San Device E682 issue

[jerrymac]

I have the E-681 updraded to E-682 running 16 pixel strings of 20 pixels each over 2 universes. I have two problems:

No color mixing (red, blue & green only) I am using, these WS2811 pixel strings.

http://www.aliexpres.../469177597.html

Other issue is I am not getting any data over sockets 2 or 4 (same as cluster numbers). I can move the same string from one socket to the next and no data on 2 or 4 and data on 1 and 3. I have setup for 2 universes (101 on socket 1 and 2 w/ the proper dmx numbers for the two sockets and universe 102 for sockets 3 & 4 with same dmx address except the universe number.). address 1 through 240 working (except color issue) and 241 - 424 0. I have set for 1804 pixels and assigned 4 sockets of 4 stings with 20 pixels per string.

seems I can't attach over 11k file (this is poor).

Clusrter String chip pixels grp str LEN RGB dmx ADDRESS range (other is 0) Referesh Gamma

1 4 1804 (f) 020 001 0020 rgb s1-001 thru 240 0597

2 4 1804 (f) 020 001 0020 rgb s2-241 thru 480 0597

3 4 1804 (f) 020 001 0020 rgb s1-001 thru 240 0597

4 4 1804 (f) 020 001 0020 rgb s2-241 thru 480 0597

Socket 1 Socket 2 Socket 3 Sockety 4 Socket 5 Socket 6

Univese Number 101 101 102 102 103 104

pakets rcvd 2422 0 2422 0 0 0

Sequence errors 13 0 11 0 0 0

Invalid Pakets 0 0 0 0 0 0

firmware 2.049

I have the pixels configered in LOR

DMX Universe 101 Circuit 1 through 480 (3 cirucits in order for each pixel of the first 8 strings and Universe 102 circuit 1 through 480 for the second 8 stings)

Any ideas?

Edited December 11, 2012 by jerrymac

[Tim Fischer]

For your second problem - have you tried running a test pattern via the device (e.g. TE 1) to see if the pixels light up? If so, you have a config problem on the LOR side. If not, something is wrong in the Sandevices config.

For your first problem, I have no idea. You're saying if you turn on Red, it works, if you turn on Green, it works, but if you turn on both you get.... ???

-Tim

[Crazydave]

Step 1.) Run a Test 4 to see if the pixels are working correctly as configured on the device.

(they should flash in sequence, there should be no simultaneous flashing on any socket

Step 2.) make a test sequence that chases sequentially through each of the channels.

then report back

[jerrymac]

Since universe's 2 and 4 are not responding full test 4 I can not watch through a coplete cycle.. It does cycle trough red (bight) pixel by pixel then to dim red.

In sequencing Adding a second color (i.e. red and green ) results in the original color remaining in this case rather than yellow only red is lit. Only seems to respond to one of the three primary colors red , blue and green, mixiing the three colors results in the original color ramaining until changed to one of the 3 primary colors.??????

Edited December 11, 2012 by jerrymac

[Crazydave]

Lets look at

Clusrter String chip pixels grp str LEN RGB dmx ADDRESS range (other is 0) Referesh Gamma

1 4 1804 (f) 020 001 0020 rgb s1-001 thru 240 0597

2 4 1804 (f) 020 001 0020 rgb s2-241 thru 480 0597 <---- why are you switching universes here?

3 4 1804 (f) 020 001 0020 rgb s1-001 thru 240 0597 <---- Is this supposed to be universe 2?

4 4 1804 (f) 020 001 0020 rgb s2-241 thru 480 0597

On the Software side....

How are your channels set up?

Edited December 11, 2012 by Crazydave

[Crazydave]

Agreed Jeff beat me to it: (i had some work to do)

It should look like this:

Cluster String chip pixels grp str LEN RGB dmx ADDRESS range (other is 0) Referesh Gamma

1 4 1804 (f) 020 001 0020 rgb s1-001 thru 240 0597

2 4 1804 (f) 020 001 0020 rgb s1-241 thru 480 0597

3 4 1804 (f) 020 001 0020 rgb s2-001 thru 240 0597

4 4 1804 (f) 020 001 0020 rgb s2-241 thru 480 0597

[jerrymac]

Thanks Jeff and CrazyDave. Between sockets, clusters and Euro jacks with a socket It gets a bit confusing.

I'll reconfigure and re try., would this also affect the color issues?

[Crazydave]

maybe. get everything in order first.. then once it's all connected and configured right... then the real troubleshooting can begin.

[n_gifford]

As far as issues with the colors, how do you have the channels mapped in the LOR software?

[WilliamS]

To fix the colors heres an easy trick.

