What is in an auto-reverser?
- Thread starter TomR
- Start date
ModelRailroadForums.com is a free Model Railroad Discussion Forum and photo gallery. We cover all scales and sizes of model railroads. Online since 2002, it's one of the oldest and largest model railroad forums on the web. Whether you're a master model railroader or just getting started, you'll find something of interest here.
When I read that it sounds like they use current limiting to detect a fault. The instant the rails get a short, they either reverse things or disconnect a section to clear the fault.
I'm just interested in how they do it.
I can think of a way to do it with regular DC control, but I don't have a loop-back that crosses the polarity so no need to worry for me.
I'll eventually do that, but I'm using up my old stuff I already bought while I learn. I like to build things, and DCC would turn me into a consumer.
later,
Tom
I'm a little surprised you haven't put one in just to mess with one!
Not necessarily - I'd say there's even more opportunity for DIY circuits in DCC land - Block detection and signaling is a whole other can of (logic!) worms for example - Sure, you could do the same with DC, but running trains can become so much more "involved" with DCC.
As always, just my 02c,
Cheers,
Ian
Great! Where is the guide at? On line?? I'd like to learn what the standards for control signals are, and how it works.
I can do surface mount, I have the proper hot air soldering equipment in my shop. I just have to work under a magnifier at my age.
Thanks, that does it!!!
Youngster.
It would be interesting to see more. Stuff like that would be useful on a web site.
That's pretty straight forward stuff, even without component values. The hard part for me would be the code for the decoder. I don't write that stuff. I just do the hardware.
I'm a little suprised they don't have a storage capacitor across the output of the bridge rectifier.
Don't systems have energy storage, so momentary loss of power doesn't glitch things up and also make the motor drop for a second?
I'm thinking a reverser would have a current limiter to prevent the momentary short when the locomotive ties across the insulated joint from faulting the supply (booster). Nothing much else needs to be done, unless I'm missing something.
I can see how it could work pretty smooth since the control system in the locomotive knows the direction rather than sensing track polarity.
I can think of how to do it automatically with a DC system too, but the hardware would be a great deal more complex. No wonder everyone went to DCC systems.
Here's the start of my first antiquated technology system.
http://www.w8ji.com/train_real_and_models.htm
I did a quick Google search and didn't find any products or conversation for auto DC reverser. Do you know of any manufactured reversers for DC systems?
It would have to be some sort of current limiter, even a big current foldback limiter. Anything that didn't overload the source while a short was there (even for an instant) would do the trick, because the power supply would think it was only maximum load and not a short.
It would have to be fast, but that's no problem at all.
I wouldn't think terribly large physically, but it would need a good bit of capacitance if it were going to continue to run the motor. Of course the motor is a generator while spinning down, but I'm not sure if spinning it down slow would hold the PLC power up long enough to get past any fault. The motors are only what, half amp or so at 10 or 15 volts??
Without keeping the motor or lights running, the PLC could be held up with a fairly small cap. So if there is something there to hold the lights up, the PLC's added load is meaningless.
A good algorthm would not do anything with bad data or dropped data lines except wait for the next good stream. More of a problem isn't the data interruption, but the fact if the PLC doesn't crash it can keep all the ouputs the same.
I don't do the code but I do the hardware. In the process I outline basic algoryths for someone who does do the code. The stuff I commonly do requires either writing status to memory if the power is going to drop, and that means keeping the PLC running long enough to write a bunch of information in. This should be such a short power interruption the cap could be reasonably small.
What chip is that? Do they have a parts list?
Thanks,
Tom
If you really are interested in building your own components, you need to take a look at this site: MERG(Model Electronic Railway Group). They have schematics, circuit board layouts, and for the components that need it, even the code, for many DCC components.
I had planned on developing and building my own DCC system, and even had a working basic command station, but there is enough affordable commercial stuff out there now that to me it is no longer worth the trouble to build my own.
They do have a capacitor on the output of the voltage regulator to keep the processor up during momentary interruptions of power, but I think for a cap to do any good for the motor it would just have to be too big.
The processor in the decoder would need to be programed, but it is possible that someone offered preprogrammed processors for those circuits.
No need for a current limiter, the reverser simply switches faster than the booster over-current protection(usually*). The booster generally does not shutdown immediately for a short, but has a built in delay. In fact some of them let you adust the delay - set it slower if you have additional circuit breakers or autoreversers, quicker for better protectioin if you don't.
