PDA

View Full Version : How I Did: GECE Snow Globe + Renard-RGB Retrofit



djulien
02-05-2012, 03:34 AM
Attached below is a writeup on a GECE Snow Globe and retrofitted Renard-RGB controller that I used this past season. The plug-in and firmware are also attached.

Here is a short demo of some of the effects that were used:

4ldsmrYhAKY

(sorry about the poor video quality - I just grabbed some quick video using a digital camera)

Here is a daytime video of the structure:

pMwQAP1CWzA

A playlist of 2011 videos with the Snow Globe in it are on YouTube at:
http://www.youtube.com/playlist?list=PLF58C3EDFF15916A9

Since the File Library doesn't seem to be available, in case something happens to the attachments there is also a copy at the following addresses:
http://downloads.eshepherdsoflight.com/HowIdid-GECESnowGlobe+Renard-RGB.pdf
http://downloads.eshepherdsoflight.com/RenardRGBFiles.zip

I did a little clean-up work on the plug-in after the show season was over, so if something is broken let me know and I'll take a look.

EDIT: fixed channel order in plug-ins 2/20/12

don

TimW
02-05-2012, 05:11 AM
Thanks for sharing this Don

Clever. Inspirational. And best of all, very DIY.

Tim

JerryPlak
02-05-2012, 05:23 AM
Don,
this is a cool set-up thanks for sharing

LabRat
02-05-2012, 08:53 AM
Fantastic write up. Thank you for taking the time to detail your work, and for sharing it so openly with us.

jrock64
02-05-2012, 12:04 PM
WOW!!!!
You continue to amaze me with what you can do.
If I did not already have the E680s on hand this would be the way to go.

The only thing it does not do is E1.31.

Real interested in the FxGen module.
Sounds like something the Dev team needs to take a look at if they are not already.

Impatiently waiting for your next writeup.

Joel

jaywalk101
02-05-2012, 12:39 PM
very nice! I never even considered using Renard controllers for pixels.... didn't know it was possible.. this is DIY all the way... :)

kychristmas
02-05-2012, 01:08 PM
Your timing is amazing Don and as always the quality of the descriptions and the write up is excellent.

Last night I was going through my stuff trying to determine what I should sell and what I need to purchase for what I want to do in 2012. While looking through my list, I realized there was a significant flaw in my designs for 2012. I will be connecting to my neighbors house using Dirks Ren-W setup. The problem is that I purchased a bunch of GE's and a E680 controller with plans to use them over there. I was thinking to myself that I was going to have to setup some kind of Wireless Bridge to make it work or bail on my plans for Pixels on the remote location. I do realize that Ren-W compatibility may not be possible, but I'm more hopeful about it than I was last night.

Cut to this morning and I read this thread. At first glance I thought it was simply another Great eSOL prop how-to. Then coming back to read, I catch the whole Renard Pixel stuff. I have read through everything and can't wait to try it out with the strings I have here. I love the whole E1.31 stuff, but for now I would like to be able to play around with Renard and have these options. I will also sit back and wait to see what Mac and Tyler do with this and the Simple concept.

For those of you coming to this thread and are not familiar, Don's website is loaded with these with these types of write-ups. If you are really into DIY props, you can easily spend an afternoon or more looking through his stuff. Of course, I encourage you to do so. You won't regret it. Even if you don't build exactly what Don has, you will learn techniques and concepts to apply to your own designs.

Thanks for your great write-ups and descriptions.

Kelly

budude
02-05-2012, 01:52 PM
A W E S O M E !

Great stuff there Don - I was hoping to see a simpler pixel controller like this for smaller display items I'm toying with.

raplass
02-05-2012, 04:52 PM
For those of you coming to this thread...

Amen Kelly, I have learned an awful lot from Don's excellent write-ups as well. Highly recommended.

dmcole
02-05-2012, 06:02 PM
Don:

Very impressive.

\dmc

michaelc
02-05-2012, 08:32 PM
Very Nice. I may want one.

djulien
02-06-2012, 02:21 AM
Thank you all for your comments!


I will also sit back and wait to see what Mac and Tyler do with this and the Simple concept.

