| Author |
 Topic  |
|
LVRALPH
Fireman
5430 Posts |
 Posted - 06/06/2010 : 07:19:03 AM
|
Well, some good news. First Steam nut came over and drilled the hole into my basement and ran the cable so I can connect the SUSIC CTC node to the SMINI nodes in the basement. It's good to have friends like Bruce, Mark F, and Hueber. A lot of my layout would be a lot worse off without them.
On line I have been offered a lot of help from a gentleman named Conrail Bill (down Wojcik). Him and Don Wood are going to be my Shell answer man so as not to clog the CMRI forum up with a lot of newbie questions. Having these 2 pros by my side is really going to make this project a success much faster than me on my own. Especially with the programming.
So keep with me guys. More updates coming.
|
Edited by - LVRALPH on 06/06/2010 5:57:31 PM |
 |
|
|
MarkF
Engineer
USA
9276 Posts |
Posted - 06/06/2010 : 10:26:35 AM
|
| I'm anxious to hear what the problem is you were experiencing. I can wait to see this thing light up the railroad! |
Mark
See my homepage at http://home.comcast.net/~prrndiv/
|
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/07/2010 : 07:09:25 AM
|
I decided to do some system wiring downstairs while I wait for my boards to come back. I decided to relocate my DCCOD's that I distributed around the layout to the Crestwood node. I installed the SMINI and the DCCOD motherboard. I mounted them with rubber grommets and # 4 screws that I had bought by the bagfull.
The output of the detectors gets fed to an input pin on the SMINI or SUSIC if your using it. So I wired up my motherboard and and connected all twelve of output terminals to the SMINI input pins. That will send the output of the DCCOD to the input of the SMINI. The white things are wire markers to keep track of things. I am also making a 3 ring binder with wiring diagrams to help remember and troubleshoot in the future, You can also see cable clamps attached to the backplane to define were my cable groups will be.
What definitions do I use to mark the wires? I use actual RR mileposts just like my prototype. So for ex: I use OS 137 for the turnout at Rockport. BK 1311 and BK 1312 for the double track blocks to the right of Rockport. Why those numbers. That's what the LV did.Number 1311 means milepost 131 track 1 (westbound main) 1312 means track milepost 131 track 2 (eastbound main). That is how I identify everything on my RR BK stands for block. These names will be used in the programming too. Use whatever makes sense to you.
Crimping all these wires is time consuming. There is a device called a breakout board that can really simplify this. It converts this to a screw terminal connection. But that will cost 8 bucks or so adding to the cost of the input/outputs by a dollar or so. I'm on a budget so I went with the crimp pins. I also had the crimp tool from prior electronic projects, so if you don't have that you will need to purchase that. It's about 15 dollars or so. Don't try to skip buying the tool if you go with the crimp pins. You will not get a good joint. Here is a picture
Also per a gentleman who just completed a sucessfull CMRI installation, he is soldering all connections. So I am soldering the pins, plus the spade connectors before I install them. He spent a lot of time chasing down problems that were related to crimped joints that were not good to the naked eye. |
Edited by - LVRALPH on 06/09/2010 06:43:01 AM |
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/07/2010 : 6:14:33 PM
|
I forgot to post the picture of the work I did in the last post.
|
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/09/2010 : 07:02:11 AM
|
Well, today Don Wood should get my SUSIC. I don't know when he will get to work on it, but I am anxious to know where I screwed up,  and even more to get it back, write a program, and hook up a device! |
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/09/2010 : 7:15:38 PM
|
Heard from Don. We will see were this goes. At least he had the same problem. Glad I built the SUSIC right.
