train-Tech signals worked the wrong way round - red=green, green=red - when I upgraded (fine when I reverted) until I reversed the polarity on all of them. Now they seem to work ok, except for the previous issue raised in https://www.hornby.com/uk-en/forum/post/view/topic_id/24291/?p=1 which is not yet fixed! 😢

 

Hope this helps someone. I seem to have been reasonably lucky compared to many others!

Paul

 

[St1ngr4y]

Hi Paul,

 

"(made more infuriating by an on-going issue that RM often doesn't close down properly, leaving a window/process open "Please wait ... backing up important files", which can only be killed by Task Manager!)"

 

Be aware that the backup procedure in 1.70 has been changed in that a lot more backup files are created than before. This is especially noticeable if you have a lot of RM program files. Next time you close down, give it at least a minute to complete this task before using the sledgehammer  😉

 

As far as Traintech signals are concerned, HRMS are still working on the 2-aspect signal issue which I reported with v1.70.

 

Ray

[RB51]

Thanks for including the bit about the Traintech signals Ray - I assume this is based on a response from HRMS? I haven't heard a peep following on from my report. R-

[pd91223]

@St1ngr4y

The issue with the backup process only happens occasionally and was a problem in earlier versions too. I usually give it a couple of minutes or so, just in case, but often I have to kill it after 20-30 minutes!! Mu laptop's not that slow!

 

Thanks for the update on Train-Tech signals. As you know this issue has been on-going for a long time now, so I'm not hopeful of a quick resolution. It's seriously holding me back from developing my layout, so much so that I'm looking at alternative controller & software solutions - but I do like RM otherwise. Thanks again.

Paul

[St1ngr4y]

Paul,

 

What I discovered with TT 2-aspect signals under RM v1.70, was that on starting a new RM session, the icons on the layout diagram, when clicked, showed the opposite aspects to the signals themselves, UNTIL each signal had been switched by a RM program. A temporary fudge is to create a RM program which sets each TT signal to the aspect you want it to be at the start of a RM session, and to run this program once only at the start of a session once all the points/signals have been set (if you have requested that to happen). You should find that, for the remainder of that session, the layout icons, when clicked, will remain in sync with the actual signals.

 

Ray

[St1ngr4y]

Thanks for including the bit about the Traintech signals Ray - I assume this is based on a response from HRMS? I haven't heard a peep following on from my report. R-

Hi Roger,

 

Yes - I have interchanged a few emails with HRMS since my original report, the last of which suggested that they are going to do further tests. Their focus seems to be on how the signals are programmed by RM when the signal is switched to "learn" mode. You may remember that, a few versions ago, they put changes into the layout designer so that, in the TT signal configuration window, the RED test button was to be the one to press to allocate the address to the signal in "learn" mode. When they did this, they also put code behind the GREEN test button, so that when this button is pressed, a warning message is displayed telling the user to use the RED test button. I believe that the manual was amended to persuade the user to use the RED test button as the recommended way of programming the signal. During my email interchange with HRMS, I also emailed the manufacturers who confirmed that, whatever value (0 or 1) is transmitted to the signal when it is in learn mode, then that value would subsequently be used to activate the GREEN aspect. The RED test button causes a value of zero to be sent to the signal, so if all users use the RED test button for a signal, then those users' signals will all show GREEN when sent a value of 0 and RED when sent a value of 1. 

However, what I also discovered during this period was that if the user clicks on the RED or GREEN sequence spots in the signal configuration window, that these, too, send out a switch signal DCC packet. So it is quite possible for a user to click one of these "spots" to change the sequence (although I don't know why one should) while the signal is in learn mode, and to send accidentally a value of 1 to the signal which would then be used for the GREEN aspect thereafter, contrary to what RM expects.

 

Ray

[RB51]

Thanks Ray, that is a very useful insight to the system. 

 

I have seen the red and green test buttons (and used them - and got the warning) but I didn't programme or synchronise my TT signals that way. I used a link wire and touched the tags under the signal which as you know causes the signals to rapidly alternate between the two aspects. I then went back to RM and operated the relevant point. That was the learning done - as far as the signal was concerned anyway. The problem though is the same using this method as using the process you outline in that the TT signal always settles on green due to the first DCC signal packet being used that way, so you could still end up out of sequence.

 

All I did then was set the point blades to the red road, put the TT signal back in learn mode and go through the process again. The point and the signal are then in tune with each other, and they have stayed that way.

 

R-

 

[St1ngr4y]

Good morning Roger,

 

The reason why they recommend using the RED test button is so that they are sure that a 0 value is sent to the signal when it is in learn mode. Thereafter, RM uses a 0 value to set the signal to GREEN and a 1 value to set it to RED. However, using your method when programming the signal, by either clicking a point icon with the same address as the signal or by clicking the signal icon itself, it is very possible that a 1 value is sent to the signal. This means that subsequently a 0 value sent to the signal will change it to RED and a 1 will change it to GREEN, which are the opposite ways that RM thinks it is switching them.

 

Ray

[Chrissaf]

The reason why they recommend using the RED test button is so that they are sure that a 0 value is sent to the signal when it is in learn mode. Thereafter, RM uses a 0 value to set the signal to GREEN and a 1 value to set it to RED.

.

Does seem rather skew-eyed logic Ray. If RM normally sends a 0 for green and a 1 for red. Why didn't they keep the same logic for the 'learn mode'? or am I missing something here.

.

[RB51]

It does seem a bit random. Admittedly I have not operated mine for a while, whilst I rebuild the layout, but up to the point of demolition they behaved perfectly. 

 

Perhaps, and given the apparent counter intuitiveness and sporadic nature of RM, I have just been lucky.

 

R-

[St1ngr4y]

The reason why they recommend using the RED test button is so that they are sure that a 0 value is sent to the signal when it is in learn mode. Thereafter, RM uses a 0 value to set the signal to GREEN and a 1 value to set it to RED.

.

Does seem rather skew-eyed logic Ray. If RM normally sends a 0 for green and a 1 for red. Why didn't they keep the same logic for the 'learn mode'? or am I missing something here.

.

 

I quite agree Chris. It would have been far better if they had chosen the GREEN test button.

 

Ray

[Fishmanoz]

Might I suggest this latest discussion be moved to an appropriate TT signal thread or to its own thread, little chance it will ever be found again if left here.