Testing of Signals at Start Up of RM - Faults Found

[PJ_model_trains]

The following information has been emailed to HRMS and includes the findings and testings by Ray and myself over the last few days.

 

We are aware of a number of problems using signals with programs in RM but, following a few frustrations, testing was then moved to looking how signals respond at Start up of RM. Here we found problems which we feel need to be addressed first, once fixed testing and fault finding can continue with the other signal issues we have found.

 

LAYOUT (OVAL) AND SIGNALS BEING TESTING 

 

----- 277 -------------------- 279 -----

!                                              !

!                                              !

!                                              !

--- 265 --- 115 <------------ 253 ----

 

All signals on screen and on the layout, for the purpose of testing, are all visible from my control position in the south west corner.

 

Signal 115, is a 3 aspect signal, the other 4 are 4 aspect signals.

 

The tests were carried out with the following settings

• Signals were set to CLEAR (Green) in the layout design section
• The box was ticked to set points (and signals) on starting RM
• We did find that signals do work, on screen and on the layout, 'when clicked independently', one by one, aspect by aspect.

 

This is only basic test of RM and signals at start up of the software to show problems experienced at the outset. We feel it best that these issues are fixed first so that other issues are not influenced or affected in anyway by them.

 

Note: We are both experiencing the same issues.

 

TESTING of SIGNALS at START UP using RM

 

TEST – 1 set all signals to RED from the screen icons (setting screen and layout to match) and exit RM

• Restart RM
• Layout - some signals are Red others are Green? – Errors
• Screen – ALL signals are Green – correct

 

TEST – 2 set all signals to YELLOW from the screen icons (setting screen and layout to match) 

• Restart RM
• Layout - some signals are Yellow others are Green? – Errors
• Screen – ALL signals are Green – correct

 

TEST – 3 set signal 115 (3A) to Y and others (4A) to YY from the screen icons (setting screen and layout to match) and exit RM

• Restart RM
• Layout - some signals are Yellow/Yellow 115 is Yellow others are Green? – Errors
• Screen – ALL signals are Green – correct

 

TEST – 4 set all signals to GREEN from the screen icons (setting screen and layout to match) and exit RM

• Restart RM
• Layout – ALL are Green? – correct
• Screen – ALL signals are Green – correct

 

THE NEXT STAGE WAS TO RUN THE SAME TEST AGAIN FOR ALL SIGNALS.

 

RESULT of second run of the 4 tests, doing exactly the same as in the first test stated above.

 

TEST – 1 set all signals to RED and exit RM

• Restart RM
• Layout - some signals are Red others are Green? BUT different signals this time – Errors
• Screen – ALL signals are Green – correct

 

TESTS – 2 & 3 gave exactly the same – some the aspect colour set prior to close and restart of RM - BUT different signals this time – Errors

 

Although the same problem happened in both tests, for myself and for Ray. We both also experienced the randomness of the aspects shown on re-start.

 

Some aspects were correct, some were not changed from what we set them before restarting RM, but the order they showed was never the same, always a different permutation following a re-start?

 

If any of the above is unclear or further information and or testing is required please do not hesitate to contact us, both our email addresses are included.

 

Thank you for your help

 

Kind regards

 

Paul & Ray

[St1ngr4y]

In my case, I have 20 Traintech signals, all configured to change to red on startup. I changed them all to green, then closed down and reloaded Railmaster. During the restart, it attempted to set all 20 signals in less than 1 second and only two of them actually changed to red.

Ray

[PJ_model_trains]

Thanks Ray

 

That reminds me of other items we discussed at the start up stage.

 

When RM starts, with box ticked to set points (and signals at start up) 

 

The first screen we see, includes the signals set as we  had set them in the layout setup part of the program for signals, so RM must read these items at this stage. I tried setting all signals top half of screen to Red aspect and all on bottom half to Green aspect (viewed from the screen), this is what happened...

 

- with the box not ticked to check all points and signals at start up, the first screen shows every signal on screen Green. This must be the default setting. Moving on...

 

- with the box ticked to check all points and signals at start up, the first screen shows them as they had been set up in layout plan signal settings, either clear or stop. The ones that were not correct can be discussed later, not standard 2,3 or 4 aspect signals.

 

Close down and restart between every test.

 

 

So back to box clicked to check points and signals each time at start up...

 

- First screen shows signals on screen as was set in layout plan settings.

