Throttle settings, speed steps, acceleration and decoder CVs

[Chrissaf]

This may well be a question only HRMC can answer, but I don't see it is something I can really send in as a support problem.

 

When you press the loco's shunt or cruise button does RailMaster simply send out a specific speed step value (based on the locos calibration)  and allow the acceleration/deceleration CV  values in the decoder to handle accelerating/decelerating to the specified speed step or does  RailMaster  manage acceleration/deceleration itself by sending out a sequence of speed steps? 

What I'm really trying to understand is the relationship between setting the throttle to a specific MPH value and the interaction with the acceleration/deceleration and speed curve CV settings in the decoder (i.e. CVs 2,3,4,5 & 6)

Hi Nick,

You seem to have similar curiosity to mine, in that I am trying to find out how the DCC protocol handles sound functions. Anyway, to answer your question, my understanding is that it is the former of your two synopses which is the case. If you get around to creating programs to run on RM, you will find the commands accelerate and decelerate with which you can gain further control than is provided by the loco decoder CV settings. Personally, I don't try to match speed steps to scale MPH - as long as the speed/acceleration/deceleratuion of the train looks correct, to me that's enough.

Ray

 

 

[Colpatben]

I have entered this Topic because it seems to be the closest to my current situation

 

I am having a problem understanding why I cannot get a loco to respond to the RM throttle output when the Elite output does what I expect it to do.

 

We have a Hybrid powered Tomix chassis that is fitted with a Zimo MX 648 sound decoder

 

http://www.ngaugeforum.co.uk/SMFN/index.php?topic=38977.msg528285#msg528285

 

Now in order to get realistic scale top speed of 35MPH I have set the following speed CVs:

 

Start CV2 = 1

Mid CV6 = 40

Max CV5 = 80

 

When Using the Elite controllers rotary knob over the full output range of 0 to 127 steps the Half Moon Speed Indicator increases from 0-100% in line with the throttle settings, and the Hybrid responds as I wish giving me a scale top speed.

 

At 50% indication (left quadrant of the speed indicator blacked out) the Hybrid moves at ½ speed 17mph and at 100% indication (both quadrants of the speed indicator blacked out) the Hybrid moves at 35MPH, and no faster because of the CV5 setting.

 

This is all confirmed by temporarily substituting a different manufactures throttle which has a digital readout of 0 to 127 (Steps) corresponding to throttle setting.

 

 

Now when we add the RM throttle to the mix it all goes awry.

 

With the Use Scale Speeds unchecked so that a generic throttle has a slider range of 0 to 127 and the  Elite Feedback =1 we cannot get the RM throttle to increase the Hybrid speed to more than a crawl ie the RM throttle is at Max (127) but the Elite Speed indicator shows less that half way.

 

We can select the Max speed allowed in the RM loco throttle setup ie 250, set the Shunting speed to ½ of that ie125 but to no avail the output from the RM still will not increase the Elite throttle to full output ie Speed indicator at maximum.

 

What are we missing?

 

What would the Loco setup figures need to be to get 100% RM throttle = 100% Elite output (speed indicator at 100%) and our Hybrid running at its maximum scale speed of 35MPH

 

Also annoying is the actual RM throttle readout is still in MPH (why is this)?

[96RAF]

Firstly think about the command chain...

RM sends commands to the Elite based on what that loco has been configured to. This data is held in RM not in the decoder unless you have set it in that loco CV listing e.g. accel/decell, etc. The Elite turns those RM commands into DCC and sends to the track. Shunt and Cruise are what you set them to in RM or if scale speed is invoked what RM has decided these will be. Selecting scale speed on or  off will toggle the throttle display between MPH to speed steps. You can override the Shunt/Cruise buttons and throttle up to 100%.

 

Using the Elite as direct control you generate the commands by button presses and loco response is based on whatever restrictions are set in the decoder such as accel/decell, etc. and the Elite again generates the DCC signal to the track.

 

These parameters may not be seen the same (by RM and Elite) It is well known that invoking scale speeds introduces problems of its own such as some twin units like Class 395 or HST can play up with it on but will behave with it off, depending upon small or large throttle in use and if first or second large throttle (Pro Only).

