[copperpen]

Two eng files in the zip, both cars are in the same DMU as a power twin set. In manual no go, in automatic runs ok.

I have sent you a message with a link to download the DMU. If I give you a link to the original, it would be quite large because it contains 7 or 5 dmu sets.

Attached File(s)

•  BR_Class105_DMBS_M50775_BRBfy_eng.zip (9.05K)
  Number of downloads: 182

[Csantucci]

First of all I am astonished by the number of people who is using gear-based engines! I thought to be almost the only one...
These are the problems I found with release 1667 (and advanced adhesion model off):
1) if the consist is made by units with different .eng files (but with exactly the same parameters) the "helper" units go out of sync in short time, see image:

if the consist is made by identical units (same .eng file) the problem does not appear
2) (this has been noticed also by other people): when accelerating in the first gear, at about 15 Kph, RPMs reduce significantly and then rise again;
3) gear-based units are not recognized if following parameters are missing in the .eng file:
GearBoxBackLoadForce( 4500 )
GearBoxCoastingForce( 2000 )
4) it would be good if gears increase/decrease only by one notch every time the related key is pressed.

[BB25187]

Hi all,

On my side, I kept on testing things. I especially tried to understand why the power of some engines was oscillating and became negative. I found at least one cause: in some engines, the parameter DieselEngineMaxRPMChangeRate was declared as follows:

DieselEngineMaxRPMChangeRate       ( 999999 )

This may have been motivated by some fancy MSTS behavior?
Changing it to a more reasonnable value (typically: 150) resolves the problems observed so far with the engine under consideration. This works with both standard and advanced adheasion systems.
Unfortunately, the engines provided by copperpen do not suffer from the same problem with DieselEngineMaxRPMChangeRate parameter. Nevertheless, the MaxPower parameter set to 1kW seems a bit unusual too...

Another things which does not behave the same with OR and MSTS is the following one: to mimic some old transmission systems, which imposed a double clutch operation (I don't know the exact word in English), the gear control was written as follows:

        Gear ( 0 4 1 0 
            NumNotches ( 8 
                Notch ( 0   0 Dummy )
                Notch ( 1   0 Dummy )
                Notch ( 0   0 Dummy )
                Notch ( 2   0 Dummy )
                Notch ( 0   0 Dummy )
                Notch ( 3   0 Dummy )
                Notch ( 0   0 Dummy )
                Notch ( 4   0 Dummy )
            )
        )

As you notice: the control of this 4 gear engine was in fact modeled with 8 gear notches, some kind of dummy neutral positions being interleaved with actual gears. MSTS honoured those neutral intermediate positions. OR just ignore them.

Regards

[Matej Pacha]

 Csantucci, on 27 June 2013 - 11:57 AM, said:

First of all I am astonished by the number of people who is using gear-based engines! I thought to be almost the only one...
These are the problems I found with release 1667 (and advanced adhesion model off):
1) if the consist is made by units with different .eng files (but with exactly the same parameters) the "helper" units go out of sync in short time, see image:
if the consist is made by identical units (same .eng file) the problem does not appear
2) (this has been noticed also by other people): when accelerating in the first gear, at about 15 Kph, RPMs reduce significantly and then rise again;
3) gear-based units are not recognized if following parameters are missing in the .eng file:
GearBoxBackLoadForce( 4500 )
GearBoxCoastingForce( 2000 )
4) it would be good if gears increase/decrease only by one notch every time the related key is pressed.

Ad 1) I'll have a look at it. It can happen if something within a copy constructor is forgotten
Ad 2) Needs more investigation too. There are two main reasons: the diesel overload algorithm and the clutch algorithm
Ad 3) OK, let's say if these parameters are missing, zero values will be used
Ad 4) This needs a little help from other OR team members - it's on the TODO list

Thanks again for the investigation!

[copperpen]

The 1Kw figure is in there as supplied. It works just fine in MSTS. I think the 1Kw figure has something to do with unpowered cab cars to get a balanced power curve and first came to light with US cabcars a few years ago. Some DMU sets are made with one power car with a driving trailer. The power twin sets were made for extra acceleration and better hill climbing with a full load. On that basis the 1Kw should not be there, but then the set will have to be balanced in a different way. I would prefer to get the gearbox ( manual ) problem sorted before moving along that path.

