Node acting as a Null Pixel, String count out

[burto51]

Hello, this has me frustrated. Have 3 singing bulb elements all daisy chained, using a F16v3 and xLights version 23.20 . Each element has 143 nodes with a count of 429 showing on the string output page of the falcon.
When doing the face definitions, end up one node short, that is, nodes 1 to 142 light up, with "node 143" out. When tested using "tools test" in xlights, and selecting "models", "singing tree left", "nodes" node 1 does not light up, however when node 2 is selected the first node (ie node 1) lights. So it appears that node 1 is being bypassed or treated as a null pixel.
Checked for null pixel selected in layout, none; checked for null pixels in Controller String ports page again none shown.
This problem applies to all three elements.
Removed all elements from controller, and replaced on controller one at a time, problem still exists in each model; when tested individually.
This appears not to be a string fault as all nodes light when tested using the controller test "colourwash".

Have tried different ports on the Falcon with no luck. Checked channel count all correct. "Tools" Test Check sequence shows no warnings or errors.

At this stage have resorted to treating each bulb as a separate model individually wired and using three separate ports on the controller, and live with the last node being out on each model.
Any suggestions ?

 

[Skymaster]

What sort of pixels are you using for them? Are they GS8208 or any other type with a redundant data line? If so, perhaps the data of the first pixel is being fed into the backup data line rather than the primary data line.

 

[burto51]

Hi, Using ws2811, standard 3 wire RGB 12 volt.
Thanks

 

[MikeKrebs]

If you use the test to do one pixel at a time, can you narrow it down?

It sounds like a configuration problem to me. I would double and triple check that the upload is working correctly.

Another possibility is the model layout doesn't include the problem node. Check the singing faces config and make sure the pixel/channel is actually setup where you expect it.

 

[burto51]

Thanks Mike,
When I do the test using Tools, Test and select nodes for the model. If I click on node 1 it does not light, if I deselect node 1 and select node 2, a node lights up only problem is that it is the first node in the string which is not node 2, if I then continue to select each node in turn I end up with a string of nodes illuminated but only 1 to 142 instead of 1 to 143.

The model does include the problem node, as the base of the bulb has nodes 1-26, if i select in the face definition nodes 1-26 only 1 to 25 light, then when i select the outline 27 to say 68, 26 lights and i am short one node and so on till the end when node 143 doesn't light because xlights thinks it has used the 143.
I have checked the channel count in the visualiser and this matches with the channel count on the string port page on the controller. The number of pixels on the controller string port page match the number on the model.

If I select the model in the Falcon test mode and enable "colourwash" all nodes light.

I have tried a new string of pixels with the same result, so it seems to be a config problem or something. I was on zoom yesterday and a gentleman looked at my controller pages, layout etc but could not see any problems.

So I am stumped, thanks for taking the time if you have any other suggestions let me know.

Burto51 (Greg)

 

[MikeKrebs]

The colorwash test would indicate that the correct overall channels are defined for the prop.

Is this someone's face or did you make it? Did you download the model or create it? How did you wire it?

IIWM, I would create a new directory, put only the controller you have having problems with and the single model of one face. I would try both manual config and auto upload and see if I could get it to work.

Can you post a screen shot of the faces config? Seems like for sure node 1 is not where you think it is or is also somewhere else causing a conflict. If you export your faces model, someone might have time to import it and see if there is a problem. Glad you tried the zoom room. Too bad they couldn't figure it out.

 

[burto51]

Hi, The model is a Boscoyo Singing bulb downloaded in the layout page. followed wiring diagram with download.

Have also tried a string of 100 new nodes totally separate to the model, laying on the floor connected to same port as the bulb would have been, colourwash works fine, when I select tools test nodes, get the same result, nodes 1 does not light, node 2 selected and the first node in the string lights ie number flamin one.

Am considering creating new model and dumping all others and starting from scratch, similar to your suggestion which maybe a better option.
Thanks mate

 

[Skymaster]

Channel mismatch between the controller and xLights? Perhaps a upload of outputs is in order, if you haven't already?

 

[burto51]

Fixed the problem, rather the brains trust on zoom did, I had a Null Pixel (which I thought was a F amp) on the controller port due to the length of the data cable. Removed the culprit and all good. Thanks to all for your suggestions.

 

[Skymaster]

Ah ha! Well found. That'd certainly have some fun with the setup

 

[burto51]

This a photo of the culprit