60 channel dmx board

It's time for another update. Again I have to say that things are progressing but the code is still a while of being finished. I'd hoped that I could get it finished and sent up to the Brissie mini but time has again got away from me. This week is another interrupted week with the long weekend ending yesterday and me having to head to Melbourne to do some work on Wednesday. That leaves me with the rest of today and Thursday morning to finish the code if I was to get a board sent off in time. I'm now enough of a realist to realise that I still have huge amount of code to write and test.
I've made a few mistakes that cost me a bit of time. As the saying goes "assumption is the mother of all stuffups". Well I made a couple of assumptions. The first was I had a quick look at the specs of an RS485 chip I had and assumed that it was going to do the same as the chip I had designed around. Oops. It turned out that the output was inverted and I didn't notice for a fair while and surprisingly enough I couldn't get data in. The second assumption was the the "lights up" program which generates dmx signals would generate a signal that is compliant with the dmx standard. After almost a full day of testing and wondering why I was getting strange results I discovered that it doesn't. See the attached pictures and you can see that there is a spike of data that shouldn't be there. As my code was checking for a break followed by a make after break (MAB) and resetting if there was a problem with the data then having a "glitch" in the MAB meant I was never getting valid code. I chose to ignore the "full on" requirement to detect the break followed by the MAB.
To make my assumptions worse if I hadn't made the assumption with the "lights up" I could have, and had tried to, invert the signal via software which would have fixed my first problem.

I am writing in the following cycles for the test mode to put things through their paces. I can't be bothered doing fades as there is enough code being used already. Each cycle will run followed by the next and it will then go back to the start.
1 SINGLE LED1 THEN 2 THEN 3 ----> 60 . @ 0.1S. TOTAL 6S
2 SINGLE R LED1 THEN 4 THEN 7 ETC @ 0.2S. TOTAL 4S
3 SINGLE G LED2 THEN 5 THEN 8 ETC @ 0.2S. TOTAL 4S
4 SINGLE B LED3 THEN 6 THEN 9 ETC @ 0.2S. TOTAL 4S
5 ALL RED LED 1 + LED 4 + LED 7 ETC @ 1S. TOTAL 1S
6 ALL GREEN LED 2 + LED 5 + LED 8 ETC @ 1S. TOTAL 1S
7 ALL BLUE LED 3 + LED 6 + LED 9 ETC @ 1S. TOTAL 1S
8 AQUA (G+B) LED 2 +LED 3 + LED 5 + LED 6 ETC @ 1S. TOTAL 1S
9 YELLOW (R+G) LED 1 + LED 2 + LED 4 + LED 5 ETC @ 1S. TOTAL 1S
10 PURPLE (R+B) LED 1 + LED 3 + LED 4 + LED 6 ETC @ 1S. TOTAL 1S
11 WHITE ALL LEDS ON @ 1S. TOTAL 1S
12 SINGLE W LED 1+2+3 THEN LED 4+5+6 @0.2S. TOTAL 4S

If anyone would like to see something different tested then now is the time to sing out. With the exception of fades I've ran through everything that i could think of.

Alan
 

Attachments

  • DMXframe.jpg
    DMXframe.jpg
    23.5 KB · Views: 16
  • LightsUp.JPG
    LightsUp.JPG
    117.3 KB · Views: 19
  • DA_DMX.JPG
    DA_DMX.JPG
    111.4 KB · Views: 18
Lights_on_Light_St_Ct said:
ok i am curious and proberly has been asked, but the 60 channel dmx controll board is it or will it be us 110V capatible? i know im new and trying to figure oput best and easiest way to get more controlled channels in my display
Click to expand...
Nah, low voltage only

From the first post:
-full 5V to 35V operating range (this may be extended to 40V)
 
The sagas that I have endured are unbelievable. As it's stated about Edison making the lighbulb "I've found 2000 ways of not making a lightbulb". Everything is working and some time later today I'll post a video of the board going through it's paces. Unfortunately while in the testing phase i noticed that when fading a channel up and down the response was terribly non-linear. At first i thought that this was just the response of the leds themselves. When I hooked up the oscilloscope to do some more definitive testing I found that it wasn't the leds. I discovered yesterday that the chips that I am driving the mosfet transistors with struggle to drive them. I was planning on using the same transistors as is used on the 27 channel boards as they are nice and cheap and will handle a pretty massive load without generating much heat at all. The downside is that they have an input capacitance of about 3000pF. When this is compared to the transistors that David_AVD uses on his DC48 that have a capacitance of around 1/10th of this there is a dramatic difference. What this actually means is that the transistors aren't turning on and off sharply or quickly and therefore aren't reproducing accurately the desired 0-255 levels. I've ordered a bunch more transistors to try so that I can get the trade off between accurate control and high current capability.

Alan (with slightly less hair than last time i posted)
 
Alan,

Thanks for the update. Sounds challenging. I will hold off buying something else for a while yet.
 
Here's the video that I said that I'd post this morning. The camera work and commentary is pretty ordinary but you can't expect too much when the commentator is making things up on the sport while filming with his camera phone in the other hand. It's just a basic demo of the board cycling through it's test cycle and a quick demo of changing one bank from 15V input to a 5V output.

http://youtu.be/foR63icQjK8
 
Board is fully working. I've got some proto boards that I'm finishing off and will offer for sale at a cheaper price as they have some mods on them. I'm expecting revised boards in from China some time real soon and once I confirm that all the hardware mods I had to make are sorted then they will be ready to ship.
For some reason my email notifications didn't kick in and I didn't see the post till now.

Alan
 
The following is a link to the user manual so far. I will probably tweak it a bit over the next few weeks but i think I included just about enough info in there for most people to be able to use the board.

https://dl.dropbox.com/u/31385205/AahMega60%20user%20manual%201.1.pdf

The proto boards that I have should be tested and ready to ship on Monday.

Feel free to comment below or send me a PM if there's something I've missed.
 
I got my test jig finished off a bit later than expected. At 10:45pm I'm a bit late to make todays mail run :)
The 1st board and the test jig did work fine though. I was a bit of a mongrel lining up the pogo pins that join the 2 boards but it means that a board can now be plonked onto the test jig and is connected to 60 leds within a few seconds rather than 5 minutes or so of screwing in dozens of connections. I also added breakout connectors on the board (and jumpers to enable them) so that I can load up as many channels as I want to an amp or 2 per channel of incandescant lamps as a load.

Testing time tomorrow morn :D
 

Attachments

  • IMAG0298.jpg
    IMAG0298.jpg
    932.8 KB · Views: 32
  • IMAG0299.jpg
    IMAG0299.jpg
    990 KB · Views: 41
The prototype boards are $150.
I tested the first 4 of them today and shipped the 1st one of them. There's a few more on the way tomorrow and I haven't got many of the protoypes left. I have 3 that I can't finish until after China gets back from holidays as i ran out of mosfets and can't load the last 3 boards.
 
Heres my new baby all setup, havent turned it on yet :)

IMG_1126.jpg
 
You must log in or register to reply here.
Back
Top