Hi Ralph
The SUSIC is working perfectly. The DIN32 card will not read. I changed all the chips out, checked the switch and checked all the traces for addressing for continuity. Nothing obvious is incorrect. The switch settings were incorrect but that is not the trouble and I suspect you tried many different arrangements to see if it would work. When I return it , they will be set to card 1 for you. Refer to Figure 2-16 in User's manual for setting the address
Best
Don
|
Edited by - LVRALPH on 06/09/2010 8:33:43 PM |
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/09/2010 : 8:31:31 PM
|
Just got this from Don![:-banghead][:-banghead][:-banghead][:-banghead][:-banghead][:-banghead][:-banghead]
Found it!
The white headers are the problem, they have too much play in them and are not making contact with the IOMBX. Pushing the card backwards makes the pins connect and it read fine. I will repalce the headers for you
Don
|
|
|
|
simon1966
Fireman
USA
2861 Posts |
Posted - 06/09/2010 : 9:52:01 PM
|
| Ralph, don't bang your head too much, it means you basically got the board correct!! |
|
|
|
MarkF
Engineer
USA
9276 Posts |
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/10/2010 : 05:54:55 AM
|
quote:
Originally posted by simon1966
Ralph, don't bang your head too much, it means you basically got the board correct!!
Simon, I know. Of all the things I could have done wrong such as poor soldering, components not inserted correctly, wrong components, I get beat by a simple mechanical connectivity issue. [:-censored]That was the last thing I suspected during my testing! But that did give me confidence that not only can I, but others can install CMRI with confidence. You can build the electronics and get it to work. I did.
With people on the yahoo group, Don Wood, Bruce Chubb, Bill Carr, Chuck Card, Chris and a whole bunch of other great guys...you can't fail.
So anyone contemplating CMRI to quote the Nike commercial...JUST DO IT! |
Edited by - LVRALPH on 06/11/2010 07:29:30 AM |
|
|
|
MarkF
Engineer
USA
9276 Posts |
Posted - 06/11/2010 : 11:08:12 AM
|
quote:
Originally posted by LVRALPH
So anyone contemplating CMRI to quote the Nike commercial...JUST DO IT!
I plan to! Heck, if you can do it, I know I can! Just kidding. Seriously, I do appreciate your diary on here and the detailed step by step. I only wish I was further along on my layout so I too could get started, but the good thing is that I can start planning for it, knowing I will install it one day. |
Mark
See my homepage at http://home.comcast.net/~prrndiv/
|
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/13/2010 : 06:32:28 AM
|
While I am waiting for my board to come back I made some progress on the hit list to get Rockport interlocking done.
1Rebuild Rockport panel
2Build Rockport signals.
3)Hookup SMINI outputs to terminal blocks.
4)Hookup SMINI inputs to terminal blocks.
5)Install detector motherboard.
6)Install node connector from SUSIC to SMINI.
7)Connect wires on CTC machine for Rockport OS section.
8)Write program to initialize both nodes and operate Rockport and operate devices.
9)Test detectors installed under layout and move to motherboard.
Not a small list, but getting there. Here is a pic of yesterdays progress.
What has been added is as follows. First I need some outputs to operate Rockport. So I connected the first half of card 0 (12 wires)down below to terminal strips. From there I will connect to RR devices. I'm told I can run ordinary telephone wire. You could connect the devices directly to the SMINI, but this gives me a test point and a way to change to a different/cheaper wire. Steam nut says he has a ton of scrap he can get for me, so I'll wait for him before running the wire.
For Rockport I will need 11 outputs. There will be two double headed searchlight signals at each end of the turnout. Each signal head will require two SMINI output pins. One for the green indication, one for red. So 4 signal heads x two = 8 SMINI pins will be needed for signal control. That will fill out one port of 8 pins. I might say here when hooking up signals it is very important not to cross port boundries. Ex: wire one LED to the top head on pin 7 on port A, and the next wire to port B pin 0. I think I know why, but that would be a long technical explanation. I don't want to scare Hueber. Just remember the rule. If a port has 5 open pins and you need 7, connect the signal to the next open port. I'll explain the ports in my next post.