Kelly, I am not too familiar with the Simple Renard, but on the schematic it shows an 18F PIC rather than the 16F688 PIC that is used in other Renard controllers. In that case, the firmware I posted for the "retrofit process" would not work as-is, and would need to be modified. I think it would still be doable, just not quite as "plug and play" as the process I described in the writeup.

Also, assuming the Ren-W doesn't affect the data stream, I think it might work but please allow extra time for testing just in case.

don

Mactayl
02-06-2012, 09:12 AM
I will also sit back and wait to see what Mac and Tyler do with this and the Simple concept.

Kelly





Kelly, I am not too familiar with the Simple Renard, but on the schematic it shows an 18F PIC rather than the 16F688 PIC that is used in other Renard controllers. In that case, the firmware I posted for the "retrofit process" would not work as-is, and would need to be modified. I think it would still be doable, just not quite as "plug and play" as the process I described in the writeup.

Also, assuming the Ren-W doesn't affect the data stream, I think it might work but please allow extra time for testing just in case.

don

Don,

This is quite an interesting project you have going with the GECE + Renard-RGB.. it sounds like it could be implemented on the Simple Renard's, I would be willing to work together if you like. I tried looking at your file but it is hex and so I stopped there. BTW very nice videos and write up.

djulien
02-06-2012, 10:38 PM
it sounds like it could be implemented on the Simple Renard's, I would be willing to work together if you like. I tried looking at your file but it is hex and so I stopped there.

Mac,

I'm PM you.

don

kychristmas
02-06-2012, 10:41 PM
Mac,

I'm PM you.

don

Ah, my work is done here.

djulien
02-07-2012, 02:00 AM
Ah, my work is done here.

Not so fast, Kelly - There's plenty of work to go around. ;)

don

kychristmas
02-07-2012, 09:46 AM
Not so fast, Kelly - There's plenty of work to go around. ;)

don

Nope, I'm management, I just get the Parties together :) Seriously though, I'm willing to do whatever is necessary to test. Once I get it working in a wired environment, I will test the Ren-W. I may also be able to help with the Ouput Plugin and later update needed for Vixen 3.0.

I did get to play around with this last night on my Ren 64x. While I didn't have too much success, it was a great distraction from the nasty stuff going on at work. I may need to pick your brain a bit.

I was simply trying to run a single 36ct GE string. I tired both Dec and Inc setting with no results. If I tried to do 108 Channels (1x36x3) It gave me Array errors. I switched to 1200 and it seems happier with that, but it still didn't work.

Thanks for the info.

djulien
02-08-2012, 01:09 AM
I did get to play around with this last night on my Ren 64x. While I didn't have too much success, it was a great distraction from the nasty stuff going on at work. I may need to pick your brain a bit.

I was simply trying to run a single 36ct GE string. I tired both Dec and Inc setting with no results. If I tried to do 108 Channels (1x36x3) It gave me Array errors. I switched to 1200 and it seems happier with that, but it still didn't work.

The plug-in assumes multiples of 1200 or 1248 channels, even if you have less. Sorry about that. I should probably fix that. If you are trying a single 36 ct string, go ahead and define 1200 channels for the plugin.

If there is just one string, connect its data line to RA4 and its ground to a ground on your Renard board or power supply.

Also, the GECE string needs to be powered up at the same time or before the Renard-RGB firmware.

If you just power up the Renard-RGB firmware with a GECE string connected (no Vixen connection), it should go into a test sequence after 2 seconds. It will run the test pattern until Vixen data is received. If you are using an RS485 receiver and the serial line is open, that seems to generate a bunch of null chars, which pre-empts the test pattern so it won't show. The test pattern will show on all strings, so you can use any of the 8 I/O pins for a preliminary test.

What are your setup details? I might have forgotten some important details or made assumptions that I shouldn't have.

EDIT: Another test would be to connect an LED and series resistor (~ 100 ohms) between another one of the I/O pins and ground. If GECE commands are going out over the I/O pins, the test pattern will show as some blinking of the LED.

don

kychristmas
02-08-2012, 10:49 PM
It does seem to run the demo on the string (Random Colors?), but it never changes. It's almost like the plugin is not sending the data. Have you used just the plugin without your FXGen plugin? My COM settings are correct because I'm using another Ren 64XC on the same com port.

I'll keep playing with it. Let me know if you find anything.