*There are some boosters that will shutdown too quick and have trouble working with some autoreversers.
Great! Thanks. I did some searches but the good stuff must come up buried in the pile of search results, or I used the wrong keywords.
That's probably the case. I'll probably buy the DCC and customize some interfaces that apply to my layout. I've got a couple good IR sensors going, including a receiver that decodes multiple IR sources with just one sensor. I thought about multiple sources spaced along the track so the sensor knows the position and direction.
Another problem with the cap would be charging current. It would increase wheel or contact sparking a great deal unless the source was current limited. It sounds like the power source (booster) has the current limiting, maybe not the reverser. Somewhere current limiting is hidden.
Tom
I don't, but I was thinking of ways to manage it. I'm not sure how practical a few of the component values would be or if the cost would be worth it.
Right now I'm buried in paying work and starting the landscaping and wiring my layout, so I won't experiment with anything else for a while.
Tom
Affiliate Disclosure: We may receive a commision from some of the links and ads shown on this website (Learn More Here)
I don't think it's a current limiter per-se - They detect a short circuit and switch the phase of the DCC signal faster than the booster can shut down.
Pretty good reading here: http://www.tonystrains.com/technews/powershield_icb.htm
Note the adjustments you can make - Some of 'em got *horrible* reviews as they'd shut down and/or flip with "no warning" - Drove people nuts!
Cheers,
Ian
PS - Tom, given your expressed love of things electronic, you gotta go DCC IMHO
Pretty good reading here: http://www.tonystrains.com/technews/powershield_icb.htm
Note the adjustments you can make - Some of 'em got *horrible* reviews as they'd shut down and/or flip with "no warning" - Drove people nuts!
Cheers,
Ian
PS - Tom, given your expressed love of things electronic, you gotta go DCC IMHO
When I read that it sounds like they use current limiting to detect a fault. The instant the rails get a short, they either reverse things or disconnect a section to clear the fault.
I'm just interested in how they do it.
I can think of a way to do it with regular DC control, but I don't have a loop-back that crosses the polarity so no need to worry for me.
Cheers,
Ian
PS - Tom, given your expressed love of things electronic, you gotta go DCC IMHO
I'll eventually do that, but I'm using up my old stuff I already bought while I learn. I like to build things, and DCC would turn me into a consumer.
later,
Tom
I can think of a way to do it with regular DC control, but I don't have a loop-back that crosses the polarity so no need to worry for me.
I'm a little surprised you haven't put one in just to mess with one!
...and DCC would turn me into a consumer.
Not necessarily - I'd say there's even more opportunity for DIY circuits in DCC land - Block detection and signaling is a whole other can of (logic!) worms for example - Sure, you could do the same with DC, but running trains can become so much more "involved" with DCC.
As always, just my 02c,
Cheers,
Ian
RW&C
N Scale with Stone Tools
I've found a DCC guide from 1994, and it's got schematics of all kinds of DCC stuff in it. Most of it's way over my head, but for someone with the electronics knowledge, it couldn't be too hard to breadboard a few parts. (Might want to just buy the decoders, though, unless you're good with soldering down surface-mount ICs.)
(Might want to just buy the decoders, though, unless you're good with soldering down surface-mount ICs.)I've found a DCC guide from 1994, and it's got schematics of all kinds of DCC stuff in it. Most of it's way over my head, but for someone with the electronics knowledge, it couldn't be too hard to breadboard a few parts.
Great! Where is the guide at? On line?? I'd like to learn what the standards for control signals are, and how it works.
I can do surface mount, I have the proper hot air soldering equipment in my shop. I just have to work under a magnifier at my age.
CSX Robert
Member
If you want to learn the what the standards are, start here: DCC Standards & Recommended Practices Index.
Thanks, that does it!!!
RW&C
N Scale with Stone Tools
What I have is a crudely photocopied guide from the 94 NMRA convention in Oregon. (I wasn't there, I was 3 at the time.) It's a bit out of date, of course, but most of the info should still be good.
Some of it's illegible due to poor photocopying, and the book has some water damage, but there's a hell of a lot of info in here. Schematics, component lists, explanations, how to interface it to a computer, written out programs for doing so. It's a good half inch thick. It's even got some primitive sound decoders in it... looking for a reverser now. (Edit: there isn't one, all the diagrams show it as part of a special booster hooked to the reverse loop.)
