Monday, December 22, 2014

December Update


This is my last post for 2014 I hope you have gotten some enjoyment from my ramblings...

This month I got the news my wife in 2015 has a job back in Adelaide and it is a little under one hour drive from home.  So after her working in the country for over a decade she will be living at home full time!  It will be great to be living together in the same house again.  Not sure how that will impact my hobby activities. tho....

Well despite a trip to Pt Augusta for my wife's work Xmas show and bringing the first carload of her stuff home I did get some modelling time. Firstly Wayne and I decided on a naming convention for the signal system components and I started learning the best way to terminate ribbon cable to 10 pin female headers. I can also report that the second version of the Ambient Light Detector PCB has arrived, more work on the design of the Standard Signaling Module has been achieved leading to s breadboard prototype, a testing module and an order for some prototypes.


Naming Convention - Signal System Components

It had become very evident that to ensure Wayne & I were always on task talking about the same bit of kit we needed to standardise!  So following some rational discussion we settled on the following:

Signal Light Module (SLM)


This is the generic Red-Yellow-Green controller - I am using Rob Paisleys 3 light signal driver and Wayne intends to use the one in the March 2001 Model Railroader Magazine.


Dual Block Direction Sensor (DBDS)


Previously known as the Ambient Light Detector PCB - the new name better depicts what it actually does.


Standard Signal Control Module (SSCM)


Previously known as the Standard Signalling Module

Breadboard Prototype
It is hoped that there will be less confusion in technical discussions from now on.


Terminating Ribbon Cable

Given there are going to be may "interconnections using ribbon cable"  I started practicing how to terminate on female headers.



I learned that the center wires need to be shorter if a neat outcome was to be achieved.


Given the outcome my initial idea of sleving every second wire with heat shrink tubing will probably not be required.

Dual Block Direction Sensor (DBDS) 

Last month I reported that Wayne has tweaked the design and a second pilot board  has been ordered to ensure we have nailed it prior to ordering the production run.  Well they have arrived.


The second Prototype


Building the second prototype

The second Prototype under construction

The second Prototype with components mounted

Both Prototypes built and tested
The first board took around 4 hours to build and the second took around 3 hours.


Standard Signal Control Module

The control logic was revisited and the desired outputs were mapped.  Then superimposed was the outputs controlled by the Signal Light Modules (Green/Red highlight) leaving the Yellow/Blue "R's" that require control through this module.  The Local Signal Heads (Ex(B) & Wx(B)), East Ahead (E+) and West Ahead (W+).  The corresponding inputs  for the Turnout conditions Normal or Reverse (TxN & TxR), Block Occupied Bx, Train Going East (BxE) and Train going West (BxW).  The relationships can be seen in the spreadsheet below.



Last months breadboard concept design was ditched and another version devised following a rethink...

The circuit concept is:



I took the design and laid it out on bread board:


I have written the code for the controller which operated successfully.

We have refined the design some more and Wayne has designed the PCB. We are about to sent it off for production of some prototypes.



I guess when the prototypes arrive and are built we can put the total system together and do final testing before placing the "production" PCB order.


Standard Signal Control Module Testing

There will be a need to make quite a number of these modules so I decided to make up a test cntre which represents all the 14 inputs and and the 10 outputs. .




So here is a video that shows how it works - not perfect but you will get the idea....




Merry Christmas and A Happy New Year to all that watch my hobby exploits.


That's it for this month.


Till the end of January or February 2015 - It depends on how moving my wife back home goes....





-ooOOOoo-

Sunday, November 30, 2014

November Update


This month for the first two weeks I was away from home travelling around 4,000 km. I visited Peter in Traralgon and got to see his Brownsville layout again and we visited Murray Scholz to look at his Bogong & Geehi Railway. There was some further work on the proof of concept Ambient Light Detector PCB's, some more on the design for the Standard Signal Module and Flintston Yard Changes.  To top off the month's activities we visited Wayne's layout.


Brownsville

Due to family issues Peter has not been able to progress very far with the layout but we did get to run a few trains and it was good to see him back in his shed again.





Visit to Bogong & Geehi Railway


Well Murray Scholz issued an invite, on a recent Model Rail Radio Podcast, to anyone passing close to his place to "call in" to see his model railway.  He lives near Culcairn NSW.

See his Blog here.