Kelly

djulien
02-08-2012, 10:58 PM
It does seem to run the demo on the string (Random Colors?), but it never changes. It's almost like the plugin is not sending the data. Have you used just the plugin without your FXGen plugin? My COM settings are correct because I'm using another Ren 64XC on the same com port.

What do you see for the demo? It should turn on about the 6th pixel one color, then 1 sec later the 5th another color, ... (I forgot to comment that out of the firmware). Then after that it should turn them all off and turn them on one at a time from begining of string to the end. It will repeat that pattern for red, green, blue, magenta, cyan, yellow, white and then repeat. Is that the pattern that you are seeing with no Vixen data?


Have you used just the plugin without your FXGen plugin? My COM settings are correct because I'm using another Ren 64XC on the same com port.

Yes, I tried it without the FxGen plugin a few times okay. What is the position of the other Ren64 in the COM chain relative to the GECE one? Everything between Vixen and the Renard-RGB firmware needs to use the custom firmware and Renard-RGB plugin.

don

kychristmas
02-08-2012, 11:24 PM
hmm. Definitely not the "Demo" pattern I'm seeing. And like I said, it doesn't seem to change when I start the sequence. I have a single Ren64 on the Com port. I was switching them out to test the COM port.

What is strange is that if I plug the string in, but leave the Ren 64 unplugged, nothing happens. When I plug the controller in, the demo starts. So it seemed to be working, but the pattern is nothing like desribed. Much more random than that.

I'll start over and reburn the pics and see where that takes me.

Thanks,
Kelly

djulien
02-08-2012, 11:32 PM
What is strange is that if I plug the string in, but leave the Ren 64 unplugged, nothing happens. When I plug the controller in, the demo starts. So it seemed to be working, but the pattern is nothing like desribed. Much more random than that.

I wonder if the 36 ct is behaving differently than 50 ct?

If possible, could you shoot a quick video of the entire string? There might actually be a pattern in there somewhere.

Another thing to try would be to go ahead and connect the COM port after the demo pattern is running. If you then play a sequence using the plugin, or use Putty or Windows Terminal Emulator and send a few Sync characters ("~") then the demo pattern should stop because the firmware thinks it is getting Vixen data.

EDIT: One other thought - if your random pattern starts off changing 1x/sec and then after about 5 seconds it speeds up to changing ~ 5x/sec, it is likely a grounding problem. I saw "random" patterns a couple of times when I forgot or did not correctly connect the ground from the Renard to the ground line on the GECE string - I did that even after looking at a diagram in the GECE info thread, so you might want to measure the voltage difference between the one you think is ground on the GECE string and the ground on the Renard just to be sure. One time I had also connected to the +5 instead of ground wire on the GECE string, but I don't remember if that caused a random pattern or just nothing.

don

kychristmas
02-13-2012, 11:23 PM
Thanks for the edit. I think I may have tried to hook to Ground, but now I actually think I hooked to positive (+) because it seemed to shut everything down. Until now, I thought there might be a grouding issue. I couldn't tell from your picture what was the Ground wire was. Just to be clear, I was simply planning on hooking Pin 7 on the Rj45 to the Ground on the GE power. Is that what you did?

Been pretty busy with Volunteer activities and work. Wish I had more time to play. Hopefully, I'll be home early enough tomorrow to take a look and verify the ground wire.

btw... I have now tried this on 2 different computer using both RS-485 and RS232 and with 2 different Ren 64s and a Ren 24LV in SSR mode. With the same results.

Thanks again for following up Don.

Edit: I couldn't stand it, so I went upstairs and stripped the other wire on the Power supply and hooked it to Pin 7. I plugged it in and then powered up the Ren64. Sure nuff, the test sequence started run. Cycling nicely through all of the colors. It took all of my willpower not to hook the controller up to the laptop. If I did, I would stay awake till 3AM playing with it. I would probably have a basic matrix built, but it would be a very long day at work tomorrow. Especially since the work day begins at 6AM.

THANKS DON. I can't wait to mess with it over the weekend.




I wonder if the 36 ct is behaving differently than 50 ct?

If possible, could you shoot a quick video of the entire string? There might actually be a pattern in there somewhere.

