Has anybody noticed this? I just put a LokPilot V5 in a Kato SD70MAC (HO scale) and it is noticeably louder. There is a distinct motor growl that none of my other models really have. It sounds like the old Bachmann decoders with RF suppression circuits. I turned off Back EMF and the famous silent Kato drive returned. I tried playing around with the settings, including the CV54 autotune feature that 'optimizes' the BEMF. Nothing really changed. The model runs very smooth with BEMF enabled, but the sound is pretty annoying, especially because it is not masked with a sound decoder. I think most people turn off BEMF to run consists anyways, but I just wanted to figure out why this is happening
I wonder if the BEMF control loop period (frequency) is in the audible range? If so, and it is oscillating about a given operating point, that could be your noise. Not sure what controls you have over the BEMF function, or whether they would help stabilize the loop. It seems like an auto-tune feature, if successful, would avoid the oscillation/noise.
There is a setting for Motor PWM Frequency, but I am not exactly sure what it does. I saw mention of it in another online thread about coreless motors. They were saying how coreless motors do not operate well on DCC, so the frequency needs to be turned up. I tried messing around with it and put in a few different values, but I could not hear a difference in the motor.
Motor drive PWM frequencies usually cause a very high pitched hiss or squeal, if not above the human hearing range anyway. What I was referring to was the BEMF control loop frequency, which is the frequency at which the BEMF processing makes changes to the PWM duty cycle (or drive strength for the motor) to keep the locomotive speed stable across varying loads (slopes, number/weight of cars, etc.) If the BEMF control oscillates, it can do so at or below half the BEMF control loop frequency, and if in the audible range, it could produce a growling noise. The BEMF control loop frequency is much lower than the PWM frequencies If nothing else, I would back off on the strength/gain of the BEMF control, to perhaps avoid the oscillation. If doing so eliminates the growl, then BEMF oscillation was likely creating the growl. Different decoders have different BEMF control loop schemes, and different registers that can be tweaked to get the behavior you want, so there may be other options for keeping the BEMF control from oscillating.
I think I see what you are saying. On the ESU decoder, I think that is called the sample time. You think I should lengthen the sample time? It sounds like there may be some kind of resonant frequency event going on here, like the BEMF sample lines up with the PWM pulses. I have been working on the Soundtraxx decoders, which are not having the same problem. I'll try it out tonight.
I'm not familiar with ESU decoders, but that seems plausible. If the pitch of the growl changes, then that is likely it. Whether it can be changed far enough to put it out of the audible frequency range while still functioning acceptably is another question... Before sound decoders, we might have even found the growling satisfying.
It seemed to have worked a bit. I lengthened the sample times and changed a few other settings. It is still slightly audible, but not as much as it was before. Coupled to a locomotive with a sound decoder, it sounds fine.
It is also possible that the growling was a resonance with something else on the locomotive (likely part of the shell). If that was the case, then even a slight change in frequency could kill the resonance, and render the noise much quieter..