Then the switch machine will need three outputs. One for normal. one for reverse, and one for local control of this turnout. If this turnout was only under dispatcher control, I would only need two pins. Most, if not all my machines will have local control. Comes in handy when restaging the RR.
I also ran the +12V and ground wires to the DCCOD mother board. This powers the detectors. I still need an inline fuse for the +12 power wire. Also the +5 and ground wires were run to the future second SMINI at this location. These are to the right of the central cable and the DCCOD.
I might also want to comment on signal type and outputs. The searchlight signals only require 2 input pins from the SMINI because I am using Oregon signal LED's. They are 2 lead LED's vs 3 lead LED's. Three lead LED's require more input pins and the SMINI configured for current sourcing operation, not current sinking; as comes standard. You will need input pins to drive each LED you have at a signal location. If you have PRR position light, then more SMINI pins will be needed, same with B&O color position. Type D traffic light types will need an output for each LED on each head. One nice thing for me, is the SMINI can give me yellow without using another pin. It does this by ocillating the red and green to make a yellow. I have to put a 500K potentiometer in the circuit at the signal for adjusting the yellow, but that's neat.
I will also talk about grounds at a later date when I install the second node at Packerton.
There is a board you can purchase called an SMC that will allow you to use only one SMINI output pin. I bought one, but am not using it here. I will explain it in my next post about ports.
As this is all coming together, I can see down the line in the distant future, my purchasing more SMINI's or even a SUSIC to give me track route alignment in my yard. It will make it so much easier for visiting yardmasters to dial up atrack nad have all the turnouts align, rather than figuring out all of the independent turnouts. The more you work with CMRI, the more you see what it can do for you. Kind of like eliminating block toggles when you convert to DCC. DCC and CMRI...perfect together.
|
Edited by - LVRALPH on 06/13/2010 07:22:29 AM |
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/14/2010 : 06:30:12 AM
|
Since I am going to soon talk about specific wiring I might take time to explain ports, bits etc.
Lets talk about the SMINI first. All the pins you see are connected to what are referred to as cards. On the SMINI pictured, the bottom is card 0, the right is card 1 and the top is card 2. Card 0 and 1 are output cards, card 2 is an input card on the SMINI. On a DIN 32 card they are all input, and a DOUT 32 all output.
On each of the cards they refer to ports. On the SMINI there are 3 ports A, B. and C. Each port has 8 pins, labled 0 to 7. So the second pin from the left on port A would be known as pin A 1, the 8th pin in port C would be pin C7.
The SMINI has 3 ports of 8 so each card has 24 pins. The SMINI has 48 output and 24 input pins.
When I labled my wires , I labeled the wires like this...1 B 3. That means card 1, port B, pin 3. Pretty easy. When I run these out to devices, I will then convert them to the device marking, such as signal 1371RT, or 1371GT. This means signal 137 westboound (odd # at end) green, top head. There is no set way to number your wires, just something that makes sense. But do do something to keep track of everything.
DIN 32 and DOUT 32 cards have 32 pins so they have an extra port labeled D. Card numbers are based on where you put them in the motherboard.
Hope this makes sense. We will be referring to this soon.
|
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/14/2010 : 1:48:43 PM
|
Well, my cards are back from Don Wood. Assuming no new problems, I am a computer program away from connecting devices and making things work!
The CTC machine came alive today. I connected the +5V LED supply and the ground block to the surplus computer supply. All the locked out LED's came on and operate with their toggles. It's not much, but I remember the excitement Hueber and I had when the first device on LV Alco C-420 # 414 came on. This is similar. |
|
|
|
LVRALPH
Fireman
5430 Posts |
Posted - 06/14/2010 : 3:52:24 PM
|
| I hooked up the DIN 32 card and it reads just like Don said it would. So I spent time reconciling the wiring on the signal and switch levers as well as the code button. I'm ready to program OS 137 Rockport and see if I can make the machine work as well as the RR. |
|
|
|
Topic |
|
Signaling my model railroad