Another thing to try would be to go ahead and connect the COM port after the demo pattern is running. If you then play a sequence using the plugin, or use Putty or Windows Terminal Emulator and send a few Sync characters ("~") then the demo pattern should stop because the firmware thinks it is getting Vixen data.

EDIT: One other thought - if your random pattern starts off changing 1x/sec and then after about 5 seconds it speeds up to changing ~ 5x/sec, it is likely a grounding problem. I saw "random" patterns a couple of times when I forgot or did not correctly connect the ground from the Renard to the ground line on the GECE string - I did that even after looking at a diagram in the GECE info thread, so you might want to measure the voltage difference between the one you think is ground on the GECE string and the ground on the Renard just to be sure. One time I had also connected to the +5 instead of ground wire on the GECE string, but I don't remember if that caused a random pattern or just nothing.

don

djulien
02-14-2012, 03:39 AM
Edit: I couldn't stand it, so I went upstairs and stripped the other wire on the Power supply and hooked it to Pin 7. I plugged it in and then powered up the Ren64. Sure nuff, the test sequence started run.

Kelly, glad you got it working! (at least the test/demo pattern) I'll be interested to hear how the rest of it works out.

That's interesting that we both made the same mistake initially. I think I originally worked from a photo I saw in this thread (3rd page):
http://doityourselfchristmas.com/forums/showthread.php?13062-RGB-LED-s-Now-Consumer-Grade-Hackable

but now that I look at it again, I see that the top and bottom photos are reversed, and so the text in the middle is ambiguous because it could refer to the top or the bottom photo. The safest approach is to double-check with a meter.


I think I may have tried to hook to Ground, but now I actually think I hooked to positive (+) because it seemed to shut everything down. Until now, I thought there might be a grouding issue. I couldn't tell from your picture what was the Ground wire was. Just to be clear, I was simply planning on hooking Pin 7 on the Rj45 to the Ground on the GE power. Is that what you did?

In the diagram on the middle of page 9 in the PDF, I tried to show the 3 ground/data/out lines from the stock controller with the ground being the wire closest to the other 2 wires from the stock power supply, but maybe that is not too clear. Here is a labelled photo that is hopefully a little clearer:
13232

For the RJ45, I can't quite tell from the photo of the Renard PCB I'm looking at as to whether it is pin 7, but if that pin goes to Ground on the Renard, then yes that is the correct pin.

don

kychristmas
02-19-2012, 06:14 PM
I had a bit of play time available today, so I revisted this GE stuff. While it's certainly better and is definitely running some sort of test pattern, I can't seem to get a grasp on what it's doing.

I have now confirmed that it is grounded correctly and the Data line is hooked to RA4 of the first PIC. Here's what I get...

Test Pattern:
It starts at the 2nd Pixel and then cycles colors start with Red, Green, Blue, Yellow, Cyan, Violet, Warm White?, Cool White. It changes the color one pixel at a time. When it reaches the end, it goes back to the second Pixel and start with the next color.

Sequence:
When I run a sequence, it is definitely reacting to it, but not as I expected. I created a sequence that I believed would light each pixel RGB. 1st Pixel Red, 2nd Pixel Green, 3rd Pixel Blue. But that's not what is happening. I end up with the first 4 pixels doing nothing and then all Pixels are either Red or Green. The good news is that they are clearly reacting to my Fades, so I know they are working.

Vixen Channel Tester:
When I start going down the Channels on the Channel Tester, nothing happens until I reach the 4th Channel. Which actually lights Pixel 26 Red, Ch 5 lights 26 Green and ch 6 lights 26 Blue. 7,8,9 light Pixel 27 in the same manner. Going down the Channels reveals that Pixel 25 is controlled by 100,101,102.

I see the above results no matter what mode I select in the Pluging on the GE Tab. "Inc...", "Dec...", or "Snowglobe..."

I'm sure I could figure it out, but I'm assuming there a simpler reason this is happening.

Any thoughts? Has anyone other than Don played around with his Plugin and Firmware?

jaywalk101
02-19-2012, 06:47 PM
I've been reading this forum and like the idea of having renard protocol running my GECE strings... I'm assuming I can't use my Renard SS16 controllers, but it sounds like my Ren48LSD could be re-purposed for this?

