Model Train Forum banner

1 - 13 of 13 Posts

·
Registered
Joined
·
56 Posts
Discussion Starter #1
If an analog (non-DCC) locomotive is sitting in a DC track block, does the DC voltage or DC current change vs an empty block?
Would throttle state affect that?
If so, could this be used by a hypothetical smart controller for train detection and control?

I am thinking of the feasibility of a possible smart DC controller, that can say OK, I have trains (locomotives) in blocks 1, 3, and 5. A UI could allow individual control of these trains by throttling each block. It could know the block layout, and keep a train moving from block to block, at least until it comes to an occupied block.
 

·
Registered
Joined
·
890 Posts
If an analog (non-DCC) locomotive is sitting in a DC track block, does the DC voltage or DC current change vs an empty block?
Sure. The voltage won't change much/any unless the loco is pulling more current than the PS can supply, but the current will increase with the voltage to match the power draw of the motor.

Would throttle state affect that?
If so, could this be used by a hypothetical smart controller for train detection and control?
Not sure what you mean by throttle state, but the current draw could be used as train detection.
 

·
Registered
Joined
·
56 Posts
Discussion Starter #3
So consider a simple track loop with two blocks. Controller is configured to know there are two and that they are connected to each other (as opposed to two disjoint loops). Controller at startup increases the throttle voltage on each block and notes whether or not the current changes, and immediately reverts throttle voltage to zero - this would be milli or microseconds worth of throttle. The changing current is logically recorded as train present or no train.

Controller (a raspberry pi say) presents a UI with the two blocks and a symbolic train in block 1 say. User clicks on train to select it and a throttle dial pops up. Mouse scroll wheel operates throttle for that train, which moves forward or backward as appropriate. Controller knows train is going to move into block 2 which is unoccupied (no train) initially, so controller allows train to proceed and connects the throttle popup to block 2 also. When it enters block 2 the changing current is detected and the throttle pop up is disconnected from block 1 which is now empty and the train keeps moving around the loop since the next block is empty. If there were two trains one in each block, the computer would not allow either one to move since the next block is occupied.

This could handle layouts of arbitrary complexity based on only enabling train movement when the next block is empty. If there is a turnout at the end of the block that connects say block 1 to block 2 (the loop) and block 3 (a siding), then the throttle popup would have a turnout indicator that could be switched by clicking on it to toggle the direction.

Each siding (dead end) would need to be two blocks and the final end block would stop the train since there is nowhere for the train to go. It would then allow a reverse throttle for the train to back out.

Each throttle popup would represent a train, and might have a series of turnout selectors (or logical layout diagram) that show the trains path through the layout. This would still require each block to be wired separately, but if such a controller were developed it could be used with existing DC layouts.
 

·
Registered
Joined
·
890 Posts
Seems feasible. I would *guess* that you could do something similar using DCC with JMRI or TCG, perhaps doing some of your customized block detection... I've never used either of those apps so don't know, but if you're going to add in the automation you might want to look into going DCC.
 

·
Registered
Joined
·
56 Posts
Discussion Starter #5
Yes DCC obviously is the (mostly) standard way of doing this but it's expensive and judging by some of the stories in this forum, error prone. And it doesn't support DC locomotives unless you can do the decoder brain surgery on them.
 

·
Registered
Joined
·
2,951 Posts
If an analog (non-DCC) locomotive is sitting in a DC track block, does the DC voltage or DC current change vs an empty block?
Would throttle state affect that?
If so, could this be used by a hypothetical smart controller for train detection and control?
analog locomotive is not moving means no voltage to the tack. empty or occupied. throttle state affects this directly - the more you turn the dial, the more voltage going to be on the track.

