It's all about the blinking...
Section: |
After
a static display in 2006 I finally mustered the nerve to venture into
animated lighting for the 2007 Christmas display. Below is a
description of each main section and pictures of each. Clicking on the
pictures will give a larger image.
The show consisted of 19,022 lights, 122 channels, and music transmitted on 107.1FM. The show was driven from a Toshiba A70 Laptop PC, 3.06Ghz, 1GB RAM and 120GB HDD running XP Pro. Initially the show ran on Vixen 1.1.11, but on about the 20th of December I decided to change to V2 beta 4, and besides a few issues when the show initially started, it ran really well. Thanks to KC Oaks for such a great program. The ControllerAfter
the initial count of lights (50 sets), and the option of Dimming or
non-dimming, a basic 128 channel Olsen 595 controller was chosen in
October 2006, but rather than pay rather expensive prices for the
boards, I chose to create a stripboard design. 64 channels per
stripboard sounded great, but rather than struggling for channels if I
decided to increase the light count before Christmas, I decided to
make 128 channels.
Darn hardware costs here are expensive compared to the USA. They say 1000 channels for under $100, well not really, but based on their hardware costs, 64 channels, complete with the triacs that are needed, ready to go, might be US$100, so thats under $150 here. Not likely - looking at Jaycar and that controller will probably cost $300 to $400 to build - OUTCH! Along came Futurlec.com.au and it looks like costing just $150 to $200 in parts - a LOT different.
Above
is a picture of the controller mounted on the top of an old PC case.
You can just see the 128 LED's that indicate the lights status. Because
of the current draw, they will be scrapped before next years show. The SSR's Because
of the stripboard design of the 595, it made sence to continue with the
same design and create a stripboard SSR setup. After researching, I
found that 16 SSR;s easily fit on a single stripboard. I toyed with the
idea of 24 on the one board, and created a design for it, but some of
the gaps between the 24V and PC side were rather small, so I
went with the larger gap to be safe.
The
24V SSR's are mounted on the baords at the back, and the 240V prototype
SSR's are mounted at the front. The end versions of the 240V ones only
had 2 of the 3 Triac pins put into the board, and the third was mounted
directly to the 240V wire & covered in heatshrink to insulate
it.
Here are the 24 & 36V SSR boards mounted in place. A fan blows cooling air directly over the SSR's, keeping them very cool, even under loads in excess of 2A. All the connections were designed to handle 2 Amps, but the loads were kept around the 1A per SSR to further reduce the chance of overheating. This is NOT a 1960's PC from NASA, although it could easily be mistaken for one lol
A closeup of the 240V SSR setup. Next year I might redesign them to be a little more robust and provide a greater distance between the control and 240V wiring. I actually tested these at 2 Amps for a few hours on at a time and the Triacs just got warm to touch. At 3A they got rather warm to hot, so a 1A limit has been placed on all outputs to avoid issues. Rope lights were tested at 0.1 sec on, 0.1 sec off for over 24hrs and no problem was encountered with any of the outputs. Each output was load tested for 5hrs at 2A and 24hrs of on/off testing . Here you can see the CAT5 connection to the SSR. The earth pins have been filled with araldite to stop any earthed plug from being inserted. A label was placed on the SSR boards stating they are unearthed and 1A max per output. Considering only a few of my rope lights draw even close to that, there will be no problems. The BoxBecause in Australia we need 24V and 36V supplies to lights, I opted for large 300VA toroidal transformers to provide power to the lights, and a plug & socket arrangement on the case to get the power out to the field. CAT5 connections to the 240V SSR's were also mounted on the front. Another reason for opting for Toroidal transformers is that rapidly switching the standard plugpacks (wall-warts for the Americans) causes them to fail - rather spectacularly actually. I tested 6 plugpacks and all 6 failed, with 3 of them actually catching fire, and that was 1 sec on 1 sec off testing.
Thats the lower section of the box, from the rear side. This side will also contain the output from the box to the lights (via the RCA plugs similar to the rear - they have been tested to 3A for a few days without issue, but max current will actually be <1A per connection).