Barnabybear
02-19-2012, 06:57 PM
I had a bit of play time available today, so I revisted this GE stuff. While it's certainly better and is definitely running some sort of test pattern, I can't seem to get a grasp on what it's doing.

I have now confirmed that it is grounded correctly and the Data line is hooked to RA4 of the first PIC. Here's what I get...

Test Pattern:
It starts at the 2nd Pixel and then cycles colors start with Red, Green, Blue, Yellow, Cyan, Violet, Warm White?, Cool White. It changes the color one pixel at a time. When it reaches the end, it goes back to the second Pixel and start with the next color.

Sequence:
When I run a sequence, it is definitely reacting to it, but not as I expected. I created a sequence that I believed would light each pixel RGB. 1st Pixel Red, 2nd Pixel Green, 3rd Pixel Blue. But that's not what is happening. I end up with the first 4 pixels doing nothing and then all Pixels are either Red or Green. The good news is that they are clearly reacting to my Fades, so I know they are working.

Vixen Channel Tester:
When I start going down the Channels on the Channel Tester, nothing happens until I reach the 4th Channel. Which actually lights Pixel 26 Red, Ch 5 lights 26 Green and ch 6 lights 26 Blue. 7,8,9 light Pixel 27 in the same manner. Going down the Channels reveals that Pixel 25 is controlled by 100,101,102.

I see the above results no matter what mode I select in the Pluging on the GE Tab. "Inc...", "Dec...", or "Snowglobe..."

I'm sure I could figure it out, but I'm assuming there a simpler reason this is happening.

Any thoughts? Has anyone other than Don played around with his Plugin and Firmware?

I can't find the post - but someone was having a problem that sounded very simlar - somthing to do with the 99th pixel data not being used to control the pixel it was a reset or something - hopefully it will jog some ones memory who can remeber the problem or point you to the post.

budude
02-19-2012, 09:03 PM
Yes - this is the extra sync character that the Renard Plug-in tosses in at the 100th channel mark - your data parser needs to recognize the pattern and ignore this. This is done to allow the Renard to "catch up" to any minor differences in Tx and Rx clocks between PC and controller.

kychristmas
02-19-2012, 10:57 PM
I can't find the post - but someone was having a problem that sounded very simlar - somthing to do with the 99th pixel data not being used to control the pixel it was a reset or something - hopefully it will jog some ones memory who can remeber the problem or point you to the post.


Yes - this is the extra sync character that the Renard Plug-in tosses in at the 100th channel mark - your data parser needs to recognize the pattern and ignore this. This is done to allow the Renard to "catch up" to any minor differences in Tx and Rx clocks between PC and controller.

I'm aware of this issue with the standard renard Firmware, but since Don created this Custom, I assume he's taken care of that.

djulien
02-19-2012, 11:04 PM
somthing to do with the 99th pixel data not being used to control the pixel it was a reset or something - hopefully it will jog some ones memory who can remeber the problem or point you to the post.

You are correct - that is related to the Pad character used in the Renard protocol. This should not be an issue in this particular case, since the firmware and Vixen Plug-in should both be handling the Pad characters correctly.


I've been reading this forum and like the idea of having renard protocol running my GECE strings... I'm assuming I can't use my Renard SS16 controllers, but it sounds like my Ren48LSD could be re-purposed for this?

In theory, any Renard with a 16F688 in it can be used, as long as you re-flash all of the PICs in that controller (and on that COM port) with the customized Renard-RGB firmware, and then connect the ground and one of the bare I/O pins (no resistor or transistor in between) to the GECE string. I only tried it with one specific type of Renard controller, though, so there may be some Renard variants that might need extra steps.

EDIT: If you have a Renard with a different type of PIC in it, let me know which one and I'll see how easily the firmware could be adapted to that other PIC.


Test Pattern:
It starts at the 2nd Pixel and then cycles colors start with Red, Green, Blue, Yellow, Cyan, Violet, Warm White?, Cool White. It changes the color one pixel at a time. When it reaches the end, it goes back to the second Pixel and start with the next color.

That sounds like the correct test pattern. The Warm White is actually supposed to be Yellow (ie, Red + Green), but it's not a very intense yellow (at least it wasn't on my strings). The Cool White is the real White (ie, Red + Green + Blue at full).


