Route control options for a yard ladder

[GaryHinshaw]

All,

I have a very simple stub-end staging yard with a compound ladder that I would like to wire for "one button" route selection.  It's a 9-track yard with 8 turnouts, all powered by tortoises.  I have a pair of DS64 quad decoders to control the 8 tortoises, but I would also like to have a redundant panel with a single push button per track for route selection (9 routes in all).  From the cursory research I have done, I can either program routes directly into the DS64's (with some effort, since each unit only supports 8 routes), or use JMRI to program the routes independently.

One specific question I have not found an answer to yet is this: if I go the JMRI route, is it still possible to operate the ladder without the computer plugged in?  In particular, if I want to use the push-button control panel via the DS64 inputs, does that require that I program routes directly into the DS64s, or can they be configures to activate the JMRI routes?  If the latter, this presumably requires that JMRI be running.  I don't mind having the computer online, but I'm a bit leery of having it be a requirement.

Are there any good intermediate-level references on this topic?  I have been looking at this pure-DS64 option:

http://www.digitrax.com/tsd/KB393/ds64-with-yard-ladder-manual-route-buttons/

and this JMRI page:

http://jmri.sourceforge.net/help/en/package/jmri/jmrit/beantable/RouteAddEdit.shtml

Thanks!

[jagged ben]

I could be wrong, but I don't think you can get the DS64 to send that kind of information to JMRI and receive a response back.  With the hardware you have described, I believe your most straightforward choice is to program the routes into the DS64 and then activate those routes either with the DS64's local inputs or remotely through Loconet and DCC (JMRI or a throttle).  This has the advantage of not requiring the computer to be running to throw the turnouts.

Disclaimer:  I have knowledge of setting up yard ladder routes with JMRI and other decoders, but not the DS64.

A less straightforward option, requiring more hardware, would be to do the programming in JMRI and set up the pushbuttons to be inputs to JMRI.  I can't see why you'd want to do it this way though.  And yes, the computer would have to be running.

Finally, FWIW...
All the routing on our club layout is done with hardware, namely diode matrices.   While requiring some materials acquisition, I'm not sure it takes much more time to make a diode matrix than to program a DS64.  And it has the advantage that you don't have to redo the work if the decoder ever craps out, or you want to rearrange your decoders, or whatever.  Once you build the diode matrix and connect it to the turnouts you can run it with whatever you like and never have to program routes again.

[GaryHinshaw]

Thanks jb, that's useful feedback.  I thought about a diode matrix, but I'd like to be able to operate the ladder remotely (throttle and/or JMRI).  Does your club have matrices set up for remote operation?

I continue to read up on turnouts, sensors and routes in JMRI and I'm going to approach this in baby steps to make sure I understand how all the components interact.  I'll probably be back with more questions soon, once they are better formulated.  Further suggestions welcome anytime.

[jagged ben]

Thanks jb, that's useful feedback.  I thought about a diode matrix, but I'd like to be able to operate the ladder remotely (throttle and/or JMRI).  Does your club have matrices set up for remote operation?

Gary, I never answered this question of yours...

Yes, you can do remote operation with a diode matrix;just feed the diode matrix with the decoders.   We have one yard ladder set up this way and will probably have others in the future.  We also have some simple diode controls for mainline turnouts that operate off of AC power in parallel with the Lenz LS150.

The advantages of a diode matrix are...
1) It usually reduces the number of drive outputs required, whether the output is via button or decoder output.  For example your staging yard has 8 turnouts and nine tracks.  To drive 8 turnouts just with decoders and route programming would require 16 decoder wire outputs (thrown and closed for each turnout)   But with a diode matrix you only need the number of tracks, which is 9, plus a return.

2) Eliminates decoder route programming: just connect the decoder outputs in the order you want, write them down for throttle control and/or program them into JMRI.  Replace a decoder quickly if it craps out, just install and program the address.

[peteski]

If I understand correctly, the diode matrix method will only supply power to the machines while being thrown?  Are these stall-motor machines reliable enough to retain their thrown position without power applied to them?

The other thing I'm not sure about is how will the matrix reverse polarity to throw certain machines in the opposite direction as-needed. Usually diode-matrices are used with twin-coil machines (no polarity reversal needed).  Seems to me that for stall-motor machines some additional logic might be needed (other than diodes). But maybe I'm just not seeing the whole picture.

