This week I was able to resume work on the terraces after the last two weeks spent working out the MERG testbed problem. As the terraces will only be seen from the front, I decided only to light up the front rooms, and then alternating between top and ground floor i.e. flats 1 and 3 are ground and 2 and 4 are top floor. I did my usual thing of using clear LEDs painted with Tamiya (transparent) Orange to give that incandescent look. Intention is hopefully to put each on its own circuit and somehow automate a on/off cycle to give the 'lived in' look. Clear plastic in for the windows and curtains in place. The doors are also in, but I've had to pait the back of the doors black for the ground floor lighted flats as the doors were mounted on clear plastic for the skylight, and as such internal lighting was showing through the whole door. This week I'll put the window ledges in and then glue the whole front walls into place.
On the MERG side of things I've been researching what various IR detector arrangements are out there. I stumbled on this one site which had a very good solution that's been used by the robotics fraternity and would equally work for model train detection. Many possible solutions entailed developing a voltage compartor circuit between the IR emitter and receiver so that it would know when something has gone over it. This robotics solution has all the necessary circuitry on one chip, an IS471F. This chip also has the ability to modulate the IR emitter's light output to the receiver circuit so that ambient light doesn't affect it. So I found a supplier in Oz and sent away for one of these chips to give it a go. If it works out, then cheaper chips can be found from Chinese electronics sites. The other thing on the MERG side was I was contemplating the AC power supply over the DC, and thought why not try the transformer for the NCE SB3a. It is 15VAC 5A (fits criteria) and I reckon I could have it supply two circuits, one for the DCC SB3a and the other for the MERG CBUS power bus as I don't think it will ever have a full 5As of consumption with my small loco fleet and the accessories to power.
On the Raspberry Pi front, I've been trying to work out a problem with JMRI integration and CPU maxing out, over the weekend. No joy, so I've posted a question on the jmriusers Yahoo forum and will see what answers I get back. I took the Pi to the Club meeting on Saturday and it certainly raised the interest of the electronics buffs amongst us. One member said to investigate using VNC on the Pi to access it without a directly connected terminal but rather from your smartphone or tablet. Something to work on later but if it can it has possibilities in the Club context.
So this week more on the terraces and maybe the IR sensor made up and tested on the MERG CANACE8C.
The first half of the week was spent dealing with the responses I received back to my questions posted on the MERGCBUS Yahoo Group, with many suggestions. On Wednesday were a number of other replies to a follow on set of questions. It was from this group of replies that the answer came. I had comms on the CBUS, I was able to get the input module to work and saw traffic on the CBUS via the FLIM Config Utility that I had connected up. But nothing on the output module. Answers around the LEDs were also sorted out. Then one of the replies re-explained a concept in layman's terms. The output pins are sinks NOT sources. They are -ve not +ve. I had wired in the test LED incorrectly to the output block (a document I was referring to appears to be wrong). DOH! Once re-seated, I tested the switch on the input module and the LED came on. DOUBLE DOH! Apparently it is a common misunderstanding. I will certainly be putting that nugget of info into my Hints and Tips! Now I can continue to test the various inputs and outputs I am looking at. I might have to invest in a couple of different output modules (solenoid motors and maybe servos). So while I am waiting on these to come in from MERG, I can go back to the Terraces as I have been really caught up in the above.
The other thing this week that kept me from progressing with the Terraces was that I had received my Raspberry Pi (microcomputer - look at www.raspberrypi.org) protective case from the UK on Wednesday. I put it together on Wednesday night and now that the Pi was protected I was able to concentrate on getting to know it a bit better over the rest of the week and into the weekend. After much research around how to connect it to an Internet source for software updates, I discovered it had been pre-configured to use the Ethernet conx. I connected it via a cross-over cable to my PC (which was being used as the Internet source) and voila' it connected first time. Brilliant. Last night I was able to download and install JMRI onto it. Excellent. That's where I left it. This week I will potter around a bit and see if I can transfer my current JMRI data files from the laptop to the Pi. This is why I purchased the Pi, a simple computer to run the train control software. I know for a fact that MERG TCC has just this week had a Linux version setup to run on the Pi, now I just have to make queries about FCU for CBUS having a Linux version.
Things are looking promising!
Spare time this week was spent re-reading the technical bulletins for the MERG CBUS modules in an attempt to try and get to the bottom of the problems. A fellow MERG member who lives locally (at least in Brisbane!) gave me some suggestions, which in the end didn't fix the problem. Come Friday I issued a post to the MERG CBUS Yahoo Group to see if a wider audience could nut it out. Many useful replies which furthered my education in these modules but still not working as they should. Hmmm. So last night I installed the USB driver for the CANUSB module and the FLIM Configuration Utility to see if the modules were 'seen' by the FCU on the laptop. The laptop could see the USB but nothing past this point. Hmmmm. Maybe another fault, or something actually wrong with the CBUS itself (wiring, ICs, etc?). So I will have to go back to the Group with this info and see if any bright sparks can get to the bottom of it!
The only thing I did on Illavearf was to 'formalise' the jury-rigged ammeter to a permanent location. I also took the opportunity to co-locate the short warning light.
This week hopefully some useful suggestions back re the MERG modules will get this sorted out, as well as I might be able to get back to the internal lighting of the terraces.