 

Regarding feedback this is just a bi-directional data flow that mimics the screen of the Elite to the RM throttle and vice-versa. As the Elite has no speed step indicator other than the arc-dial there is no way of accurately comparing Elite and RM speeds to each other, without some sort of DCC monitor device. Also to be taken into account is which speed curve is the decoder set to on the Elite - 3 point or complex, and is there a motor algorithm that can be changed, and what about BEMF effects. All these would affect 50% on the Elite dial not equalling 50% speed steps or 50% road speed.

 

How to get 35 MPH at 100% throttle both showing as such on the Elite screen and RM. i would think CVs 5 & 6 set by the Elite to the appropriate speed step for 35 MPH. RM should readback the same values. Scale speed off so RM is reading speed steps and no RM programming limitations are affecting the speed curve.

 

Thats about all I can offer based on logic and no guarantees they will work.

Rob

[Chrissaf]

This may well be a question only HRMC can answer, but I don't see it is something I can really send in as a support problem.

 

When you press the loco's shunt or cruise button does RailMaster simply send out a specific speed step value (based on the locos calibration)  and allow the acceleration/deceleration CV  values in the decoder to handle accelerating/decelerating to the specified speed step or does  RailMaster  manage acceleration/deceleration itself by sending out a sequence of speed steps? 

What I'm really trying to understand is the relationship between setting the throttle to a specific MPH value and the interaction with the acceleration/deceleration and speed curve CV settings in the decoder (i.e. CVs 2,3,4,5 & 6)

[BagEndJct]

I'm going to tack onto this thread too as I was inspired to try loco control by program by Ray's comments above.

 

I use programs now for route setting and for platform announcements. Today I thought I'd try and control my Hornby King George I TTS with a test program, mainly to see if I could get it to smoothly accelerate away from stop up to cruising speed. However I could not get the "Accelerate forward [a] to [ b ]" to work. I put 0 in the square brackets for a and 15 in for b. Nothing happened. I also tried placing a "forward to shunt" before the accelerate command but, though the forward to shunt command worked, again, nothing happened with the accelerate command. I think I must be missing something. Not much in the RM guide about this except the explanation that a and b are steps rather than speeds.

 

I would also be interested to learn how others organise their programs as mine have got into a total muddle. Some clues about a naming regime would be welcome. And, do programs have to be placed in the Railmaster main directory or can they go into a sub-directory, like sounds, which I can keep nicely organised in sub-directories? 

 

Thanks, as ever.

(Ha! I just found that putting a b in square brackets causes the text to format as Bold! So a strategic space should help get my text back to normal!)

[St1ngr4y]

Hi BEJ,

Was the accelerate command the last command in the program ? Remember that an accelerate 0-15 will take about 15 seconds to complete. If the program finishes immediately after the accelerate, it won't have had time to start, never mind complete the acceleration. Try putting a stop command for the loco, say, 20 seconds after the accelerate. Also, there will be a definite speed step at which the loco begins to move, depending on the loco and its motor. You may need to increase the start speed to, say, 2 or 3 - worth an experiment. I usually start my sound locos with something like Forward to [2] followed by an accelerate starting at 2 about 3 seconds later. This gives the sound time to complete any brake release etc before the accelerate really begins.

 

Ray

[St1ngr4y]

Hello again BEJ,

I had to delay answering your questions regarding program names in order to pick up my granddaughter from school.

First of all, the programs (currently) must reside in the main RM folder. It has been mentioned in the Desirable Features Thread that it may be beneficial to allow the use of subfolders to organise programs, but due to privacy settings on the RM folder, these subfolders may need to be created elsewhere e.g. in the Documents folder, on your disk drive.

Personally, most of my programs are confined to moving one train from A to B, and I try to make my program names reflect this by containing the loco, start and end points. So for example, I have hidden sidings called HS1 to HS9, I have a station with six platforms numbered 1 to 6. Both the hidden sidings and the station platforms are bi-directional, so the hidden sidings each have a north and south end, and the platforms have an east and a west end. So, for example a train might travel from HS3S to P4W. meaning from Hidden siding 3 South to Platform 4 west. I use the loco running number to identify it, so the full program name would be 60004 HS3S to P4W. I have other named places on my layout also. For example, a signal stop would be called Snnn where nnn is the address of the signal. So, the previous program could be split into two new programs 60004 HS3S to S210 and 60004 S210 to P4W.

Occasionally, I merge two programs together, running different locos in opposite directions around the layout. I simply call these Mnnn, the M meaning merged. Then I keep a list showing which pair of train movements each Mnnn program runs.

