[ErickC]

Csantucci, on 26 September 2021 - 12:24 AM, said:

Eric, try the Unstable release. The compressor should cycle there.

I just tested in U2021.09.26-1236. It seems the problem is not that the compressor isn't running, rather, the compressor is always running whether it ought to or not. I was erroneously assuming that the compressor never started because I was not hearing the air cleaner blowoff when initializing the brakes or when the system had finished charging. It did not occur to me to check the HUD until today for some odd reason.

[Csantucci]

Yes, that's a change that has been introduced. It has been stated that mechanical compressors run continuously when the engine is on.

[ErickC]

Csantucci, on 26 September 2021 - 12:38 PM, said:

Yes, that's a change that has been introduced. It has been stated that mechanical compressors run continuously when the engine is on.

They do and they don't. I think there's a fundamental misunderstanding here in both the way mechanical compressors actually work and the purpose of the sound triggers here. It's much more complex than "the compressor always runs" and the sound triggers reflect the more complex reality and are much more appropriate for the simulation.

Prime mover driven mechanical compressors are always "running" in the sense that the parts are always moving, but are in actuality not always "running" because they don't always compress air. There is a major difference in the sound between these two states. In this case, "on" simulates a condition where the compressor is compressing and "off" simulates when it is not (see: http://www.rr-fallen...l/f40sm-s05.pdf for more details on how this works on EMD locomotives). This allowed us to simulate the very audible difference. As it currently stands, it now sounds like it's always running and the blow-off does not operate as it should (when the system stops charging). This is completely inaccurate. So one of four things needs to be done:

1. We need a new set of triggers to reflect the "pumping or not pumping" state and every SMS file ever made for any locomotive needs to be updated (as you can imagine, this would be a prolonged undertaking)
2. The current triggers need to be adapted to reflect the "pumping/not pumping" state rather than the state of the compressor "running" or not
3. Acknowledge that the way the sim was set up previously more accurately depicts the output of the system rather than whether or not the parts are mechanically moving, which is relevant to real life if you are a maintenance tech who needs to lubricate the compressor and keep it working but not relevant in the simulator where the only purpose of the compressor being "on" or "off" was to determine if the brake pipe should be receiving air (which the sim already knows) or whether the sound of the system actually delivering air should be playing (which is now broken).
4. Sound developers need to delete all compressor streams because the sound should not always be playing. This would be a major step backwards in realism.

Options 2 and 3 are functionally interchangeable. Either would be fine.

TL;DR: whether the compressor parts are moving or not is irrelevant to the sound simulation. All that matters is whether the compressor is pumping air because that's what determines the sound that it makes. We used to be able to simulate that correctly. Now we can't.

[railguy]

Erick hit the nail on the head. Option 2 or 3 would seem workable to me. The other parameter that would be nice, as I noted elsewhere, would be a "MainReservoirLeakRate" parameter in addition to the current TrainPipeLeakRate. There are specific rules on how much of train pipe leak is permissible and it's not a lot. When one hears a locomotive air compressor running, say, when a locomotive is simply parked, it is more likely because the main reservoir has some leakage. As an aside, both leak rates should affect ALL trains--player, AI, and loose consist--with sound triggers in all responding accordingly.

[Csantucci]

I see your points, and to me also solutions two and three are the best ones. IMHO for stable release 1.4 the best solution would be to revert all changes inserted about the mechanical compressor (which should be solution 3 mentioned above), and to use the Unstable Release to work further on the mechanical compressors, basing on solution 2 for backwards compatibility with existing .sms files.

[steamer_ctn]

Csantucci, on 27 September 2021 - 12:28 AM, said:

I see your points, and to me also solutions two and three are the best ones. IMHO for stable release 1.4 the best solution would be to revert all changes inserted about the mechanical compressor (which should be solution 3 mentioned above), and to use the Unstable Release to work further on the mechanical compressors, basing on solution 2 for backwards compatibility with existing .sms files.

Perhaps the short term solution is to put the sound triggers back to what they were (ie stopping/starting air flow into the reservoir), and leaving the feature that the air fill rate would vary if it is a mechanical compressor due to engine rpm.

This would be a relatively simple temporary fix for the purposes of v1.4.

Would this solve the issue for v1.4?

[Csantucci]

This is solution 2 envisaged above. This solves the sound problem, but I wonder what state you display in the HUD when the compressor is running but not pumping.

[sim-al2]

I think the previous status quo where it simply displays "Off" is technically accurate enough, as for HUD (and sound trigger) purposes the compressor basically is off, even if in reality it's merely unloaded and not generating any air pressure.

[Weter]

Compressor idling/compressor actuated

[steamer_ctn]

Csantucci, on 27 September 2021 - 02:30 AM, said:

This is solution 2 envisaged above. This solves the sound problem, but I wonder what state you display in the HUD when the compressor is running but not pumping.

A further patch has now been committed so that the compressor effectively operates on the original "electric" compressor code, so in effect, as far as turning on and off, it should be the same as it was. HuD display should indicate as per the electric compressor.

The only enhancement retained is to vary the charging rate with engine rpm for mechanical compressors (appropriate parameter needs to be set in ENG file).

As suggested the compressor code will require a more detailed review at some future stage. For example, the current electric compressor code operates the compressor regardless of whether the diesel engine is running or not.