[Csantucci]

Peter Gulyas (gpz) (thanks to him!) just committed (rel. 1669) a patch by myself extending sound features partly only for gear-based engines and partly for all locos/trainsets. It contains following items:
- restore of events 23 - EnginePowerOn and 24 - EnginePowerOff for diesels;
- Variable1 proportional to throttle for gear based engines, to allow for a variable that can be used for volume curves (motor sound volume linked to effort)
- addition of events 68 - GearUp and 69 - GearDown for gear-based engines (missing in MSTS)
- addition of events 70 - ReverserToForwardBackward and 71 - ReverserToNeutral (valid for all loco types): MSTS has only a trigger that indicates that the reverser has changed position, but you don't know the position reached, which is sometimes necessary for sound
- addition of events 72 - DoorOpen and 73 - DoorClose (valid for all loco types, missing in MSTS)
- addition of events 74 - MirrorOpen and 75 - MirrorClose (valid for all loco types, missing in MSTS); for the many locos without mirror animation it is a useful general-purposed couple of sound triggers.

As already stated, for gear-based engines Variable2 is proportional to RPMs.

[copperpen]

OK. In an attempt to see why these UK DMU sets are not behaving as expected using the manual gearchange, I made sure that both eng files were identical in all power and gearbox parameters, and corrected the wheelradius as well. The result is that the PowerTwin set I am using as the test set will now move and accelerate normally, however, the rear engine is not putting out the same levels of power as the lead unit. When I hit 4th gear the power levels tail off, but that may be caused by too low a setting for TE in that gear.

Therefore, the problem is, why does the trail unit not generate the same power level as the lead unit?. Modified files attached.

Edit: While using the Power Twin set again, I notice two other things that are not quite right. I stopped the train to change cabs, and put the gears to neutral. In the second cab, gears were in first, so I moved them to 4th and went back to the lead cab, where gears were still in neutral. Seems to be a mismatch in gearshift between units in the train.

Second item is diesel RPM. This looks to be tied to the train/gear speed when it should be linked to the throttle setting. It is rather strange to pop outside the cab and hear an engine idling when the HUD says it is doing 1800 RPM.

Attached File(s)

•  BR_Class105_DMBS_M50775_BRBfy_eng_2.zip (9.04K)
  Number of downloads: 183

[Csantucci]

Release x1670 is available, where the GEARS cabview animation works now.

[James Ross]

 Csantucci, on 27 June 2013 - 10:13 PM, said:

- addition of events 68 - GearUp and 69 - GearDown for gear-based engines (missing in MSTS)
- addition of events 70 - ReverserToForwardBackward and 71 - ReverserToNeutral (valid for all loco types): MSTS has only a trigger that indicates that the reverser has changed position, but you don't know the position reached, which is sometimes necessary for sound
- addition of events 72 - DoorOpen and 73 - DoorClose (valid for all loco types, missing in MSTS)
- addition of events 74 - MirrorOpen and 75 - MirrorClose (valid for all loco types, missing in MSTS); for the many locos without mirror animation it is a useful general-purposed couple of sound triggers.

Where were these extensions to MSTS file formats discussed and agreed upon? Please can people not extend MSTS file formats without consulting the rest of us. These extensions need to be formally proposed or removed; to propose them, please start a new thread and explain:

• Why they are needed (admittedly, fairly obvious in this case) now, as MSTS file format extensions and not OR-specific files.
  
• How successfully MSTS ignores them.
  
• How successfully MSTS Bin ignores them.

While I'm all for small enhancements like this, they should be properly considered - in public - so that we don't, for example, get people authoring new MSTS files that they believe are compatible with OR and MSTS but in fact just crash or misbehave horribly in MSTS.

Thanks. :)

[copperpen]

 Csantucci, on 28 June 2013 - 11:12 AM, said:

Release x1670 is available, where the GEARS cabview animation works now.

Thank you to the team.