Some of the back connections. Each transformer has 3 outputs (Well I make it 3), with a 5A fuse in each. The 3 outputs go to the connections on the back, and the other side will eventually go to the SSR's when they are mounted at the top of the box.
The rear of the case has a fused, but unswitched output from each transformer, 3 power inputs (one per pair of transformers), and one of the exhaust fans. I guess with 1800VA of transformers, 128 channels of 595, 96 SSR's as well, it's going to get a little "toasty" in there at times. I have one intake at the bottom front, an exhaust on the rear, there will be one more at the rear (where the power supply is), and another inlet blowing over the SSR's at the front, so the airflow should be more than adequate.
Thats an overall of the box. Only one 36V and three 24V transformers are currently installed, and there is room on the upper support, for two more. That support is a 6mm steel bar which is supported and screwed down at each end, and easily holds the weight of the transformers.
All this stuff needs some cooling too. I added a front & back fan down low, and thought one blowing overt the SSR's would be a good idea too. Behind the case is a pieve of board angled up. This board also has two "Fins" to further direct airflow up through the SSR's to keep them cool. With the sides of the case on, theres a huge airflow through the case - perfect for keeping all this stuff cool.
Here is a great image of the side connections. Theres 96 SSR's connected to this side section, and each group of 16 has 19 cables in total running to this side. I know that the cable is WAY too big, but because of the bundling here I wanted to ensure that the cables had minimal resistance, thus at 1A they would heat up only a very small amount. Each SSR is connected to the output terminal with 1mm cable, and each group of 4 or 5 return cables is conected back to the neutral of the transformer with 1.5mm cable.
Well
theres the completed connection block on the side. It actually took 2
people to push that side down so it could be closed, so I dont want to
open it again in a hurry! The ShowWell being the first person in my town to have animated lighting, it was a huge drawcard. The lights went live on 1 December 2007 with a single song (Aussie Jingle Bells) after a few days of testing and tweaking. On the first few days I got 1 or 2 cars, then it slowly increased. After the first week I added my version of Wizards In Winter plus a short 30 second song where all the light are on. The cars picked up to 10 a night. We had one minor vandalism issue, so I purchased and installed a Swann wired camera setup. 4 cameras covered the entire yard, and the PC-independent recorder records for 4 days in a loop. Next year I might install a 500GB drive so it manages to record for more than a week at a time. There were no more problems during the show. By about the 15th of December we were getting about 20 to 25 cars a night and we were giving free Candy Canes out to people when they arrived to watch the show. Most had heard about the show from people who had come in the previous days, and some people were actually ringing people and telling them to come and watch the show while they were parked outside. On the 19th I added the last song - a 2.5 minute cut-down version of Christmas Canon Rock. We were interviewed on the local news (and by my local paper) and it was played a few days later, further increasing the traffic to about 60 cars a night. The interview is located at the bottom of the page. By Christmas Eve the traffic had gotten a little out of hand, with about 200 cars a night, with people waiting in line 250 meters down the street and slowly moving up - often they had heard the entire 9 minute show before they got to see it close up. Although we were not accepting donations to a charity, a few people insisted on giving money, and the total was $32 & a few cents, so a $35 donation was given to Careflight on their behalf. The lights were turned off on New Years Eve at 11:50pm after what I consider a great month. I learnt a lot this year, and next years show will be bigger and better (I HOPE!!!)
This is the PC and control box in the corner of the loungeroom. The small FM transmitter can be seen beside the PC. I had purchased another transmitter, but it turned out to be a piece of junk, so this one was used for the entire show - 1 set of batteries lasted the whole 4hr show, and you can see a few spare sets there fresh from the battery charger. The Video'sThis year I did not spend much time at all getting video's of the display, so I apologise for the very poor quality of the ones placed online. The first story is my local news doing a story on the lights. I expected a "Filler" at the end of the news lasting maybe 30 seconds, but instead I was given just under 2 mintues in the first 10 minutes of the news.
|
