Wanted - new project ideas!

[David_AVD]

Have you ever had a Christmas related electronic project that never saw the light of day?

Was it because you lacked experience in circuit design, PCB layout, PIC code, etc?

ACL has several members (including myself) with these skills. Maybe we can help you with these projects.

So, post your project ideas here and we'll see what's possible.

 

[AAH]

What I'd love is something like the DC48 that you're developing David but it would 100% emulate a LOR controller. Probably have to emulate 3 controllers really as they are locked to 16 channels. If I could get away from having to put a DMX adaptor between my LOR software and some new controllers it'd be sweet. This has been in the back of my mind as a project for the future but with your DC48 nearing completion and with the amount of work i keep not saying no to it could be never before I even started looking at doing this.
You're a glutton for punishment David.

 

[David_AVD]

"glutton for punishment" ? Nah, just passionate about nerdy stuff.

As for the LOR emulation, maybe hang off until LOR release their new DMX dongle?

 

[dmoore]

What I'd love is something like the DC48 that you're developing David but it would 100% emulate a LOR controller. Probably have to emulate 3 controllers really as they are locked to 16 channels. If I could get away from having to put a DMX adaptor between my LOR software and some new controllers it'd be sweet. This has been in the back of my mind as a project for the future but with your DC48 nearing completion and with the amount of work i keep not saying no to it could be never before I even started looking at doing this.
You're a glutton for punishment David.

There is a guy here in the US that did write firmware for a controller he developed that was completely LOR compatible, including the 512 extensions. It also did DMX just like the LOR firmware. To my knowledge, it still sits on the shelf to this day. He indicated that while it only took him less than a week to write DMX firmware (having never seen DMX before writing it), it took him several months to completely reverse engineer and write the LOR firmware.

Actually the LOR protocol isn't limited to 16 channel controllers, they are limited to 512 channels per controller – this is why you can use the iDMX as a single 512 channel “LOR” controller. You can see more about how the LOR protocol operates here: http://www.holidaycoro.com/2010LSHWorkshop/DMX%20and%20LOR%20Protocols.pdf

While the LOR protocol is wonderful for lower levels of channels, say around 700 when you start getting complex commands sent to a lot of channels all at once (twinkles, shimmers) you can end up with "stickys" as the network becomes saturated and commands cannot go out. While DMX may not be the most efficient protocol, as it re-sends the same data over and over 44 times a second, even if it has not changed, you at least know what you will be getting at the other end.

 

[dmoore]

I'm still all for a portable tester. Many have talked about it, a few have gotten close (http://diylightanimation.com/index.php?topic=2697.0) but still no DIY tester.

http://forums.auschristmaslighting.com/index.php/topic,888.msg7142.html#msg7142

 

[joelrose]

What I have wanted for years is a high channel count individually adjustable constant current driver for LEDs.

I could see a DMX front end driving some control chip that did DMX in, serial out, to a STP16CPS05 or a similar constant current driver.
A jumper would set the current to 20 or 100ma on each chip.
By high channel count I would see 64 channels as the minimum.
The interface connector would be the biggest issue.
Even using a .1 header it would need over 6 inches of space.

You asked, so just throwing it out there.

Joel

 

[chilloutdocdoc]

I'm still all for a portable tester. Many have talked about it, a few have gotten close (http://diylightanimation.com/index.php?topic=2697.0) but still no DIY tester.

http://forums.auschristmaslighting.com/index.php/topic,888.msg7142.html#msg7142

Let me finish your darn program first david!

Edit: (serious time)

I think constant control Development is definitely in the future, and possibly more for this year, I'm still interested in a RGB(W?A?) flood controller using DMX or PoDMX.

Another idea I've had is a Constant Current "Breakout" for lack of a better term, basically it would be put inline with whatever existing controller that is used, and it would provide constant current. My idea was to mainly pair these up with a $6 dmx controller for an ultra cheap DC LED box. (Just a thought.

 

[David_AVD]

All interesting ideas guys.

Sounds like Josh is on the case with the DMX tester.

Joel, there are dimmable (internal pwm), constant current chips available, but they tend to be either low channel count, low current or a bit expensive. The STP16CPS05 doesn't have the internal pwm, so would need to have pwm applied via the data which I think would make the firmware tricky for large channel counts.