Sequence:
When I run a sequence, it is definitely reacting to it, but not as I expected. I created a sequence that I believed would light each pixel RGB. 1st Pixel Red, 2nd Pixel Green, 3rd Pixel Blue. But that's not what is happening. I end up with the first 4 pixels doing nothing and then all Pixels are either Red or Green. The good news is that they are clearly reacting to my Fades, so I know they are working.

If all of the pixels are reacting to the same channel, that sounds like the GECE node addressing is messed up or a problem with the sofware. Although, it might also be possible to get this if you restarted the GECE string without restarting the Renard-RGB firmware (because they would assign addressed based on the next command they saw).

EDIT: maybe I misinterpreted. Does the sequence turn on one channe at a time, or do you mean all channels and one color?


Vixen Channel Tester:
When I start going down the Channels on the Channel Tester, nothing happens until I reach the 4th Channel. Which actually lights Pixel 26 Red, Ch 5 lights 26 Green and ch 6 lights 26 Blue. 7,8,9 light Pixel 27 in the same manner. Going down the Channels reveals that Pixel 25 is controlled by 100,101,102.

I see the above results no matter what mode I select in the Pluging on the GE Tab. "Inc...", "Dec...", or "Snowglobe..."


This sounds like the zig-zag order, and the fourth instead of 1st GECE string. Try RA3 to see if it changes to the third instead of 4th, or RA0 to see if it becomes the first one. Maybe I wrote down the wrong I/O pin order.

I had the zig-zag hard-coded for the show, and then un-hard coded that before posting. Maybe I left it on - I will go back and check.

EDIT: Actually, it doesn't sound exactly like the zig-zag order, so I might have messed up the channel orders during the cleanup. After node 26, 27, ... does it go all the way to node# 36 (assuming this was with a 36 ct string)? Which node turns on for channels 94 - 99, and 34 - 39?

Also, if you select All On, do they all turn white?

don

djulien
02-21-2012, 02:31 AM
This sounds like the zig-zag order, and the fourth instead of 1st GECE string. Try RA2 to see if it changes to the third instead of 4th, or RA0 to see if it becomes the first one. Maybe I wrote down the wrong I/O pin order.

I had the zig-zag hard-coded for the show, and then un-hard coded that before posting. Maybe I left it on - I will go back and check.

EDIT: Actually, it doesn't sound exactly like the zig-zag order, so I might have messed up the channel orders during the cleanup. After node 26, 27, ... does it go all the way to node# 36 (assuming this was with a 36 ct string)? Which node turns on for channels 94 - 99, and 34 - 39?

Whoops - not enough testing after "clean up". :blush2: I removed the hard-coding in one place but not the other. Sorry about that. I've updated the ZIP attachment in the first post of this thread and the external web site.

I've verified that the increasing and decreasing node address ordering is working okay now, at least for 50 ct strings. It assumes there are always 8 strings present, so assign 3 * 8 * #nodes per string to the plug-in. For example, with 36 ct use 864, for 50 ct use 1200, etc. The plug-in will divide the number of channels by 24 to get the GECE string length.

don

kychristmas
02-21-2012, 10:40 PM
I certainly do not mind Don. Thanks for checking into it and correcting it. I have a 7AM meeting and It is taking all I have not to go up to my office and hook the controllers up again. I'll let you know how it goes.

btw... When the sequence stops, will the strings go back to Demo mode?

Kelly


Whoops - not enough testing after "clean up". :blush2: I removed the hard-coding in one place but not the other. Sorry about that. I've updated the ZIP attachment in the first post of this thread and the external web site.

I've verified that the increasing and decreasing node address ordering is working okay now, at least for 50 ct strings. It assumes there are always 8 strings present, so assign 3 * 8 * #nodes per string to the plug-in. For example, with 36 ct use 864, for 50 ct use 1200, etc. The plug-in will divide the number of channels by 24 to get the GECE string length.

don

djulien
02-22-2012, 12:01 AM
I have a 7AM meeting and It is taking all I have not to go up to my office and hook the controllers up again. I'll let you know how it goes.

btw... When the sequence stops, will the strings go back to Demo mode?

No. There is a command that can be sent to the firmware to do that, but the plug-in does not use it. I should probably add it. You can, however, send the command using Putty or Windows Terminal Emulator (or any other serial program). I can provide more details if you want them.

