My Aristo RS-3 has separate connections for the heater and fan. It appears that they both just got full track voltage. I’m wondering what the appropriate behavior would be, since I have an opportunity to customize it. For example, would the heater run continuously, and the fan just come on during hard acceleration? Or can I just tie them together and expect the heater to heat up quickly enough to get a puff of smoke when I fire them both up? I’m thinking about the steam engines that have chuff sensors and have both a fan and heater, but don’t have any experience with this.

The electronics to sense the acceleration are not an issue. I’ll build something custom for that. It’s the implementation of the actual smoke unit that I’m curious about.

Not modulating the heater and having slow speed will equal melted plastic.

The right operation is to sense load and motor speed, and kick up the fan speed and heater both under higher speed and load. Note I mean the motor speed the prototype would be using, and so you want to incorporate the idea of “notches” in your electronics. I’d use BEMF to sense load, and motor rpm.

Greg

p.s. figure out black smoke like a real AlCO and you will have permanent hero status!

Thanks. That’s exactly what I needed to know. So actually modulate the fan speed, as opposed to a binary on/off. I was going to just use the DCC speed packets for control, but BEMF is another option.

Eric,

the real thing only really smokes until the turbo kicks in. but of course no one has come up with the magic black smoke that is repeatable yet.

years ago in model railroader there was a fiction piece on real black smoke for our models.

Al P.

If you don’t have BEMF sensing, you could sense the speed of the motor with a tach, and then use the difference between the commanded speed and the actual speed to figure out load.

Greg

AirWire has BEMF sensing (I assume) for their cruise control, but it is not externally available. Since I plan to use this as a switcher with short trains, most discrepancy between commanded an actual speed will be the result of the momentum setting, which is predictable. So I think I can cobble something realistic together using DCC. I’ll see what I can come up with.