Watch a teaser video below



Some photos I took can be found here.


Ambient Light Detector PCB

Wayne has tweaked the design and a second pilot board  has been ordered to ensure we have nailed it prior to ordering the production run.

Standard Signaling Module

More thinking on optimising the design for integrating the various module components has led me to decide how to deploy and interconnect them.   Below is the proposed circuit and a breadboard concept layout for a standard signal module

Proposed circuit


Breadboard concept layout

The principles agreed after discussion between Wayne & I are that the Detection & Control modules will be centrally located with wires running to detection devices and signal masts rather than locating the modules locally.

Centralising the modules allows for control wiring between modules to be at 3.3v and direct connection between control inputs.  This means that the number of components can be reduced through removal of optocouplers used to isolate module inputs and outputs for simple logic interconnections. The associated number of module connection points can also be reduced.  Another spin off will be the reduction in power consumption from regulated power supplies to.


Flintston Yard Changes

In the August update I detailed the need to replace a crossover with a double slip. I have commencing wiring the track and control panel.

Control Panel Changes
Holes marked and drilled for two toggles, a push button and four LED's.


Wayne's Layout

On 27 November Chris, Barry, Rossco & I visited Wayne's layout.  The purpose of the visit was to get an appreciation of his layout and proposed operating regime and then have a chat and get to know each other a bit better over a BBQ & a few glasses of red wine. We the day was a great success and we all had fun and stayed way longer than we imagined.

He is some introductory stuff..



I also put some photos into a slideshow.




Thanks Wayne!


That's it for this month.


Thanks again to those that continue to help me achieve my ambitions and hello once again to all my followers!


Till the end of December.





-ooOOOoo-

Tuesday, October 28, 2014

October Update


This month has seem some interruptions to normal proceedings as my wife and I got sick and our second Granddaughter has arrived - welcome Alice!

Apart from that, we have received the proof of concept Ambient Light Detector PCB's, planning has commenced on the Standard Signaling Module, and made some more progress on the Signaling Concept Design, Pierre had another training session and the final Grand Opening Video has been completed.

So once again, absolutely nothing was done in the layout room this month!  I might have to change that next month as my medical plan says I should do more exercising - I think that all the movement working above & below the layout qualifies for that!.


Ambient Light Detector PCB

Last month I informed you that Wayne and I completed the design of the Printed Circuit Board and two proof of concept boards were ordered - They have arrived!

Murphy is alive and well too - I realised that I had made a couple of errors in the circuit diagram which found its way onto the board but I am sure it will be easy to fix (very lucky!!).  The current draw was quite a bit higher than I expected then I discovered that the 3.3V voltage regulator "was a radiator" as it was drawing 0.5A - we put it the wrong way round!  The high current circuit check also revealed that I left out 10 current limiting resistors on signal lines (10 x 20mA = 0.2A - Ah that will make a big difference).  The trimpot holes were one size too small too.

I was really impressed with the finish of the boards, which included solder masking, see below.

Board as received



Board with components mounted
Now for the modifications to address my design issues and then testing began.....


Adding current limiting resistors


10 x 330 Ohm between MSP Pins & Optocouplers

Traces broken with Dremel & Stone bit

3.3V regulator modifications


Due to some irrational MSP behaviour and after research I found a recommended practice on the TI MSP forum site in reltion to using a Voltage Regulator that had not been implemented.  We did not have the Capacitors and we used a 10k resistor not 47k.


Previously I had always used Dave Loman's standalone implementation which uses a 10k Ohm resistor between Vcc & Pin 16 when battery powered.


Regulator installed the correct way round, 47k Ohm resistor and 2 x 10uF 50V Caps.

Programing tweak


Whilst reviewing the TI MSP forum I also found some other expert advice (Posted by  on Nov 21 2013 10:04 AM)  in this link ....

// If you're not using the 32kHz crystal then you need to clear the oscillator fault flag and take it out of crystal mode like so.

IFG1 &= ~OFIFG;
  while (BCSCTL3 & LFXT1OF)
      BCSCTL3 = LFXT1S_2;
I do it right after
WDTCTL = WDTPW + WDTHOLD;                 // Stop WDT


We got 3 outa 4 right - duh....


The design for ambient detectors E1, E2, E3, E4, W1 & W2 was perfect but we stuffed up W3 & W4.

So I id further mods to prove the rest works....



