Okay, then I understand it
Quote:
Okay, then I understand it correctly. FWIW, TCS WOW sound decoders have a brake function, but each press of the brake function key applies 20% braking.
I do remember someone mentioning this a while back. Thanks for the reminder. I don't have any TCS decoders at the moment so my experience with them is about zero, so this is good to know.
Quote:
So I'm assuming you could send the appropriate number of function presses per how far the brake lever is pulled over with a TCS decoder.
Yes and no. While it is certainly possible to do this, there are some practical issues. This is the same reason using manual notching to control the sounds seems appealing at first, but ends up being problematic in the end. Fundamentally, with the ProtoThrottle, the brake (and throttle) levers are absolute inputs - they are in a certain position, remain in that position, and expect the decoder to be in the corresponding state. On the other hand, the TCS brake function (and manual notching) are relative controls - each activation advances the control up or down but there is to way to say "go to 45% brake" or "go to notch 6". Therein lies the problem. The ProtoThrottle has no way to know if the decoder is actually in the desired state (and take action if it is not), nor any way to reinforce the state to the decoder.
So, for example, let's say we are using the common example brought up on these forums of "just use manual notching" to decouple sound from speed. Let's say you are moving along in notch 6. The locomotive briefly loses power and loses its brain. You'd expect, when it comes back on line, to be moving again in notch 6. However, the PT already sent the manual notching commands to put the decoder in notch 6. It cannot keep resending those "notch up" commands without adversely affecting the state of a good decoder. It also has no knowledge that the decoder in question lost its brain. There is just no way to reinforce that the decoder should be in notch 6.
Speed, on the other hand, is defined by DCC to be an absolute control. The desired speed can be (and is) continually refreshed to the decoder so when the decoder comes back on line, it resumes the previous speed. There is simply no way to do this for relative notches, or in the case of this thread, relative brake steps.
Yes, there are ways to somewhat work around this. Things like sending eight notch up commands when the throttle is put in notch 8, eight notch down commands when the throttle is put in idle, or a bunch of brake down commands when the brake handle is fully released, but that only resynchronizes the throttle and decoder when the PT control enters that position AND the decoder is ready to accept the command.
These relative controls on the decoders will make the PT interface less than ideal. I've mentioned this to the decoder manufacturers, and I think they understand, but we're not going to see any quick fixes soon. So, my only option left may be to build some wizardry into the PT to try and emulate the desired behavior. Maybe using manual notching and/or TCS style braking will end up being the solution, with the knowledge that there are some corner cases, like mentioned above, where things can get out of sync. In the meantime, my opinion is that using absolute (on/off) functions that can be reinforced periodically is the best, most reliable solution and gets most of they way toward the proto-experience. It's also the most universal among decoder manufacturers.
Quote:
I'm curious how this differs from what SoundTraxx has implemented in their Tsunami2 for braking. It seems like LokSound braking may be the most rudimentary at the moment across the three leading sound decoder makers.
The Soundtraxx brake is just on/off, at least the form of it that I was using. There may be other modes, like the TCS, that I'm not aware of yet. This is the same as the Loksound brake. The main difference between the two was the feedback that the user got from the Soundtraxx decoder, both in terms of the brake's effect on speed and on sound. The ESU decoder, at least as configured, did not provide the immediate feedback that the brake had been applied. Once you got used to the Loksound brake, it worked well, but it was clear the Soundtraxx decoder had a shorter learning curve with the brake than the Loksound.
Michael