 

- Next it checks the points, one by one with the expected pause between each point it checks and sets. 

 

- Then it checks the signals and 'in a second' it has done it, or has it, only HRMS can confirm this, we appreciate checking setting signals does not take long unlike checking and setting points?

 

This information is included mainly to show what we have seen during testing. Signals are set for the first screen, points are checked and set next, signals are checked and set after that. This would mean signals are checked twice, or so it appears? It is possible it does read the data twice, we do not know so add this in case it may make someone think we better check this out.

 

[PJ_model_trains]

As my initial test was on a single basic loop, I have since done a check on all signals on my layout (excluding temporary feather control signals or other makes of signals some with feathers incorporated.)

 

Therefore keeping this test to just Train-Tech signals basic 2, 3 or 4 aspect.

 

There are 15 basic Train-Tech signals on my layout. Changing all aspects to the opposite of what they should be, so that on re-start of RM all signals have to be reset/changed... only 4 of the 15 changed correctly. On a second test of the same only 3 changes and not the same signals?

 

[Fishmanoz]

PJ and Ray, explained well, I have no problem following each of you.  Very interesting, particularly the random nature of the problem.

 

Can I ask a couple of things: first, is my memory correct that these signals are all controlled by custom Train-Tech decoders?  Then are they 100% reliable at changing when you do so manually from the RM screen?  And last one, with the previous faults reported of switching sequences back down the line to simulate block operation, was there a random element to the faults you found, or were they wrong consistently?

[St1ngr4y]

Hi Fishy,

I can answer the first two, I'll leave the last one to PJ.

Yes these signals are all controlled by custom TrainTech decoders

Yes they are 100% reliable at changing when switched manually via the RM screen.

I have reported another fault with 1.62 which hasn't been mentioned yet. These signals only change about 10% of the time when switched by a RM program. HRMS are trying to reproduce that fault at present.

You may remember I had a problem several months ago with these signals which turned out to be that the signals were only using 1 byte addresses. I was offered a free firmware upgrade for all of these signals which would allow two byte addresses, but I declined this and decided just to keep my addresses < 256 to overcome that one.

Ray

[RB51]

Can I ask either Stingray or PJ a couple of questions please?

 

I have started to look at signalling using 3 aspect Train Tech signals but with Tawcraft decoder/controllers. The Tawcarft decoders as far as I can see are essentially the same as the Train Tech version except they have their own power supply - they don't take their power from the DCC bus system. I should also point out that the signals exist only within RM at present. There are no signals on the layout - as yet as I am still trying to decide whether to buy any.

 

I have set up signals on the RM screen and if I associate one/or more of them with a point ID the points will alter when I operate the signal. I have failed to far to get a point icon operation to alter a linked signal state.

 

If I want to go "down the line" changing signals it seems you have to put all the "down the line" signal changes into the one signal you are operating. So, in this example --

/media/tinymce_upload/78ebf4cf3c0ee5bba0c429bdf36cb365.jpg

If I set signal D's options to change points 1 & 2 and alter the state of signals E & F, I find that I also have to put the signal change for B & C into signal D's option table. I thought that B & C would be "informed" of the instruction "When A is RED" and change according to their settings in signal A?

 

From your experience - am I missing something?

 

Thanks R-

[RB51]

Where I have put ....

 

I thought that B & C would "see" the instruction "When A is RED" and change according to their settings in signal A?

 

I should have more correctly have said..

 

I thought that B & C would be "informed" of the instruction "When A is RED" and change according to their settings in signal A?

 

Anyway, you get the drift I hope.

R-

[RB51]

Have noted that the image in my original post on this subject is not loading - on Windows anyway. It is fine on a Mac, so my apologies to Windows users. I have informed Admin. R-

[Fishmanoz]

Not working on iPad either Roger. 

[RB51]

It seems as though the original image didn't work properly - and my attempt to do it again has failed "Image type not allowed". Drawn in Keynote on a Mac and saved as PDF. Have converted to JPEG - awaiting Adam's response.

 

EDIT: Seems that was OK - if a little large. R-

 

 

[RB51]

This attempt with JPEG. Is this OK Adam? If so can you put it with my first post? R-

 

/media/tinymce_upload/78ebf4cf3c0ee5bba0c429bdf36cb365.jpg

[Yelrow]

ROG, will need magnifying glass/ binoculars. Image is fine, just needs Guy Fawkes, to blow it up. john

[RB51]

Ha-ha-John. Felt like blowing all this tech up I can tell you.

