[Matej Pacha]

BB25187, on 24 June 2013 - 10:37 AM, said:

So I have two questions:
- Did you succeed in moving the Kiha?
- Is there any new operation to perform to operate such an engine in OR?

Well,
yes and, no. May I ask you for the FPS information (F5 HUD view)? If you are running at low FPS, you may get the oscillations. You can try to disable the advanced adhesion model by menu setting or by pressing Ctrl+Alt+X. I'm afraid any low-end PC with FPS below 15 should always run with the advanced adhesion disabled.

Anyway, I've succesfully tested the KIHA31 (default) with no problems. I've also tested some Czech gearboxed vehicles with both manual and automatic gearbox and all worked fine.

Thanks,

Matej

[BB25187]

Hi,

Matej Pacha, on 24 June 2013 - 11:58 AM, said:

Well,
yes and, no. May I ask you for the FPS information (F5 HUD view)? If you are running at low FPS, you may get the oscillations. You can try to disable the advanced adhesion model by menu setting or by pressing Ctrl+Alt+X. I'm afraid any low-end PC with FPS below 15 should always run with the advanced adhesion disabled.

Anyway, I've succesfully tested the KIHA31 (default) with no problems. I've also tested some Czech gearboxed vehicles with both manual and automatic gearbox and all worked fine.

Thanks,

Matej

As for the FPS, it is stable and pretty high: around 60. But in fact I set the VerticalSynch to true. When I set it to false, the FPS is much higher.
I will try to disable the advanced adheasion model and let you know. In any case the fact that the same operations may lead to different behaviors is puzzling, somehow!

Regards

[copperpen]

While I can get the default Kiha to move it requires at least 70% throttle and a wait of over a minute, and eventually heaves into motion. This behaviour is similar to the current steam engine physics where a light engine requires at least 60% forward gear and the same or more of the throttle just to get moving, just a bit on the unbelievable side.

My British power twin however will not get above 0.7 mph with full throttle, where the max first gear speed is 14mph. Change the operation to automatic and it is away like a racing whippet. In manual operation, these DMU sets should start to move and accelerate on the first notch of the 5 stage throttle.

FPS on the route with the power twin are 63, on the route with the Kiha twin, 156.

Just tried the power twin with advanced adhesion disabled and managed to get to 2 mph in 4th gear at full throttle. Although the rear unit was running, no motive power was observed in the force information window.

[Matej Pacha]

Now I see - it is the adhesion model what causes the problems. I've tried something to eliminate another negative effect and this came with v1664. Please be patient, I can do some investigation how to get the best from it. In mean time, please use the simple adhesion model instead.

Thanks

EDIT: Try v1666 - it should be fixed or improved.

[copperpen]

No improvement in v1666



This shows what is going on. Note the negative values for the lead unit. The strange thing is, if I change the two cars from manual to automatic gearbox, they work flawlessly.



Note the lead car still has negative motive, but the rear car now has a lot of positive

.

[BB25187]

Hi!

In a sense, I am happy that somebody else is facing the same type of issue: it may be easier to understand the point.
No problem: we will be patient, and we perfectly understand that the tuning of this feature is not easy.
After the suggestion you made yesterday, I disabled the advanced adheasion model. The global picture is improved, but remains unperfect:
- The Kiha31 is operated in conditions which look very similar to MSTS ones.
- As for the other two engines I already tried, the behavior remains a bit strange: I still observe some oscillations and negative power when speed is low and when using gear 1. But suprizingly, the problem becomes negligible when the speed increases, and using gear 2 to 4. We can switch to the more stable mode not only by increasing speed and switching to gear 2, but also by stabilizing the oscillations thanks to a temporary reduction of the throttle.
I noticed that you delivered a tentative fix at rev #1666. I hope I have time to test it today. Otherwise I will let you know tomorrow.
Thanks anyway for your work!

Regards

[BB25187]

Hi,

Unfortunately, I confirm that revision #1666 does not significantly change the situation: oscillations with negative power are still to be observed, whatever the adheasion model option.

Regards

[Matej Pacha]

Hi guys,
1666 is not the right number, try v1667 :unknw:

Thanks

[copperpen]

1667 does not fix the problem either. It is not a problem with changing gears, that works properly in both auto and manual mode. What seems to be happening is to do with how Motive is applied to the gearbox model. When in manual mode the lead car has relatively high negative Motive, and any other power cars in the DMU have a lower negative value, all move forward but not very fast. In automatic mode, the lead car still develops negative Motive but the trailing cars develop positive Motive to three or four times the value of the lead negative and so the train moves forward at a far more rapid rate.

All rather strange, because it is only the change of one word in the eng file.

[Matej Pacha]

This is weard - I just tested all three types of gearboxed vehicles I have, manual and automatic and worked fine - I was able to start the train even with 10% of throttle. Can you please provide the eng file? Or a link.

Thanks