I am thinking of the feasibility of a possible smart DC controller, that can say OK, I have trains (locomotives) in blocks 1, 3, and 5. A UI could allow individual control of these trains by throttling each block. It could know the block layout, and keep a train moving from block to block, at least until it comes to an occupied block.
this will take multiple externally controlled variable supplies to control each section separately.

theoretically you could supply small voltages (0.5V) to the track that will not break the loco stiction and detect those currents, but why bother?
if you insist on analog control you could do optical detection.


bottom line feasible? yes , somewhat feasible. but not practical at all. i'd stick to digital

good luck
 

·
Registered
Joined
·
3,444 Posts
And it doesn't support DC locomotives unless you can do the decoder brain surgery on them.



some DCC systems do support DC loco, by using zero stretching ... without adding a decoder ..

but in response to your main question, while a DC loco is sitting there is no difference to an empty block, once it starts moving the current is different ..
that's why all the different types of 'detectors' that can be mounted underneath where they can't be seen ..
 

·
Registered
Joined
·
56 Posts
Discussion Starter #8
analog locomotive is not moving means no voltage to the tack. empty or occupied. throttle state affects this directly - the more you turn the dial, the more voltage going to be on the track.


this will take multiple externally controlled variable supplies to control each section separately.
I think there's a $5 chip that does the switching part - at least 8 throttle channels switched to 16 output blocks, and it's cascadable if more are needed: https://www.futurlec.com/Zarlink/MT8816AEpr.shtml

theoretically you could supply small voltages (0.5V) to the track that will not break the loco stiction and detect those currents, but why bother?
if you insist on analog control you could do optical detection.




bottom line feasible? yes , somewhat feasible. but not practical at all. i'd stick to digital

good luck
I think optical detection would require effectively duplicate wiring and extra logic - there are expensive components available to do that.

I guess I am looking for a cheap effective way to control a DC layout that leverages cheap electronics and computers and open source software.
 

·
Registered
Joined
·
1,030 Posts
i think what you asking is how to detect that a locomotive has moved from one block to another. At least with DCC, occupancy detector indicate that a block is occupied by a locomotive or cars.

linn westcott's twin t circuit can detect a block drawing current. can't find a schematic on the web but it may be as simple as putting the base-emitter path of a transistor in series with the throttle and track. the transistor conducts when there is current and there needs to be a 2nd circuit for the opposite polarity

before releasing a block that a locomotive has exited from, you may also want to consider when the entire train has exited the block before allowing another train to enter it
 

·
Registered
Joined
·
890 Posts
Cheap is relative. But, if you have a lot of spare time and are using it as a learning experience, then OK.
 

·
Registered
Joined
·
2,951 Posts
I guess I am looking for a cheap effective way to control a DC layout that leverages cheap electronics and computers and open source software.
to me it looks you headed in the wrong direction then.

equipping a locomotive with ~20$ worth of electronics will go a very long way.

good luck
 

·
Registered
Joined
·
56 Posts
Discussion Starter #12
i think what you asking is how to detect that a locomotive has moved from one block to another. At least with DCC, occupancy detector indicate that a block is occupied by a locomotive or cars.

linn westcott's twin t circuit can detect a block drawing current. can't find a schematic on the web but it may be as simple as putting the base-emitter path of a transistor in series with the throttle and track. the transistor conducts when there is current and there needs to be a 2nd circuit for the opposite polarity

before releasing a block that a locomotive has exited from, you may also want to consider when the entire train has exited the block before allowing another train to enter it
Thanks for the pointer to that circuit - seems there is a long history of DC control tech I know nothing about! And good point about needing the entire train to leave the block - may have to have some kind of current draw in the caboose such as a 10k resistor across the last pair of wheels so the controller can detect that tiny change in current when the caboose leaves the block.

Cheap is relative. But, if you have a lot of spare time and are using it as a learning experience, then OK.
This would not be cheap for the initial development obviously, but once it exists the cost for adding it to a DC layout would be minimal. Cost of raspberry pi plus maybe a circuit board or bread board and a couple of $5 chips. Commercials would not develop it since there would be no profit (this is the main problem with pure capitalism - good cheap solutions don't get implemented because there's no profit). There may also be patent issues, although I expect most of those relevant expired long ago.
 

·
Registered
Joined
·
890 Posts
If you're going to add a resistor to your caboose, just light it. 2 birds, one stone, and the current will likely be higher than a 10k resistor.
 
1 - 13 of 13 Posts
Top