Hope this provides you with food for thought.

Ray

[BagEndJct]

Hi Ray

 

Yes, thanks for the thoughts on naming programs. I was using something similar for my announcements programs. There was a gap of a week between starting writing these & carrying on with a few more, and I forgot what my system was. I need to write it all down these days! I think I need to group different types of programs too, so I could have maybe A- preceding all announcements, R- for Routes, O- for operations etc?

 

As an aside, I'm enjoying producing announcements in the program I use for video editing (another hobby). I record "clips" (putting on a strong Bath accent, for that is from where I hail), render the various clips (platform number, destination, arriving or standing at etc), then apply a heavy level of reverberation so that my voice is barely recognisable, and rendering again as uncompressed WAV.  I have a small, cordless Bluetooth speaker under my board and my visitors are quite impressed by these almost unintelligible announcements. Just like what I used to hear while waiting for my train home from school on Bath Spa station!

 

As far as getting my loco to accelerate is concerned, I'll try your suggestions. Not tomorrow as it's "shopping" day in our local town of Mazarrón. So it will be Friday morning, and I'll report back.

 

A bit more info I just remembered. When I test ran the program within the Program window, I saw a green arrow flash up against the 'forward to shunt' command but not against the 'Accelerate forward [a] to [ b ]' command.

 

Thanks.

[St1ngr4y]

If you put your sound files into the RM folder, then you can play them through your pc speakers by running them from within your RM programs. One of the entries I added to the Desirable Features thread was to allow announcement-type messages to be "spoken" from within a RM program. This wouldn't use sound files, but the text entered by the user would be "spoken" by one of the synthetic voices built into Windows. I have done this myself using an Excel spreadsheet, so it can't be that difficult to add it into RM. Rm already uses these voices in its Voice Control subset. Only trouble is, most of the voices in Windows have American accents  ☹️

 

 

 

Ray

[BagEndJct]

I did try that Ray. For most effect, I find it best to apply the reverb to the whole announcement rather than play separate "reverbed" sound files strung together. I does mean quite a bit of work, though, and a lot more sound files than would be the case. 

[St1ngr4y]

@BagEndJct

 

Have you had a chance to retry the accelerate / decelerate commands ?

 

Ray

[Pietrarsa]

Hi all.

Pulling up this thread with measurements made with a dcc sniffer.

I wrote a program to test speed steps for a very reliable loco and this are results:

(in the "note" column the measurement recorded)

/media/tinymce_upload/027bc846db3862b6ccdb0b38e25ba5ee.jpg

 

See discrepancy from values programmed and values reported. Particularly the accelerate and decelerate instructions.

[96RAF]

Are the ‘speed steps’ being set in the instructions column uctuslly SS or MPH/KPH and the reported SS in the Note column are actually speed steps from the sniffer.

Rob

[St1ngr4y]

Hi Pietrarsa,

 

There are a couple of things wrong with your program. There is an extra parameter at the end of the accelerate and decelerate commands which you haven't used (which is OK but...). This specifies the rate of acceleration or deceleration, and it defaults to 1 speed step per second. So if you accelerate from 20 to 50, it will take 30 seconds (roughly) to reach 50. So putting a Forward to [50] command at 40 seconds is (1) too soon and (2) unnecessary. Similarly with the Forward to [20] at 70 seconds.

I advise removing both of the Forward to [xx] commands, move the decelerate command a further 10 seconds on, so that the loco "cruises" at 50 for about 10 seconds, then retry your tests.

 

I think Rob may be right about KPH being involved in the numbers.

 

Ray

[96RAF]

Uctuslly - sorry that should obviously be actually.

Rob

[St1ngr4y]

Uctuslly - sorry that should obviously be actually.

Rob

 

That's actually obvious  😉

[BagEndJct]

@BagEndJct

 

Have you had a chance to retry the accelerate / decelerate commands ?

 

Ray

 

Ray

 

Sorry for my absence. No, I'm afraid I haven't. To quote a much earlier PM: "Events, dear boy, events". I'm afraid a few matters cropped up non-railway related which required some urgent attention. I'll try and get back to the Train Shed over the next couple of days.

 

Thank you for thinking of me!

John 

[BagEndJct]

@S1ngR4y

 

Ray,

 

OK. Even though I've read all of the above, I still can't get my head round the accelerate command. What exactly are the two parameters and could we see some working examples, please? Does it need a stop at the end for the whole program to function? I think you implied that something was needed after the accelerate command?