If you guys are interested, I'll look through it and scan out some more of this stuff.
Some of it's illegible due to poor photocopying, and the book has some water damage, but there's a hell of a lot of info in here. Schematics, component lists, explanations, how to interface it to a computer, written out programs for doing so. It's a good half inch thick. It's even got some primitive sound decoders in it... looking for a reverser now. (Edit: there isn't one, all the diagrams show it as part of a special booster hooked to the reverse loop.)
If you guys are interested, I'll look through it and scan out some more of this stuff.
Last edited by a moderator:
Youngster.
It would be interesting to see more. Stuff like that would be useful on a web site.
That's pretty straight forward stuff, even without component values. The hard part for me would be the code for the decoder. I don't write that stuff. I just do the hardware.
I'm a little suprised they don't have a storage capacitor across the output of the bridge rectifier.
Don't systems have energy storage, so momentary loss of power doesn't glitch things up and also make the motor drop for a second?
I'm thinking a reverser would have a current limiter to prevent the momentary short when the locomotive ties across the insulated joint from faulting the supply (booster). Nothing much else needs to be done, unless I'm missing something.
I can see how it could work pretty smooth since the control system in the locomotive knows the direction rather than sensing track polarity.
I can think of how to do it automatically with a DC system too, but the hardware would be a great deal more complex. No wonder everyone went to DCC systems.
Here's the start of my first antiquated technology system.
http://www.w8ji.com/train_real_and_models.htm
Last edited by a moderator:
...
I can think of how to do it automatically with a DC system too, but the hardware would be a great deal more complex. No wonder everyone went to DCC systems.
...[/URL]
I did a quick Google search and didn't find any products or conversation for auto DC reverser. Do you know of any manufactured reversers for DC systems?
RW&C
N Scale with Stone Tools
What can I say, I started young. 
As far as I know, a DCC reverser just senses the short, then switches before the short trips the breaker in the booster. A limiter does sound logical to minimize the short.
I've never seen a commercial decoder with a capacitor there. I suspect that a cap of sufficient size would be pretty bulky. Also, if the board remains powered, but only gets half a command, what would it do? How would it respond to low current (as the capacitor charges and drains)? I guess it can't hurt to try, but I suspect you'd be better off putting a good flywheel on the motor, and smaller capacitors on the headlights to kill the flicker. (That'd make the on/off more dramatic, too.)
The book doesn't mention programming the decoders. The only reference in the schematic section is
There's a lot of stuff in here. Stationary decoders, animation, coach lighting, computer interfaces, and that's just in this one section. I'll scan some of the better stuff and upload it to another thread when I get a good chunk of time to burn.
As far as I know, a DCC reverser just senses the short, then switches before the short trips the breaker in the booster. A limiter does sound logical to minimize the short.
I've never seen a commercial decoder with a capacitor there. I suspect that a cap of sufficient size would be pretty bulky. Also, if the board remains powered, but only gets half a command, what would it do? How would it respond to low current (as the capacitor charges and drains)? I guess it can't hurt to try, but I suspect you'd be better off putting a good flywheel on the motor, and smaller capacitors on the headlights to kill the flicker. (That'd make the on/off more dramatic, too.)
The book doesn't mention programming the decoders. The only reference in the schematic section is
From that, I assume that the circuits detailed in the book will work without any special programming. (The spec sheet for the one I posted says that it'll take an address from 0-32, so clearly it's not terribly sophisticated.)
There's a lot of stuff in here. Stationary decoders, animation, coach lighting, computer interfaces, and that's just in this one section. I'll scan some of the better stuff and upload it to another thread when I get a good chunk of time to burn.
It would have to be some sort of current limiter, even a big current foldback limiter. Anything that didn't overload the source while a short was there (even for an instant) would do the trick, because the power supply would think it was only maximum load and not a short.
It would have to be fast, but that's no problem at all.
I wouldn't think terribly large physically, but it would need a good bit of capacitance if it were going to continue to run the motor. Of course the motor is a generator while spinning down, but I'm not sure if spinning it down slow would hold the PLC power up long enough to get past any fault. The motors are only what, half amp or so at 10 or 15 volts??
Without keeping the motor or lights running, the PLC could be held up with a fairly small cap. So if there is something there to hold the lights up, the PLC's added load is meaningless.