EDIT: I just re-read the last post and I see that the machines are powered using a positive/negative supply and a common return. That makes more sense but I still can't quite visualize the diode matrix. I guess I would have to try to draw this out on a piece of paper. 

[jagged ben]

If I understand correctly, the diode matrix method will only supply power to the machines while being thrown? 

The way we do it at the club, yes.   But that's up to the power source.  Using a diode matrix doesn't determine that.  If you are using decoders then it's typically a decoder setting.

Are these stall-motor machines reliable enough to retain their thrown position without power applied to them?

Yes. 

The other thing I'm not sure about is how will the matrix reverse polarity to throw certain machines in the opposite direction as-needed.

With diodes.  And resistors.

Usually diode-matrices are used with twin-coil machines (no polarity reversal needed).  Seems to me that for stall-motor machines some additional logic might be needed (other than diodes). But maybe I'm just not seeing the whole picture. 

This diode matrix is designed specifically for stall motors. [EDIT: Incorrect links removed.  See below.]

EDIT: I just re-read the last post and I see that the machines are powered using a positive/negative supply and a common return.

Not sure where you are getting that from. 

[jagged ben]

PS:  After perusing the instructions for the DS64, I'm not sure that it can be set up for use with the diode matrix I linked to.  We power our matrices with the CVP AD4, which has the advantage of being configured either as 4 paired outputs or 8 independent outputs.  It's also considerably cheaper.

[peteski]

Ben, your statement "But with a diode matrix you only need the number of tracks, which is 9, plus a return." (specifically the "return") was what made me think that all the motors have one side hooked up to a common return and the other side is supplied positive/negative voltage (in reference to the return) to drive them in each direction.

The circuits you showed are much more than a simple diode matrix.  No wonder I couldn't wrap my head around using diode matrix for controlling stall-motor machines.  In my mind I was envisioning a simple diode-only matrix (like the twin-coil machine matrices of the old days).  Those consisted only of a power supply, diodes and pushbuttons.

You examples have lots of ancillary electronics. The diodes are simply acting an logical AND gates which feed the electronic flip-flops that power the machines.  Now I get it.   

[jagged ben]

Sorry Peteski, I should have looked more carefully at those links.  They are not the ones we are using, so I removed them.

The diode matrices we use require only diodes and resistors, nothing fancier.  The return is simply the common (ground) side of the circuit, all outputs are the same polarity. 

If I can't find a corresponding diagram online then feel free to PM me with your e-mail address and I'll send you one.  Same goes for you Gary.  (Sorry, not going to set up pic hosting for myself yet just for this.)   Everything Google turns up is needlessly complicated.

[peteski]

Sorry Peteski, I should have looked more carefully at those links.  They are not the ones we are using, so I removed them.

The diode matrices we use require only diodes and resistors, nothing fancier.  The return is simply the common (ground) side of the circuit, all outputs are the same polarity. 

If I can't find a corresponding diagram online then feel free to PM me with your e-mail address and I'll send you one.  Same goes for you Gary.  (Sorry, not going to set up pic hosting for myself yet just for this.)   Everything Google turns up is needlessly complicated.

No Problem Ben.  But if you use a common (ground) I still don't picture how you reverse the polarity of the switch motors.  I don't need a diagram of this circuit right now - I was simply curious.  If you ever come up with a diagram I'll be interested in seeing it (just to satisfy my curiosity).

[jagged ben]

Here's a description:

Each pushbutton (or rotary switch position) connects 12V + through diode(s) to the side of the tortoise(s) you want to drive for that route.

Both sides of each tortoise are also connected to Gnd through a 1K resistor.

That's all there is to it.   

[peteski]

Here's a description:

Each pushbutton (or rotary switch position) connects 12V + through diode(s) to the side of the tortoise(s) you want to drive for that route.

Both sides of each tortoise are also connected to Gnd through a 1K resistor.

That's all there is to it.   

Now I understand - thanks Ben!  So only about 12mA is "wasted" by each 1K resistor which ends up with 12V directly applied to that side of the circuit.  That make sense.

[GaryHinshaw]

Thanks for the responses.  I'm not seeing the setup immediately from your description, but if you have a diagram, please feel free to email it to me at my gmail address (I think you already have that).  No rush BTW - this project is on the back burner while I work on the building up the "Vortex", which is needed to connect my yard to the rest of the layout.   I'll have some progress to report on that front in my TBC thread soon.