 

I had put an explanation on here - but it seems it has been moderated out. R-

[PJ_model_trains]

Hi guys

 

Thanks for your input, Ray and myself after testing the signals and sharing our thoughts and findings, had a chat as to whether to add these details to the forum at this stage. We didn't want HRMS to feel we had done so for the wrong reason, our reasons were to make known the issues so that others could test if the wished to do so and comment if they wish to do so. So we made the correct decision.  ;o)

 

It was coincidence we found the issues at start up, quite a few issues showed in testing with programs, then there was the lets go back to the start thoughts and test from the beginning. This basically was to make sure everything was working up to a point, it is good we did and important to clear these initial issues before going forward to the others. This is the reason this thread started at this stage and so far has not included other issues.

 

I think, it is better that we look at the faults in stages, hopefully HRMS will agree, these are...

Stage 1 - fix the initial issues

Stage 2 - look at the other issues in programming

Stage 3 - finally then we should look at the back down the line problems (By they the initial issues may clear the down the line problems, if not it is not a problem we can then look into those issues with the knowledge everything else is working)

 

Our concern at this stage was not just that the issue is when initally starting RM it was also that the results each time were 'random'.

 

You will now see why the additional faults, programs and programming down the line are not mentioned at this stage.

 

Ray and myself are, whilst we wait a fix on the start up issues, sharing our experiences with program issues and testing the items we find so that we can then report to HRMS as individual personal issues or collectivvely we both replicated the same faults. Hopefully this should help HRMS in their fault finding and fixing too.

 

 

[RB51]

Well - that's put me in my place I reckon. R-

[PJ_model_trains]

PJ and Ray, explained well, I have no problem following each of you.  Very interesting, particularly the random nature of the problem.

 

Can I ask a couple of things:

first, is my memory correct that these signals are all controlled by custom Train-Tech decoders?  Then are they 100% reliable at changing when you do so manually from the RM screen?  

And last one, with the previous faults reported of switching sequences back down the line to simulate block operation, was there a random element to the faults you found, or were they wrong consistently?

 

As Ray has confirmed the signals we are testing with are Train-Tech signals with their own build in decoder.

 

They are 100% reliable and are the ones HRMS recommended we should use.

 

In my initial report I did add, the test were only with Train-Tech 2, 3 and 4 aspect signals, I have other signals on my layout from other companies, these were not included, the aim is to test with Ray and myself, using just the signals above at this stage.

 

We can confirm that all 2, 3 & 4 aspect Train-Tech signals with built in decoders all work correctly from the icon in RM when clicked to change the aspect (one by one). They may not start correctly and may need clicking to change them into sequence but after that, 2 aspect R/G, 3 aspect R/Y/G, 4 aspect R/Y/YY/G all toggle in the correct sequence on screen and on the layout. 

 

Ray and myself have not included problems in programs, nor have we included problems down the line, we didn't include them at this stage so as not to cloud the issue, a lot of time has gone by, a little longer will not hurt. The aim for HRMS and all users has to be to find and fix the problems we are experiencing, we felt it best to take it in stages, programming back down the line, although important, isn't as important as getting these initial issues sorted first.

 

Yes Fishy, the 'random' issue at start up was our main concern, problems at such an early stage are not good to provide random results makes them more of a concern, we feel.

 

[St1ngr4y]

If I set signal D's options to change points 1 & 2 and alter the state of signals E & F, I find that I also have to put the signal change for B & C into signal D's option table. I thought that B & C would be "informed" of the instruction "When A is RED" and change according to their settings in signal A?

 

From your experience - am I missing something?

 

Thanks R-

I'll just add one thought to your question for now. If it was my layout, I would do it the other way around - i.e. have the signals react to the points settings, instead of having the signals change the points.

Ray

[PJ_model_trains]

Hi Roger

I have started to look at signalling using 3 aspect Train Tech signals but with Tawcraft decoder/controllers. The Tawcarft decoders as far as I can see are essentially the same as the Train Tech version except they have their own power supply - they don't take their power from the DCC bus system.

 

Does this mean you are only starting to look into purchasing them, you haven't done so yet?

 

I do wonder why you are considering the Tawcroft decoder, the power draw from the built in Train-Tech decoder is so small. I also wonder why have a decoder under the layout when your signal is the same regardless, the Train-Tech decoder is in the base of the signal.

