[Doug]

I was reading the manual of the DCC controller that I have been recently playing with. I see that it has 28 speed steps with pre determined speed values. These values from 2 to 255 look like this:

Obviously programmed this in for a train.

This is the acceleration curve of a Porsche Carrera:

So we need to reprogram the DCC chips for cars.

Something like this:

For better acceleration.

The programming is done by changing values of CV 67 through to CV 94. This can be modified for each chip in each car so one can adjust the acceleration characteristics based on the original car.

 

[astro]

good find!

Don't forget the DCC supplies voltage or current to the train motors, a given voltage/power for each step, which has been chosen to result in those speeds.

However, a train has a different inertia, mass, motor and gearing to a car, so the transformation from the train graph to the porsche graph will not be simple.

Also to get the porsche graph for a given slot car will depend on the motor, mass, gearing etc of that particular car, so different slot cars will need different graphs, in an ideal world.

 

[RC45]

QUOTE (astro @ 27 Jul 2004, 00:28)good find!

Don't forget the DCC supplies voltage or current to the train motors, a given voltage/power for each step, which has been chosen to result in those speeds.

However, a train has a different inertia, mass, motor and gearing to a car, so the transformation from the train graph to the porsche graph will not be simple.

Also to get the porsche graph for a given slot car will depend on the motor, mass, gearing etc of that particular car, so different slot cars will need different graphs, in an ideal world.
And the same would be true for every single electric train model out there - there are many gear/rpm/weight combos.

Prototype vs scale weights are also so far apart that even in model railroad speed table udjustsments it is a trial and error process.

This is not a "train vs car" thing - but simply, a "how best to simulate this car vs that car and scale speed" thing..

For fear of repeating these things, but this is one of the first custom configurations I have spoken about when ever I arrived at a slot car forum.

 

[Doug]

QUOTE (RC45 @ 27 Jul 2004, 01:36)For fear of repeating these things, but this is one of the first custom configurations I have spoken about when ever I arrived at a slot car forum.
I didn't see that Paul, not here I presume...
Please don't hesitate to repeat yourself if it will make things more clear.

What values have you used to reprogram the Speed steps?

This whole exercise explains why the car was not accelerating and decelerating correctly using the default settings. I must now go and reprogram the chips.

This is what is should (roughly) look like:

Here's the data:

CV#	Step	TrainValue	Car Value
67	01	002	002
68	02	004	040
69	03	006	066
70	04	009	090
71	05	013	110
72	06	018	124
73	07	024	140
74	08	031	153
75	09	039	164
76	10	048	173
77	11	058	182
78	12	069	190
79	13	080	197
80	14	090	205
81	15	100	212
82	16	110	220
83	17	120	226
84	18	130	230
85	19	140	235
86	20	150	240
87	21	160	245
88	22	170	248
89	23	180	252
90	24	195	254
91	25	210	257
92	26	225	258
93	27	240	259
94	28	255	260

 

[Xlot]

Two comments :

A ) The "Speed" command is really a "Volts to motor" command. A train/car will accelerate until it reaches equilibrium @ that voltage. I believe there is a fundamental difference here, trains would seem to be at equilibrium most of the time, while slot cars are notoriously always accelerating unless you have a 10 m long straight and short gearing

B ) One should consider that a slot car has no gears and that a brushed DC motor's torque vs. speed curve is totally different from an internal combustion engine (max torque @ 0 rpm, then decreases linearly)

So, my reasoning (I might well be proven wrong) is that the latest curve would give you too few "low" steps - where the motor response is torquey - and too many "high" steps - where response is sluggish

 

[Doug]

Xlot, give us a sketch of how it should look please.

 

[Xlot]

Hey, you know I'm semi-literate - it took me six months to learn how to post pictures !

I would say the default curve is more appropriate
Rather, have you set Vo=start voltage=CV2 ?
Also, have you tried going to 14 speed steps ?

Ciao
Beppe

 

[astro]

I don't understand what you are trying to do with these graphs, or how they are used with DCC.

the DCC has no idea about what speed the car or motor is going, it can just give voltage, and perhaps detect current draw.