I envisioned having an uploadable "background" sequence that would play between sequences or when Vixen was not active. I just didn't finish working through all the details on that yet, so I didn't expose it in the plug-in.

BTW, if the firmware or plug-in don't quite do what you want them to and it isn't too difficult I can try try to change it.

(I hope you did not read this until after your meeting :wink2: )

don

kychristmas
02-23-2012, 12:00 AM
Don,
I don't want to appear unappreciative or treat you like my personal developer, but I assume you shared this so other could benefit from your hard work. I downloaded the latest plugin. It definitely made a huge difference, but it still doesn't seem correct.

In Incremental mode, there is an issue where any pixel that is turned on is the same color as the last pixel colored. So if I try to turn the first Pixel Reand Second Pixel Blue and Third Pixel Green, all three pixels will be Green. If I then make the 36th Pixel Purple, I end up with 1st,2nd,3rd and 36th Pixels All Purple.

One other strange thing is that the first 9 channels are skipped. The first Pixel in the string starts at Channel 10. The Blue channel for the 36th Pixel is at Channel 18.

Thanks so much for working on this. I know it's very close.

Kelly

djulien
02-23-2012, 12:28 AM
but I assume you shared this so other could benefit from your hard work.

Definitely!


In Incremental mode, there is an issue where any pixel that is turned on is the same color as the last pixel colored. So if I try to turn the first Pixel Reand Second Pixel Blue and Third Pixel Green, all three pixels will be Green. If I then make the 36th Pixel Purple, I end up with 1st,2nd,3rd and 36th Pixels All Purple.

This is the palette size limitation. I only needed bi-tonal patterns for my sequences so the plug-in only sends 2 colors currently (off counts as one of the 2). I'll be removing that limitation some time in the next few months because I want to do fancier graphics for the next display season; the current version only does 2 colors at a time because I ran out of time for the 2011 display season.

EDIT: I could try to move that earlier if it's critical.


One other strange thing is that the first 9 channels are skipped. The first Pixel in the string starts at Channel 10. The Blue channel for the 36th Pixel is at Channel 18.

How many channels do you have assigned to the plug-in? Are you still using 36 ct strings?

don

kychristmas
02-23-2012, 07:19 PM
Definitely!

Thanks, I'm glad to hear it.



This is the palette size limitation. I only needed bi-tonal patterns for my sequences so the plug-in only sends 2 colors currently (off counts as one of the 2). I'll be removing that limitation some time in the next few months because I want to do fancier graphics for the next display season; the current version only does 2 colors at a time because I ran out of time for the 2011 display season.

EDIT: I could try to move that earlier if it's critical.


Not a problem. I would love to play with in realy Pixel mode, but my e680 will be here soon enough for. I would also be willing to help do some of the work if that helps. Is this a change needed for the Firmware, Plugin, or Both? Not so good at Firmware, but Plugin should not be a problem.




How many channels do you have assigned to the plug-in? Are you still using 36 ct strings?

don

I'm using 36 ct strings. I incorectly stated taht the blue channel for the last Pixel (36) was at 18, its actually 118. So things are shifted by 9. I have it setup for 864 Channels as per your formula :)

Again, thanks for the quick responses. I have enjoyed playing with the GE strings.

btw... I

djulien
02-23-2012, 10:18 PM
Is this a change needed for the Firmware, Plugin, or Both?

Both - there are some things I'd like to fix up in the firmware also, as well as some restructing to support some other changes I'd like to make to it (I'm going to try to merge my other stuff into one source file).


I'm using 36 ct strings. I incorectly stated taht the blue channel for the last Pixel (36) was at 18, its actually 118. So things are shifted by 9. I have it setup for 864 Channels as per your formula :)

I don't understand why that is happening. I tried setting 864 channels here and got the first 36 of my 50 ct nodes to light up correctly, and from what I've read the 50 ct nodes work the same as the 36 ct nodes, so I'll need some debug info. Could you set a trace ("data logging") file via the Setup window, then use the Vixen Test window to turn on/off channels 1 - 12 and 117 - 120 (one at a time) and then post or send that file? Or, if you prefer to just wait for the E680 that's okay.


btw... I

??

don

kychristmas
02-23-2012, 10:59 PM
Both - there are some things I'd like to fix up in the firmware also, as well as some restructing to support some other changes I'd like to make to it (I'm going to try to merge my other stuff into one source file).

