I thought I would see if anyone here has ideas about this... Our club has been talking about replacing our 5 old analog radio throttles, which have been becoming increasingly unreliable. We also have a DCC system (Digitrax Super Chief) and one idea we're toying with is using DCC to control analog throttles. This would reduce problems with radio interference and possibly have other benefits such as being able to eventually do computer control. Before you suggest using large scale decoders directly as analog throttles, let me say that we have already thought of that. And it has some problems, mainly with regard to how decoders handle shorts. I'm looking for other ideas. That is, some way build or convert an analog throttle so it can be controlled either by Loconet or by a decoder, but still handle shorts as analog throttles normally do. (Note, we have a common rail system so short handling really has to be robust.) I really have no idea where to start. I'm sure it's possible to do. Just not sure if it would be worthwhile to try. Have at it...
sounds like a real rube goldberg contraption you are wanting to build .. you will need some sort of computerized voltage regulator ..
I think that this idea is reasonable, assuming that the output is on an isolated track outside of the DCC-controlled zones. What you want to do is use the decoder as a variable voltage source (or a digital-to-analog converter, if you wish), and then use a shunt transistor with current-limiting electronics around it (and a big heat sink) to handle the higher current sent to the track. So basically you're ending up with a digitally-controlled analog throttle. But this would have to be a build from scratch. Jury-rigging or modifying existing throttles would be a bit dicey unless you know exactly how the innards are designed.
I'd think it might be easier to tack a bullet-proof short protection circuit on the backend of a large scale decoder than trying to build/modify a DCC controlled 'analogue' controller. Bear in mind that many of the later 'analogue' DC controllers are similar to the output end of a decoder anyway - eg. PWM is not uncommon. You could go really old-fashioned of course and use a motor-generator. Decoder drives a motor, motor drives a DC generator (another motor basically) which is connected to the track. It's hard to get more isolated than that
If you use a Zimo large scale decoder they have short circuit protection and overheat protection on the motor connection side of the decoder. This may be easier then trying to design something from scratch.
As I interpret what you're saying, that won't be possible at the club. It's a common rail system. Can you explain why this would be a problem, though? Okay, now I'll have to look up 'shunt resistor'. This is a little over my head, but thanks.
So how would we know if something is 'bullet proof?' We'd need to be sure that we're not going to fry decoders. Well, that's a fascinating idea! Is it actually practical? There would be an efficiency drop, of course, so I'm wondering if that would be manageable. Also, would the generator have any strange 'loading characteristics'?
Thanks for the response (even though I said I was looking for other ideas ). We've discussed similar short protection features on the Lenz Silver+, but the problem is that when the decoder shorts, it apparently sets a CV that has to be reprogrammed before it will work again. Given the number of shorts we experience on the mainline, taking the time to reprogram a decoder after a short isn't really acceptable. How does the Zimo handle shorts? And would it stand up to repeated shorts over a long lifetime?
It would need designing and testing. The overload protection isn't hard - it's ensuring other kinds of 'short' that for example connect one side of the decoder to another power source (decoder/power supply) don't cause problems that will stretch the imagination, but that is a problem for any system including 'analogue' controllers. Before power electronics became common it was a very common system for converting power/voltage, especially on DC railways where a transformer is no use. Yes, there will be efficiency loss and it will be quite bad with the small motors involved as they are pretty inefficient anyway. As such to get say 5A output you might need to input 7 or 8A (at 12V) so the 'motor' might need to be heftier than the 'generator'. The only significant 'loading' characteristic would be a bit of lag due to the additional physical inertia of the rotating parts. Still a good idea to put a circuit breaker on the output though. It's easy enough to test if you have a couple of DC motors available (even small ones). Just connect the two shafts together, controller wires to one motor, wires to test track from the other, and try running a loco on the test track. It'll be fun to do even if it goes no further
I admit that this would take a bit of R&D and testing to get it to work properly and reliably. I work in electronics and I know a lot about that aspect of design work - an idea is good on paper but needs debugging and testing. That and this project looks a lot like a test instrument that I assemble and had a large part in designing - a microprocessor-controlled current injector for the electric utilities. Trade the current source for a voltage source, and voila... But it's not a bad idea - someone who likes to tinker and experiment would have a ball with this one.