 

Another point is you may have to reset your decoder, if it is under the layout it is harder to get to, with the decoder in the base of the Train-Tech decoder a small piece of wire to touch the two terminals is all that is needed. It is simple and well thought out. I am sure these pluses are also why HRMS recommend them.

 

I should also point out that the signals exist only within RM at present. There are no signals on the layout - as yet as I am still trying to decide whether to buy any.

 

Most of the issues Ray and myself will talk about when the start up issues are fixed will be that what happens on the layout is not what is happening on the RM screen. If youo are only looking at what is on the screen you cannot test or replicate the issues as checking the two work together in sync is what we are looking at.

 

Ray will no doubt confirm this also.

 

[PJ_model_trains]

ROG, will need magnifying glass/ binoculars. Image is fine, just needs Guy Fawkes, to blow it up. john

 

Nice twist on words John ;o)   Blow it up!!!

Back on topic.

 

[PJ_model_trains]

If I want to go "down the line" changing signals it seems you have to put all the "down the line" signal changes into the one signal you are operating. 

 

Hi Roger

 

Are you really desperate for the setting of down the line when there are known issues? Can it wait?

It is an issue that has bugged me for some time (I think you noticed) but even I think the other issues are probably more important at the present time. I hope you agree.  ;o)

 

You would be best just sticking to signals for the time being or you could set signals and points. Leave the back down the line signal setting until we get the initial issues fixed. HRMS have confirmed they are working on them, Ray got a fast response from them on his most recent report to them.

 

[RB51]

Fair enough PJ - I'll butt out and leave you and Ray to your exclusive conversation. R-

[RB51]

I'll just add one thought to your question for now. If it was my layout, I would do it the other way around - i.e. have the signals react to the points settings, instead of having the signals change the points.

Ray

This is what I seem unable to do Ray. Signals alter points but not the other way around. I have clearly intruded where I shouldn't be. I'll leave you both to it. R-

[PJ_model_trains]

Fair enough PJ - I'll butt out and leave you and Ray to your exclusive conversation. R-

 

Butt out/exclusive conversation!    I hope the message wasn't taken the wrong way Roger, it was meant with good intentions as this topic will cover many faults as we progress. I sensed a different tone, not one normally from yourself Roger. ;o(  

 

It just seems important that the issues at start up, especially as of a 'random' nature need to be fixed before getting into the deeper issues.

 

I have issues with 'back down the line' signalling, going back to early this year, a large thread if we remember rightly, but until signals are working, especially these start up issues, even I find I would be giving myself more head aches to start messing with them. We don't know, but the problems at start and in programming could fix the back down the line issues, or worded differently, these issues with 'back down the line' signallng may be affected by the ones we are facing now. Probably not but who knows, we can only hope. At least HRMS are working to fix the issues.

 

The discussions are not exclusive, they are put on the forum for anyone to test, to take part, to contribute to, so that together we can let HRMS know and they can fix for us. The reason Ray and myself are at this stage is that since signals were added to RM, approx. October 2014, we are the ones who went straight in to adding them and experienced issues from the start. We reported fault after fault between January and March this year in the hope HRMS would find and fix the problems. A lot of water has gone under the bridge since then, not to mention more recently Windows 10 issues they could have done without. But now it is back to business HRMS are working on the issues. Together we can test the faults some have experienced and reported, if you test and find different signalling faults Roger, please tell HRMS, please add them on here.

 

[RB51]

Thank you PJ for your response. You cannot be surprised that I was a little "miffed" surely? Your comments were the first in over 2½ years on two forums where my question was dismissed as untimely coupled with a vague suggestion that it was not worthy of your (joint) attention. To answer some of your points either for yourself or anyone else who may be interested please see below.

 

Yes you are correct - I have yet to buy any Train Tech signals. I say exactly that in my post.

 

Whilst I am not desperate to have "down the line" sorted out, it is I think a legitimate question. It it not up to you to ask me to wait. My queries are as important to me as yours are to you.

 

I am not just "considering" Tawcraft I am using them because

* I have 4 of them on the layout already

* They are essentially the same as the Train Tech Version

* They are cheaper than theTrainTech version

 

I dont recall saying where my decoders are located so why you are talking about them as if they are under the layout I am at a loss to understand.

 

Thank you for your invitation to place my queries on the forum - I thought that was what I was doing.

 

R-