I don't understand why that is happening. I tried setting 864 channels here and got the first 36 of my 50 ct nodes to light up correctly, and from what I've read the 50 ct nodes work the same as the 36 ct nodes, so I'll need some debug info. Could you set a trace ("data logging") file via the Setup window, then use the Vixen Test window to turn on/off channels 1 - 12 and 117 - 120 (one at a time) and then post or send that file? Or, if you prefer to just wait for the E680 that's okay.



??

don


Darn A.D.D, I'm not even certain what I started to write at the bottom of that message. I also can't remember what distracted me.

I'll try to send the Data File. Of course, it seems like you are encouraging me to wait for the 680. So I'm happy to leave it alone if you like as well. :) I'm enjoing messing with it. It's a been a quick little distraction from my crappy work life lately.

Thanks Again
Kelly

djulien
02-23-2012, 11:48 PM
Of course, it seems like you are encouraging me to wait for the 680. So I'm happy to leave it alone if you like as well. :) I'm enjoing messing with it. It's a been a quick little distraction from my crappy work life lately.

I wasn't intentionally encouraging it. I just didn't want you to feel like you had to spend more time on this if you had other things you wanted to do instead (especially if you have an E680 coming anyway). But, if it is still of interest, then I would prefer to track down what's happening and fix it so the debug file would be helpful.

Sorry to hear about work. Maybe give out your co-worker's numbers to Rachael to liven things up a bit? :)

don

jrock64
02-24-2012, 12:23 AM
I am hoping that Kelly gets the bugs worked out by the time I want to use it.:wink:

Really do not want to waste an E680 on 3 x 18 count pixel strands
GECE 18ct snowflakes

Joel

djulien
02-24-2012, 12:48 AM
I am hoping that Kelly gets the bugs worked out by the time I want to use it.:wink:

Do you have a specific timeframe for that?

Are the 18 ct like a 50 ct just shorter?

don

jrock64
02-24-2012, 02:20 AM
Assuming the same control chip.
Diffuser is a flat 6in snowflake.
http://www.menards.com/main/p-1691175.htm

August - September time frame.

It was one of those can't find what I want
so gust grab whatever they have left purchases.

If your controller works, I may have a place to incorporate them into the display.

JOel
.

kychristmas
02-24-2012, 09:14 AM
Assuming the same control chip.
Diffuser is a flat 6in snowflake.
http://www.menards.com/main/p-1691175.htm

August - September time frame.

It was one of those can't find what I want
so gust grab whatever they have left purchases.

If your controller works, I may have a place to incorporate them into the display.

JOel
.

I can tell you works for sure. Just with the limitations he described about the colors. It would still enabled you to control each pixels on/off state and one color. I haven't tried it yet, but I'm pretty sure if they are all on, you get two colors. Not ideal, but certainly a step beyond the Dumb RGB string control and definitely usable. I figure anything Don provides above what I have now will be a nice to have, but not necessary.

My next step is to see if I can get it to work with Ren Wireless. I'm guessing I can, but I'll know for sure this weekend. My plan is to use some of these strings to replace my Neighbors C9 Incans.

If you have a Controller Ren 64 or Ren24LV (and possibly others), I would give it a try. At least half of my trouble was self-inflicted. Since I have not had any Pixel control in my 4 years, it was just fun to play with. I have a e680 on the way that I will use to control them in my display.

djulien
02-24-2012, 11:00 AM
I haven't tried it yet, but I'm pretty sure if they are all on, you get two colors.

Yes.


My next step is to see if I can get it to work with Ren Wireless. I'm guessing I can, but I'll know for sure this weekend.

I'll be interested to hear how that turns out! There might possibly be some timing issues to address with the plug-in, depending on how the Ren-W handles bursts of data, etc. BTW, I forgot to ask - what baud rate did you try the Renard-RGB at?


At least half of my trouble was self-inflicted.

That's not bad. I think > 90% of mine is. :)


August - September time frame.

That should work out fine.


It was one of those can't find what I want
so gust grab whatever they have left purchases.

I've got lots of that also!

don