I havent seen anything on the market for slot cars which would give acurate enough slot car telemetry to tell what its acceleration curve was. RC45's method of experimentation is more likely to give perfect results, imho.

Even a speed trap, so you could send the car off at different DCC constant steps, and tell what scale speed that gave, would not show what that step was corresponding to under acceleration.

If the graph is something you can input into the programmer, the new graph WILL give better response than before, but it is unlikely to have anything to do with the speed graph of the resulting car performance.

Also, the porsche acceleration graph response is for someone flooring it and going up the gears. I THINK (but you have the manual!) that the train graph shows constant speeds resulting from each step of the controller. Ie making the travel of the speed knob or trigger smooth and responsive all the way along. If this is what that graph represents, then when you floor the controller, it will applt voltage step 28 straight away (or dependent on some other graph). The reason I dont think this is the graph for that is because it is a graph of steps to speed, and there is no mention of time.

 

[Xlot]

Some confusion here - not helped by DCC manuals quite improperly using "speed" and "power" instead of " Volts to motor"

Let's recap :

- CV 2 sets the voltage @ step 1 (just enough so the car moves). This is a % of available (decoder inlet minus rectifier drop) voltage, and for some NMRA reason is expressed on a scale from 1 to 255

- CV 5 sets voltage @ full throttle (step 14, 28 or 128 to your taste) - again as a % of available voltage and again on a scale from 1 to 255. So if you want your kids playing safely, you set it as say 190 [for the same reason, IMO "racing" decoders should have this nailed at 255, forget yellow flags and handicapping]

- the curve we've been dealing with assigns a voltage to each intermediate step, so yes Astro it's related to trigger response - one would mess with it according to driving style, motor winds/gearing and type of track [Doug, if you are bypassing CV 2 and CV 5 and establishing the entire curve directly, it should start with 50-70 @ step 1 - but still "sag" as it goes up to 255 @ step 28]

Now we can go back to the more confortable realm of electro-mechanics : once a voltage has been set, a motor has a well known torque/speed characteristic, while the car has a given inertia and a (less well known) drag/speed curve - it will accelerate and reach steady speed on that basis

 

[Doug]

CV2 is minimum speed: set to 0 (or no volts)
CV5 is maximum speed: set to 255 (full voltage applied)

CV3 is acceleration: value 0 to 255, one unit corresponds to 32/1000 seconds, the time it takes until the loco/car decoder internally switches over to the next higher speed step.

CV4 is brake delay: value 0 to 255, one unit corresponds to 32/1000 seconds, the time it takes until the loco/car decoder internally switches over to the next lower speed step.

Note: initially my CV3 did not equal my CV4 and turning the speed knob say a quarter turn to the right would take it to a certain speed, but bringing it back by a quarter would not bring it to where it started. This is very disconcerting for slot cars as the control is much more erratic than trains.

CV67 to CV94 are the speed steps. Hence the graph.

Within CV29 there are configuration settings such as direction, speed steps (27 or 28), whether or not to allow analogic running and and whether or not to read from the speed table (CV67 to CV94) or whether to calculate it automatically.

The 14 speed steps is an old setting, now widely replaced with 27 or 28 steps.

To come back to what I was saying:

I think they are important and the shape is relevant to the vehicle. One can say that the voltage applied is in proportion to the speed. No volts, no speed; max-volts, max-speed - it's not going to go faster.

Think of a big steam train: The fireman opens her up from a standing start at the station; chu, chu, chu; The Revs slowly increase; it slowly starts moving gathering speed as momentum gathers and the cylinders start working to the optimum.

Think of a car: Light and nimble; hurtling off the start line at high revs; peaking 20 to 30 seconds down the road going flat out. We measure acceleration within the first 60mph/100kph because after that it really does take a while to go still faster.

 

[Xlot]

QUOTE CV2 is minimum speed: set to 0 (or no volts)

Respectfully, it's not speed, it's voltage - at which step does the car start to move ? You should see which "speed (grr) value" this corresponds to and input that as CV 2 (or CV 67/step 1) in the new curve

 

[Doug]

Understood. I'll try it out.

 

[astro]

QUOTE I think they are important and the shape is relevant to the vehicle. One can say that the voltage applied is in proportion to the speed. No volts, no speed; max-volts, max-speed - it's not going to go faster.