Run test 01 write down the color that the pixels turn should be red but likely isnt

Run test 02 write down the color that the pixels turn should be green but likely isnt

Run test 03 write down the color that the pixels turn should be blue but likely isnt

This will be the proper RGB order of your strip. From here reprogram the appropriate clusters to the proper color order. So for example if test 01 came back blue, test 02 came back red, test 03 came back green your RGB strip is actually BRG order. So your command to fix them on say cluster #1 would be

RGB 1 5 <--- then enter

The RGB command defines the order in which the 3 color channels are assigned to the

pixels in a given cluster. Most pixels use the same sequence that the name RGB would

imply: RED first, followed by GREEN, then BLUE. Some pixels however assign their color

channels in a different order. BGR is fairly common. To cover all possibilities, the

RGB command lets you define the RGB pixel order of a cluster’s strings to be any of the 6

possible combinations, using a number from 0 to 5. The numeric codes, and the sequence

they assign, is as follows: 0=RGB, 1=RBG, 2=GRB, 3=GBR, 4=BRG, and 5=BGR.

RGB, n, a Defines the pixel color order for strings in group n

RGB 1,5 would define cluster #1 pixels to be BGR.

As far as issues with the colors, how do you have the channels mapped in the LOR software?

Dont use LOR to fix your colors, use the E681-682's software to fix it.

[jerrymac]

it is set for RGB, the test sequences proplery light red or green and blue (bright chase over dim color). I have confirmed LOR by using the same lines copied to some of my dumb pixels (DMX) which work with the colors expected.

I did test changeing to GRB and the colors still did not mix and displayed the wrong primary colors as expected.

The problem that the pixels will not mix these three colors (or any color using combinations of colors (orange, pink,yellow etc) and ONLY respond to the three primary (additive colors). when changing to any color that is a mix the pixels remain at the prior color until changed to R / G / B ?

has anybody purchased these pixels

promotion!!! WS2811 IP68 led pixel module,256scale gray,IP68;3wire(red/blue/black)DC5V input;50pcs a string

I read the review which were all great. I am using a computer power supply as a power source (actually two for the E681).

[WilliamS]

The link didnt work? I have 18 of these on my house in total not including my matrix

http://www.aliexpress.com/store/product/5m-led-digital-strip-DC12V-input-WS2811IC-256-scale-10pcs-IC-and-30pcs-5050-SMD-RGB/701799_568458133.html

The were all in # BRG format. On the test they were all wrong, then changed the clusters to #4 RGB and fixed.

Are you remembering to save after you make changes?

[Crazydave]

If it's testing properly for Red Green and Blue when using the E681 Test functions... then the strip is working correctly, and maybe something screwy in LOR set up.

Can you expand a few of your channels and show the settings for one of the pixels?

Screen shots?

[n_gifford]

Dont use LOR to fix your colors, use the E681-682's software to fix it.

I never said anything about using LOR to CHANGE the colors. The OP said that they had no color mixing, and were only getting red, green and blue. I'm just asking how the channels are mapped, to try and troubleshoot from the beginning.

[n_gifford]

it is set for RGB, the test sequences proplery light red or green and blue (bright chase over dim color). I have confirmed LOR by using the same lines copied to some of my dumb pixels (DMX) which work with the colors expected.

I did test changeing to GRB and the colors still did not mix and displayed the wrong primary colors as expected.

The problem that the pixels will not mix these three colors (or any color using combinations of colors (orange, pink,yellow etc) and ONLY respond to the three primary (additive colors). when changing to any color that is a mix the pixels remain at the prior color until changed to R / G / B ?

has anybody purchased these pixels

promotion!!! WS2811 IP68 led pixel module,256scale gray,IP68;3wire(red/blue/black)DC5V input;50pcs a string

I read the review which were all great. I am using a computer power supply as a power source (actually two for the E681).

I have 1,000 of these same pixels in this year's show and they've worked great so far! Can you upload some screen shots, to show the channel configuration in LOR?

[jerrymac]

!!! YEA !!! finally got it. The problem(s) were the confusion between clusters and sockets and euro plug sockets (still don't understand and could REALLY use a more english explination). and (color issue) was had data and ground wires for the pixels backwards. (I soldered waterproof connetors to all of the pixel strings and plugged in all of the data wires to the ground (and visa versa) to the euro plugs.

After checking the jumpers (thanks jeff which were correct) and resetting the confifturation page on the 681-6 (thanks to jeff and crazy dave ) and stil the same issue with colors i took out a yet unused string of pixels ( i was testing with a single string rather than all 16 strings on the tree). When i plugged directly into the 681-6 it worked fine on all sockets. Started chasing backwords until i noticed the cross over in data -vs- grouind.

A GREAT BIG THANKS . It is boards like this that get us together in a way that only a few years ago could not have happened (except for those of us who still remember the old bulletin board days (like Comp-u-Serve0 and long distance 300 baud modem communications).

Again THANKS !!!!!!!!!!!!

[jerrymac]

!!! YEA !!! finally got it. ............ and (color issue) was had data and ground wires for the pixels backwards. (I soldered waterproof connetors to all of the pixel strings and plugged in all of the data wires to the ground (and visa versa) to the euro plugs.

...........

Type O I had grnd from the pixel strings going to the clock conection on the board . these stings are 3 wire type and no clock wire. Apparentlly the lights were getting some grounding from the clock circuit on the board which allowed them to light up but not correctly.

(now reverse 16 pugs and all should be good again).