 

I can't get the King with Hornby TTS Decoder to respond at all but I got another loco fitted with an ESU to do something, but not accelerate!

 

Thanks!!

[St1ngr4y]

OK let us concentrate on the King with TTS decoder. Before running the program, have the loco sound off then try this...

/media/tinymce_upload/44b03fc102e7b0f304ce33c677589d4d.png

Before I go through the program step by step, I will point out that I changed F17 into a latching function by adding "on/off" to the description. The first couple of commands are self explanatory. The Forward to [2] at 8 secs is to get the loco moving and to get the sound files to perform the brake off/steam sound. It might be an idea to try the Forward to [2] command by itself first to see if the loco moves at all. If it doesn't, increase the value until you find the point where it just starts to move. Then the accelerate command is obeyed. If you have changed the Forward to [x] value on the previous command, put the same value as [a] in this command. As I mentioned in an earlier post above, the default rate of acceleration is 1 speed step per second, so the time taken to reach the end of the acceleration is, roughly, b-a seconds. During the accelerate, the loco sound should be chuffing. 10 seconds after the accelerate has finished, the F17 will stop the chuffing, then 2 seconds later the declerate command will start. I have added to optional parameter of 0.5 to the end of the decelerate, so that it will slow down twice as "quickly" as it accelerated, so it should take (b-a) x 0.5 seconds to get down to the final speed which I have set, arbitrarily, to [5], so in this case about 17-18 seconds. Then it will "crawl" at 5 until the stop command at 83 seconds. Then the final commands are self explanatory.

 

Hope this helps

 

Ray

[96RAF]

Ray

Without cranking RM up to check I presume F17 is wrongly listed as a play once function when in fact it should be as you say a toggled function requiring an on/off suffix.

 

One thing to note is BEMF invokes the effect of TTS F17 automatically as throttle is reduced, in that when you slow the loco by hand or via a programmed event it will coast until BEMF senses it has reached the requested speed whereupon it will chuff again as the load comes back on.

 

RM of course requires an end event listed for every event started in a program, regardless of any automatic function occurring.

 

Rob

 

Rob

[St1ngr4y]

I presume F17 is wrongly listed as a play once function when in fact it should be as you say a toggled function requiring an on/off suffix.

 

 

Quite right Rob. I think I had to make similar adjustments with other TTS decoders I have installed.

 

Ray

[BagEndJct]

Ray / RAF96

 

F17 is correctly labelled as on/off toggle command in my King George I sound commands.

 

Ray, I have created a new program, based on yours, ready for testing tomorrow. Looking at yours, I think i may have not been allowing enough time for the accelerate to get going and also, I recognise that I may well have to try different values to get the loco going & adjust the accelerate command accordingly.

 

Thanks for your support.

[BagEndJct]

@St1ngr4y

 

Ray

 

Success! I had to up the parameters slightly to get my King going but then it all worked beautifully. Amazingly, the train almost ended up in the platform where it started! I can now get a bit more ambitious with programs. I have all my announcements to sort out to and..............and..........................

Thank you for all your help.

 

John

[St1ngr4y]

I have great fun developing programs to run the trains. I like to start off a schedule of programs, then sit back and watch, without having to bother with knobs on a controller, or even the throttles in RM. Having points changed automatically, and signals changing at exactly the right time .......

 

Ray

[Pietrarsa]

Hi there.

 

I modified the test program according to your suggestions. I also read in RM manual that "Forward to X" command should be expressed in SpeedSteps, while "Accelerate/Decelerate Forward X to Y" wants Mph/Kmh values (I set Kmh scale speed in RM settings)

In the following image, values reported from the sniffer are visible in the "note" column. It appears RM being "calibrated" to mph scale speed, even if I check Kmh in RM settings.

/media/tinymce_upload/b1b2c7810819e5cd34fadb4c6e4b170f.jpg

 

 

This is the DCC sniffer report for "Forward to" commands (actual SS in note column):

/media/tinymce_upload/8ae5cad093ee0ce68fa6fd14d594a724.jpg

It's clear that speed steps are not involved.

There is another issue regarding top speed. Using RM throttle as well as commands in programs the top SS value is never 127, it always stops at 109. This happens with all locos/dec I use.