The graph will be important and make a difference. But if this is in fact a volts to steps graph, then

a - it has nothing to do with what the car does when u floor the throttle (CV3 will, but not the graph; In most cases you may want CV3 delay = 0, to use the car's natural gearing and torque etc.)

b - changing the curve is more like chnging from a 45 ohm to a 20 ohm resistor, or making linear to log curve changes with potentiometers. It is more to do with how much control/throttle distance you have over the car at different speeds.

The porsche acceleration curve is a red herring and not related to what the graph represents

 

[Doug]

QUOTE (astro @ 27 Jul 2004, 19:18)The porsche acceleration curve is a red herring and not related to what the graph represents
I beg to disagree. It represents the convex shape of a vehicle that has a better power to weight ratio than a heavy cast iron train that need momentum to get moving.

The correct graph may need to be flatter or more pronounced, but I'm sure that it's convex as opposed to concave or linear.

 

[Xlot]

I believe Astro is saying that, if you really want to match the Porsche's acceleration curve, you need a quite different set of tools - mainly one or more speed traps (Alberto Elli makes them as an accessory to Wincrono) - and then work on motor and gearing

But first you would have to decide what scale is the track (we all know lengthwise it's not 1:32) and ditto for the speed

 

[RC45]

This is the way I present it.

I have 4 Ferraris - a Carrera Enzo, a Policar F355, a ProSlot 360 and a ProSlot 360 Challange.

The Carrera has it's stock electric motor, the F355 and ProSlot 360 Challange each a Slotit V12 and the ProSlot 360 has the motor from the ProSlot 360 Challange.

In this scenario, the Enzo should be the fastest at full throttle.. so I determined what I would like it's max voltage to be.- which would hopefully be the max speed - an appropriate 212mph.

Then I compared (all by eyeball I might add, considdering I don't have a long straight) how the other cars rated at their new max voltage to give them top speeds ranging from 180mph to say 165mph for the 360 Challange (the lower top end representing the lower gearing a track 360 may have over it's road counterpart)

Then I spent some time jacking around with acceleration and brake delay to simulate the fact that the 360 Challange would have awesome brakes - in essence no brake delay so that rolling the throttle closed had an appropriate instant reduction in volts.

A similar result is possible with assigning different voltages to the same speed steps for the differing cars.

This would simulate the shorter time in which the Enzo would reach the same "speed" (different voltages may be required by different electric motors to show the same "speed" for 2 different cars) than it would take an F355.

Make any sense?

 

[astro]

I think RC45 has the only way to work this.

The porsche curve is speed against time

The DCC graph is 'speed' (or probably voltage) against throttle position.

I agree the shape of this graph should probably be convex like the porsche curve, but this is due to voltage overcoming inertia and friction, not to do with acceleration. But any use that these graph shapes have is purely coincidental.

 

[Tropi]

A fascinating discussion but . . .
inevitably raises the question as to whether chip programming skills might displace driving skills in competition.

The implications for 'slot racing as we know it' are intriguing.

 

[Xlot]

Hi Tropi,

I wouldn't worry about the speed curve - after all, it's very much like what committed racers already do with their electronic (say NSR ) controllers, or even screwing/unscrewing diodes on PM ones

I do worry a lot about the yet untapped possibilities : using back-emf for traction control, and overriding general speed limitations (yellow flag, handicaps, etc ) one might think of - that's why I'm basically suspicious of "bells & whistles"

Ciao
Beppe

 

[MarcusPHagen]

Hi, Xlot:

As for "untapped possibilities", I'd be curious to know whether the chips can be reprogrammed "on the fly". That is, must one use a dedicated controller & track segment to reprogram a car, or is it possible that someone with a hacked controller could send signals or instructions to other vehicles? If so, this introduces some real "black hat" scenarios! For example, what if I don't need to nerf another car, but simply send out a series of random "full power" signals for a range of ID numbers? If that's too obvious, then how about intermittent "brakes" or "lower speed" in the same way?

Marcus

PS: despite my participation in the "Cheat" thread, I don't see this as anything I'd want to be doing. However, I think that it is at least necessary to determine whether such things might occur, and/or how they could be prevented. mph