This week I spent getting things finalised for the MERG test bed I've been building. Last night I screwed the USB card into place and wired it up to the CBUS. I put together some LEDs and a toggle and a push button switch to use as part of the 'learning experience'. It was getting late when I tried to 'learn an event' but it didn't work. Might be too tired. Try again in the morning. Come Sunday morning I tried again but still nothing. I referred to the Technical Bulletins and I seemed to be doing everything correctly. It was at this point I had to go and do some chores, so I will refer off to the MERG forums to see if these cast any ideas my way as to what I am doing wrong.
Another thing I did through the week was get the ammeter wired up (jury rigged!) to test on the layout. After consulting Mr Mackie about a switch problem, everything came good. It was sitting on 1/4 A just turned on. Today I put a sound loco on and nothing changed on the ammeter. So I am assuming I might have to 'zero' the meter back a bit so that 1/4A is registering when a loco is on it?
I wasn't able to get down to the Gold Coast Model Train Show this weekend so I am waiting on reported from the guys about how it went.
As the TDF finishes tonight (pity about Cadel) and I have only a few nights worth to catch up on, I am hoping that I will get back to the terraces now and sort out the MERG issue this week.
I had a quick look on the MERG internal forum this morning and there was a posting that someone had the same problem and the trouble was diagnosed as the voltage regulators has been transposed. Had I done the same? So I went and had a look and "doh" I had! So I unsoldered both. Swapped to where they should have been. Connected up the power supply and voila (sorry to much TDF!) the voltages were now correct at the checkpoints. After disconnecting the power I inserted the various ICs, powered back on and the green LED came on to show all was well. Hooray! I then connected the 5VDC output to the 5VDC input on the 'input' board and its green LED came on. Great.
This evening I connected the USB to the 5VDC and its voltages were correct. I completed the assembly by inserting the various ICs and USB port. Power restorted and its green LED came on. I am a happy camper!
Now I just have to refresh myself with how to program the boards with events, which will require a switch and a LED. Then I have to setup the USB drivers on the laptop.
I did note on the USB maker's site that they currently have no USB drivers for Linux on ARM processors. Well that rules out the Raspberry Pi connecting directly to the port.
Well a good start to the week.
I ordered the 12VDC transformer earlier in the week, so while waiting I started on putting together the USB interface kit from MERG. Once again it was an easy assembly up to the point of needing to check for voltages, and this is where I stopped as I needed the transformer. No further work on the terraces as I had a busy week apart from checking that I had the various components for the LED lighting.
The transformer arrived Friday. I had to get a 2.5mm DC connector to connect the transformer plug to the test bed / bus. So off to Jaycar in the morning to get this and a 6A bridge rectifier for the ammeter mentioned in the previous post. Saturday afternoon was done at the club for a few hours. Firstly a working bee on the HO layout to make a start on cutting the various straights and curves for around the helix blob. Then the monthly meeting followed, and afterwards some more cutting. Later in the arvo I spent some time in the shed getting the DC connector mounted to the test bed.
Saturday night I spent wiring up the CAN / Power bus into the test bed and into the Input and Output kit boards. To my surprise the boards where not getting 5VDC power! 12VDC was going into the Output board which would then supply 5VDC to its own electronics as well as the power input to the Input board. Nothing. I followed the instructions again to make sure everything OK, and it appears to be. I'm stumped! I'm going to have to do some research through the forums to see if this has happened before and if not make contact with a fellow MERGer to see if he has any ideas what might be happening.
Sunday afternoon I quickly did a temporary hash-up of the ammeter to the bridge rectifier and track power (3Amps max through the SB3A booster). Immediately the 'short' light came on and the ammeter was registering about 1.5Amps. I wonder if this was supposed to happen? I'll have to check with Mr Mackie.
Well this week will be spent trying to sort out the MERG kits' power supply problem and maybe install the LED lighting into the terraces so that I can continue the build.
As expected the TDF is getting in the way of modelling as I am spending the evenings catching up with the previous night's recorded session. However I was able to start the second MERG kit over the weekend and I am now up to the point of needing the power supply, so I will be ordering one of these this week. After my conversation with Ian Both, a fellow MERG member in Brissie, I have decided to accept his suggestion and use a 12VDC 5A transformer. It is cheap enough and if I need to revert to the AC based adaptor then I will.
No movement on the terraces as I have to sort out some LED lighting to go into them before I seal the front walls. I only have to worry about the front windows as no one will see the back windows.
The other thing of interest was I posted a query on the MERG yahoo forum querying the pronounciation, is it like 'berg' or 'merge'? Well did that throw up any interesting chain of events. Apparently both could be interchangeable as MERG was originally formed from the merger of two existing clubs, but you could use it as 'berg' as the 'G' is a hard G as in 'Group'. I think I will sound it out like 'berg'.
Mr Mackie took a delivery of 12VDC ammeters from ebay last week and a cuple of spares. I will purchase one of him and setup the DCC circuit as per Marcus Allman's page for Illavearf because since upgrading to an SB3a I have lost the inbuilt ammeter feature of the PowerCAB.
So this week I might get around to the LED lighting in the terraces OR start on the MERG USB kit. Not sure. Club monthly meeting this Saturday so maybe a working bee on the new layout before the 2pm meeting. We will see!