Just to clarify, I don't have time to completely design and build every project that comes along. This thread is for getting those new project ideas out into the open and moving them forward where practical. Hopefully I and others can share technical advice, promote discussion and get some of these projects completed. I can do some schematic and PCB design work if that helps too.

 

[chilloutdocdoc]

Maybe... We'll see how much time I wind up having, I do want to have my own display done this year, but having a DMX tester for myself would be quite useful.

 

[mrpackethead]

What I have wanted for years is a high channel count individually adjustable constant current driver for LEDs.

I could see a DMX front end driving some control chip that did DMX in, serial out, to a STP16CPS05 or a similar constant current driver.
A jumper would set the current to 20 or 100ma on each chip.
By high channel count I would see 64 channels as the minimum.
The interface connector would be the biggest issue.
Even using a .1 header it would need over 6 inches of space.

You asked, so just throwing it out there.

Joel

http://www.linear.com/product/LT3746 heres something that might be semi useful

 

[AAH]

What I'd love is something like the DC48 that you're developing David but it would 100% emulate a LOR controller. Probably have to emulate 3 controllers really as they are locked to 16 channels. If I could get away from having to put a DMX adaptor between my LOR software and some new controllers it'd be sweet. This has been in the back of my mind as a project for the future but with your DC48 nearing completion and with the amount of work i keep not saying no to it could be never before I even started looking at doing this.
You're a glutton for punishment David.

There is a guy here in the US that did write firmware for a controller he developed that was completely LOR compatible, including the 512 extensions. It also did DMX just like the LOR firmware. To my knowledge, it still sits on the shelf to this day. He indicated that while it only took him less than a week to write DMX firmware (having never seen DMX before writing it), it took him several months to completely reverse engineer and write the LOR firmware.

Actually the LOR protocol isn't limited to 16 channel controllers, they are limited to 512 channels per controller – this is why you can use the iDMX as a single 512 channel “LOR” controller. You can see more about how the LOR protocol operates here: http://www.holidaycoro.com/2010LSHWorkshop/DMX%20and%20LOR%20Protocols.pdf

While the LOR protocol is wonderful for lower levels of channels, say around 700 when you start getting complex commands sent to a lot of channels all at once (twinkles, shimmers) you can end up with "stickys" as the network becomes saturated and commands cannot go out. While DMX may not be the most efficient protocol, as it re-sends the same data over and over 44 times a second, even if it has not changed, you at least know what you will be getting at the other end.

It'd be nice to know how/where to get the info on that LOR compatible controller. Failing that a nice LOR->DMX adaptor that wasn't $260 like LOR wish to charge would be handy. I personally hate developing microcontroller software in particular anything to do with comms. The hardware side of development is a pleasure but the comms does my head in.

 

[David_AVD]

I personally hate developing microcontroller software in particular anything to do with comms. The hardware side of development is a pleasure but the comms does my head in.

I find comms is pretty easy. The shimmer / twinkle effects would be reasonable CPU intensive and not altogether trivial though.

 

[dmoore]

It'd be nice to know how/where to get the info on that LOR compatible controller. Failing that a nice LOR->DMX adaptor that wasn't $260 like LOR wish to charge would be handy. I personally hate developing microcontroller software in particular anything to do with comms. The hardware side of development is a pleasure but the comms does my head in.

The guy that wrote the firmware never released the controller. He showed it to LOR and that product is what you know as the Signal-Over-Power solution LOR is about to sell in their upcoming controllers. If you really need it, PM me and I'll put you in touch with him. Be aware that he won't get it away for free.

Again, personally I think the LOR protocol, while perfect for what we were doing several years ago was great - moving forward I find it hard to believe that it will be a good solution given it's complexity, limited support and other technical issues.

 

[David_AVD]

One idea I had was for a low channel count, but higher current SSR.

I imagine that some displays use a bunch of DC channels and need just a few high current ones for a big run of non-digital strip, etc.

The new SSR would actually get it's input from your normal DC controllers. This way you can add a few high current channels to any regular DC controller.

 

[David_AVD]

Well, it looks like what I was thinking of doing (the SSR) has already been invented.

For some reason I was thinking the standard SSRs used a switched positive on the inputs. I was going to do a switched negative one, but it looks like the existing ones are that way anyway.

So, feel free to ignore this and my previous posts!