How Much Acceleration and Deceleration?

[BCR 570]

I am close to completing the first phase of speed matching my diesel locomotive fleet, which has involved matching scale miles per hour at speed steps 1, 5, 10, 15, 20, 25 and 28.  The fleet comprises Atlas/China RS-3, C-420 and C-630 mechanisms, Hobbycraft Canada C-424 mechanisms, and an Atlas/China GP9.  All engines are fitted with ESU LokSound Micro Select V4 decoders.

Before testing and adjusting each locomotive, I set the acceleration and deceleration to zero.  I am now wondering how much to dial back in.  With a friend over we tested acceleration/deceleration settings of 20, 30, 40 and ended up settling on 55 for the GP9.  This was based on ability to approach and couple onto a string of cars, and stopping distance from 10 smph and 20 smph.  I know that the end result needs to be a compromise between prototypical accel/decel curves, and not taking so long to come to a stop that impedes operating sessions.

The bigger question for me is this:  having speed matched the locomotives at various speed settings, will a common accel/decel setting yield consistent results throughout the fleet?  While the engines have been speed matched at various settings and all have the same decoders, I suspect that mechanical performance will come into play as a potential variable.

I would appreciate hearing from those with advice and/or experience in this field.  if a common accel/decel value is not going to yield consistent rates across the fleet, then I will have to work out a procedure for matching accel and decel.   


Thank you,

Tim

[jagged ben]

Setting all your locomotives to the same acceleration and deceleration should yield common results because as you stated they all have the same brand of decoder .   The function is not part of the NMRA standard, so if you have any different brands of decoders then you'll have to search documentation and do some math. 

For example, Digitrax acceleration is based on settings from 0-32 where each increment stands for 1/10th of a second to increase a speed step.  So 55 is not even a setting with Digitrax.  But you're using ESU, where the setting is 0-255 and every four increments stands for one second that it takes to go from stopped to full speed (or vice versa).  So 55 would mean it takes about 19 seconds to go from stopped to full speed.  Seems high to me, but then you probably don't ever do that in one throttle motion anyway.

As far as the physical part of it, the decoder is just using the acceleration setting to determine when it gets to your matched speed steps.  So I'm pretty sure any discrepancies you notice between locos would be just about as apparent whether you had acceleration turned on or off.  In my experience it's not that easy to keep locos precisely speed matched in any situation, but probably easier for you with a limited number of model types with similar mechanisms. 

[Mark W]

Ben shared some good insight on how the accel/decel formula is calculated differently between brands.  I also found that ESU changed their formula between V4 and V5.  (One uses the EU standard formula, the other uses NMRA, I forget which is which)  @RBrodzinsky helped me a lot when I was programming them.


As far as "how much" to apply, that one is a big fat 'depends'.  Not only on your personal preferences, but which throttle you are using!

When I run knob throttles, I prefer very little acceleration and just dial up/down manually.  However after switching to the Proto Throttle, I dial up acceleration almost to the max!
I've settled on what amounts to about 60 seconds acceleration; that is, your train will go from stop to full speed in 60 seconds*.    With the new V5 decoders, this is the max setting, with the old V4, it's about 60%. 
(*I think this is still slightly on the fast side, but at shows, it's a nice balance when you have impatient crowds watching.)

Another thing to keep in mind is the independent break function.  You don't have to sacrifice deceleration when worried about stopping trains in time.  I prefer to keep deceleration high so that your trains can 'coast' as long as possible, and with the break, I'll still always hit my mark when stopping. 

[RBrodzinsky]

The LokSound V4, LokSound Select, LokSound V5 Multi-protocol and LokPilot V4 all use the EU multiplier (0.25), while the LokSound V5 DCC line uses the NMRA standard multiplier (0.896)

[Cajonpassfan]

Tim, another thing that plays into this that you didn’t mention is the top speed, or available speed range, over which to accelerate or decelerate. I set my freight engines @ step 28=30 mph as I run a mountain railroad, and that’s pretty much as fast as they could ever run. With a dramatically reduced speed range, I find speedmatching and acc/dec rates much easier to match, and even if two locos are a bit off, it’s far less critical.
As to how much acc/dcc, that’s very personal; some guys freak out when they turn down the throttle and their train keeps going. I like lots of momentum on my road engines, especially with decoders that have a brake function. For yard switchers, not so much.
Just food for thought...
Otto

[peteski]

. . .
The function is not part of the NMRA standard . . .

Actually, as Rick mention, acceleration/decleleration timing is a standard defined by NMRA.  But some European brands do not adhere to it.

[jagged ben]

The LokSound V4, LokSound Select, LokSound V5 Multi-protocol and LokPilot V4 all use the EU multiplier (0.25), while the LokSound V5 DCC line uses the NMRA standard multiplier (0.896)

Actually, as Rick mention, acceleration/decleleration timing is a standard defined by NMRA.  But some European brands do not adhere to it.

Interesting.  I stand educated.  However, when I went to research this a little more ... Wow, kind of a mess.  I'm regretting the time I spent on this.   

Anyway, some observations:

Here's what the NMRA standard says: "The formula for the acceleration rate shall be equal to (the contents of CV#3*.896)/(number of speed steps in use)."   This keeps the total the time to accelerate to full speed the same regardless of what the number of speed steps is, which is mathematically the same as how ESU describes it's behavior.  And which I think makes sense. 

How'd they come up with .896 though?  Why not just do 1 second to make the math easier? 

But here's the other thing:  Not all US manufacturers follow the NMRA standard anyway.    Here's the ones I found that do (besides ESU V5):
- NCE (I'm gonna assume when the manual says "each unit = 7mS between speed steps" that they mean for 128 speed steps since 128*.007=.896
- SoundTraxx
- Zimo

Now here's Digitrax:  "As you increase the CV value [for acceleration or deceleration], the rate of speed step change is approximately 1/10 second per increment in acceleration value. ... For example a value of 01 ... means that it would take approximately 2.8 seconds to reach full speed."  So the 'multiplier' here is 2.8 seconds.   Oddly though, the way they word it implies that if you change the decoder speed step mode to 14 speed steps, it would take 1.4 seconds to reach full speed.  I wonder what really happens?   

TCS: couldn't find any documentation that put numbers on anything.