A good algorthm would not do anything with bad data or dropped data lines except wait for the next good stream. More of a problem isn't the data interruption, but the fact if the PLC doesn't crash it can keep all the ouputs the same.
I don't do the code but I do the hardware. In the process I outline basic algoryths for someone who does do the code. The stuff I commonly do requires either writing status to memory if the power is going to drop, and that means keeping the PLC running long enough to write a bunch of information in. This should be such a short power interruption the cap could be reasonably small.
What chip is that? Do they have a parts list?
Thanks,
Tom
CSX Robert
Member
If you really are interested in building your own components, you need to take a look at this site: MERG(Model Electronic Railway Group). They have schematics, circuit board layouts, and for the components that need it, even the code, for many DCC components.
I had planned on developing and building my own DCC system, and even had a working basic command station, but there is enough affordable commercial stuff out there now that to me it is no longer worth the trouble to build my own.
They do have a capacitor on the output of the voltage regulator to keep the processor up during momentary interruptions of power, but I think for a cap to do any good for the motor it would just have to be too big.
Last edited by a moderator:
RW&C
N Scale with Stone Tools
Clearly you know more about this stuff than I do.
The parts list for that circuit (hand written, no less!):
IC1=LD10 (Digital Plus)
IC2=78L05 (TO-92 Case)
T1-T2=BD680
T3-T4=BD679
T5-T8=BC547
X=4.19 MHz Resonator
D1-D4=PH77D (or similar 1A)
D5=24v Zener diode
C1=22 microfarad (or two sixteen microfarad) 16v ELcap
R1=47k ohms 1/8 watt
R2-R6=3.3K ohms
D6=IN4148 Diode
LA1-2=14-16v/55-50 mA bulb
Motor HO=M
P1-P2=track connections
A0-A4=Address code 0-32
STV1-2 = 4 steps start voltage
(Is and 1s, 0s and Os may be switched around.)
I have no idea how much of this helps or is useful, but this is all that was there. I imagine finding the ICs will be the hard part.
Edit: CSX, thanks for the link! Looks like a much better source.
The parts list for that circuit (hand written, no less!):
IC1=LD10 (Digital Plus)
IC2=78L05 (TO-92 Case)
T1-T2=BD680
T3-T4=BD679
T5-T8=BC547
X=4.19 MHz Resonator
D1-D4=PH77D (or similar 1A)
D5=24v Zener diode
C1=22 microfarad (or two sixteen microfarad) 16v ELcap
R1=47k ohms 1/8 watt
R2-R6=3.3K ohms
D6=IN4148 Diode
LA1-2=14-16v/55-50 mA bulb
Motor HO=M
P1-P2=track connections
A0-A4=Address code 0-32
STV1-2 = 4 steps start voltage
(Is and 1s, 0s and Os may be switched around.)
I have no idea how much of this helps or is useful, but this is all that was there.
I imagine finding the ICs will be the hard part.Edit: CSX, thanks for the link! Looks like a much better source.
Last edited by a moderator:
CSX Robert
Member
The processor in the decoder would need to be programed, but it is possible that someone offered preprogrammed processors for those circuits.
CSX Robert
Member
No need for a current limiter, the reverser simply switches faster than the booster over-current protection(usually*). The booster generally does not shutdown immediately for a short, but has a built in delay. In fact some of them let you adust the delay - set it slower if you have additional circuit breakers or autoreversers, quicker for better protectioin if you don't.
*There are some boosters that will shutdown too quick and have trouble working with some autoreversers.
Great! Thanks. I did some searches but the good stuff must come up buried in the pile of search results, or I used the wrong keywords.
That's probably the case. I'll probably buy the DCC and customize some interfaces that apply to my layout. I've got a couple good IR sensors going, including a receiver that decodes multiple IR sources with just one sensor. I thought about multiple sources spaced along the track so the sensor knows the position and direction.
Another problem with the cap would be charging current. It would increase wheel or contact sparking a great deal unless the source was current limited. It sounds like the power source (booster) has the current limiting, maybe not the reverser. Somewhere current limiting is hidden.
Tom
I don't, but I was thinking of ways to manage it. I'm not sure how practical a few of the component values would be or if the cost would be worth it.
Right now I'm buried in paying work and starting the landscaping and wiring my layout, so I won't experiment with anything else for a while.
Tom
Affiliate Disclosure: We may receive a commision from some of the links and ads shown on this website (Learn More Here)