Underside trace cuts and jumpers to what it should be...

Final Test


Well after all the above it now works as it should and the board consumes 110mA when operating.



Where to from here


Redesign the W3/W4 section of the board to fix the design error and given all the changes it will be will be sensible to do a second pilot board  to ensure we have nailed it prior to ordering the production run.  I think extra $50 and the wait will be worth it.



Standard Signaling Module

Planning has commenced for the "backbone" module see below.










The next stage will be to build a breadboard prototype.

Signalling Concept Design

Last month I mounted the development board on a frame and had to revamp the "Junction Logic" board to isolate the 3.3V microprocessor circuit from the 12 V signal operation circuit.

I have refined the microprocessor programming to include "error checking" and undertaken extensive testing.  I can report I think I have it working fairly reliably now.

Signal Trial


Below are links to 3 videos, 2 that show the normal operation of the 4 blocks in each of the inner and outer loops of the pilot and the third shows the "Junction Routes".


Outer Loop (Top - East & West signals)


Inner Loop (Bottom - East & West signals)


Junction Routes (Top & Bottom, East & West signals)

Pierre in training

Pierre is gaining more confidence and I think he is enjoying the experience. He has completed a pick up run and a separate set out run.  Now for the combination!!  I have put a small presentation together for him to learn the different types of rolling stock.

Well I can report that he decided that it was harder than expected and he needed some guidance!  A "try again" has been requested however it has not happened as my wife & I have had the lurgy and been put on antibiotics and grounded by the Doc.

Grand Opening Video

Part 3 is now available.



BIG thank you must again go to Don for all the work required to put the 3 videos together for me.


That's it for this month.


Thanks again to those that continue to help me achieve my ambitions and hello once again to all my followers!


Till the end of November.





-ooOOOoo-

Wednesday, October 1, 2014

September Update


This month I have been away in the country to be with my wife for a week and managed to get the 3 light signal driver PCB's built, helped Ros dismantle Pam's Benmoor & Waitaki River Railroad layout, ordered the proof of concept Ambient Light Detector PCB's, made some progress on the Signaling Concept Design, Pierre had another training session and the final Grand Opening Video will be available next month.

So absolutely nothing was done in the layout room this month!


3 Light Signal Driver Circuits

I have finished building the 120 or so PCB's!

Parts box at start

Production Line

Soldering seemed to never end.... 

Completed board

A pile during construction!

Depleted parts box...


I got the boards from Rob Paisley - link

Ambient Light Detector PCB

Wayne and I have completed the design of the Printed Circuit Board and two proof of concept boards have been ordered. We decided to get it professionally made and if successful we will order more!

Here is a sneak peek of the layout.

Signalling Concept Design

This is where most of my time has gone this month.  I mounted the development board on a frame to make the task easier.




The "Junction Logic" board had to be amended following the decision to isolate the 3.3V microprocessor circuit from the 12 V signal operation circuits using optocouplers.


Version 1

Version 2



Adding the "turnout State" wiring (left) & Control RED outputs (right).

In the July I indicated that I wanted the Full use of the TI LaunchPad MSP430G2553 which means re-configuring the PIN use from "Communications" to "I/O".  For those that may be interested following research I can report that you need to configure pins P2.6 & P2.7 using the following:

   P2SEL &= ~BIT6; // Port P2.6 as I/O 1 of 2 steps
   P2SEL2 &= ~BIT6; // Port P2.6 as I/O 2 of 2 steps

   P2SEL &= ~BIT7; // Port P2.7 as I/O 1 of 2 steps
   P2SEL2 &= ~BIT7; // Port P2.7 as I/O 2 of 2 steps

before and in addition to the usual configuration definitions.

I am now refining the microprocessor programming to include "error checking" whilst undertaking extensive testing.

I also got some 10, 20 & 40 way ribbon cables for interconnection of all the baords.

10 way

Pierre in training

Pierre is gaining more confidence and I think he is enjoying the experience. He has completed a pick up run and a separate set out run.  Now for the combination!!  I have put a small presentation together for him to learn the different types of rolling stock.

Grand Opening Video

Part 3 should be available next month.

BIG thank you must go to Don for all the work required to put these together for me.



That's it for this month.


Thanks again to those that continue to help me achieve my ambitions and hello once again to all my followers!


Till the end of October.





-ooOOOoo-