[k6ccc]

Jerry, I sent you a PM with my cell number. I should be able to help you out on the clusters vs sockets issue.

[jim st john]

!!! YEA !!! finally got it. The problem(s) were the confusion between clusters and sockets and euro plug sockets (still don't understand and could REALLY use a more english explination).

A bit belated, and definitely long-winded, but possibly of help to others.  The term 'socket' as used in the E682 setup can be confusing because we think of a 'socket' as a physical electrical connector.  In this case 'socket' is a networking term that refers to a "virtual" connection, an imaginary wire that carries one universe of DMX data from our show PC to the E682.  The E682 supports 6 sockets, or 6 DMX universes.  The first step in configuring the board is to determine which universes your board needs to respond to.  Typically, but not always, these will be sequential universe numbers, and typically, but not always, the first controller will start with universe number 1.  Once you have identified which universes you need, you use the UNiverse command to map those 6 universe numbers to the E682s sockets S1 thru S6.

 

Let me give an example of configuring 2 E682s for a total of 32 50-pixel strings, beginning at universe 1 channel 1, with all pixel addressses being assigned contiguously (no gaps).

 

Since we start with universe 1, we will map the first 6 DMX universes to the first controller.  This is actually the default configuration that ships with the controller, so we don't have to change anything, sockets 1-6 will already be mapped to universes 1-6.  What this is telling the controller is that those 6 virtual DMX wires between our PC and the E682 are bringing in universes number 1 thru 6 to this E682.  If we did have to do this manually we would type 6 UNiverse commands: UN 1 1, UN 2 2, UN 3 3, UN 4 4, UN 5 5, UN 6 6 (and of couse we would end with a SAVE 0!)

 

Now the 2nd step of configuring an E682s addresses is to map the individual clusters of pixel strings to those 6 sockets (and by doing that we are mapping them to the DMX universes that are assigned to those sockets).  Again, since this is board number one,and our configuration matches the factory default, we don't have to do anything.  Cluster 1 starts with socket 1 channel 1 and runs thru socket 2 channe 90.  This is a total of 600 DMX channels (510 in the first universe <remember 511 and 512 are NEVER used> and 90 in the 2nd, total 600).  600 channels = 4 strings * 50 pixles * 3 channels per pixel.

 

Cluster 2 begins at 2-91 and runs thru 3-180.  Cluster 3 begins at 3-181 and runs thru 4-270, and cluster 4 begins at 4-271 and runs thru 5-360.

 

So, even though we've assigned universes 1-6 to the 6 DMX sockets on this board, as it turns out we're only using the first 5. 

 

Now, let's continue with our remaining pixels on the 2nd E682, and this is where it gets interesting.  First we need to decide which 6 universes to map to our 6 sockets on the 2nd E682.  Remember, we only used a portion of universe 5 on the first controller, and we want all of our addresses to run with no gaps, so the 2nd controller will begin by using the last part of universe 5.  That means controller #2 will be mapped to DMX universes 5 thru 10.  To do that we use the UNiverse command:

UN 1 5, UN 2 6, UN 3 7, UN 4 8, UN 5 9, UN 6 10.  (I have typed that as 1 long line whereas when actually entering those commands each would be on its own line.)

 

This would be a good time to point out that YOU SHOULD NEVER DUPLICATE A UNIVERSE NUMBER when mapping sockets to universes.  Start out with the specific universe numbers that you need, and if you don't need all 6, then just map the unused ones to the next universe numbers in order.

 

OK, now we will set the DMX start addresses for the clusters on the 2nd E682.  The first board ended with cluster 4 using up thru socket 5 (universe 5) channel 360.  So, the 2nd E682 should start with universe 5 channel 361.  Remember we enter the DMX addresses as socket numbers 1-6, not universe numbers, so we will use the command: DMX 1 1 361 to set the first dmx address of the 1st cluster to Socket 1 channel 361.  Since socket 1 on this board is mapped to universe 5 this will give us the result we want, a start address of universe 5, channel 361.

 

Cluster 1 will run from 1-361 to 2-450.  So we use the command DMX 2 2 451 to assign socket 2 (univ 6) chan 451 thru socket 4 (univ 8) chan 30, to the 2nd cluster.  Notice here that this cluster actually uses 3 universes, parts of universes 6 and 8, and all of universe 7.

 

DMX 3 4 31 sets cluster 3 to use 4-31 thru 5-120, and DMX 4 5 121 sets cluster 4 to use socket 5 chan 121 thru socket 6 channel 210.  (If you notice, it's easy to know what address to use in the DMX command because we just go 1 more than the ending address of the previous cluster).  And please don't forget to SAVE 0 when done...

 

As a check, board #2 uses chans 361 thru 510 of universe 5 (150 channels), all of universes 6 thru 9 (2040 channels) and 210 channels of universe 10, for a total of 150+2040+210=2400 channels.

 

And the 2 boards together use a total of 4800 channels (all of universes 1-9 = 9*510=4590 plus 210 of universe 10, total 4800).  4800 channels = 1600 pixels, so everything seems right.

[Shubb]

Can you repeat that please?

(wink, wink)