Pixie4 First pixel blinks randomly when computer is off

[toddm1919]

I have a Pixie4 on a network with other controllers.  It is set to unit ID 88.  I have 50 pixels on  all four ports.  With the computer off, the first pixel on the string attached to the first port (88) will blink red randomly.  It may be 10 seconds it may be 30 seconds.  All other pixels on the that port and the others (89,8A,8B) remain off. It will continue in this fashion until I turn on the computer.  

When the computer is on, I have LOR in the system tray, but the schedule is not enabled. When the computer is on, the pixel will not blink...at least not for the 5 minutes that I was staring at it.  If I turn the computer off, the blinking returns.  

I have made sure that the First Pixel as Indicator box is unchecked in the config.  (I know that this is supposed to only last for a minute and be white, but I figured I would check it anyway.)

When I run sequences through the sequence editor, everything runs fine.

Any ideas on what would cause this?

Thanks,

Todd

[dibblejr]

19 minutes ago, toddm1919 said:

I have a Pixie4 on a network with other controllers.  It is set to unit ID 88.  I have 50 pixels on  all four ports.  With the computer off, the first pixel on the string attached to the first port (88) will blink red randomly.  It may be 10 seconds it may be 30 seconds.  All other pixels on the that port and the others (89,8A,8B) remain off. It will continue in this fashion until I turn on the computer.  

When the computer is on, I have LOR in the system tray, but the schedule is not enabled. When the computer is on, the pixel will not blink...at least not for the 5 minutes that I was staring at it.  If I turn the computer off, the blinking returns.  

I have made sure that the First Pixel as Indicator box is unchecked in the config.  (I know that this is supposed to only last for a minute and be white, but I figured I would check it anyway.)

When I run sequences through the sequence editor, everything runs fine.

Any ideas on what would cause this?

Thanks,

Todd

Id check for a loose wire or something else making contact  at the green connectors. With all of my pixies I have never had a light stay on, unless I set it up that way for testing.

JR

[TheDucks]

There is a setting in HU to have a Pixie blink the 1st led when first  powered on  and no net (so you don't have to open the case with the card).

Could this be what you see

[toddm1919]

4 minutes ago, TheDucks said:

There is a setting in HU to have a Pixie blink the 1st led when first  powered on  and no net (so you don't have to open the case with the card).

Could this be what you see

Already checked this.  The box is not checked in the HU.  Also, the blinking continues well after that option should stop..in excess of 15 minutes after turning off the computer and the timing between blinks is not consistent as it would be if it were the indicator.

[dibblejr]

Just now, toddm1919 said:

Already checked this.  The box is not checked in the HU.  Also, the blinking continues well after that option should stop..in excess of 15 minutes after turning off the computer and the timing between blinks is not consistent as it would be if it were the indicator.

Since you built it , look for that wire. You may want to ensure you have the wire going in to the green terminal lug in properly. 

JR

[toddm1919]

33 minutes ago, dibblejr said:

Since you built it , look for that wire. You may want to ensure you have the wire going in to the green terminal lug in properly. 

JR

JR,

Are you referring to the connection at Port 1 on the board? I will have to wait until the morning to check it.  

Any ideas as to why the computer being on or off would change the activity of that pixel?

Edited November 30, 2019 by toddm1919

[dibblejr]

10 hours ago, toddm1919 said:

JR,

Are you referring to the connection at Port 1 on the board? I will have to wait until the morning to check it.  

Any ideas as to why the computer being on or off would change the activity of that pixel?

Not being a programmer I recently discovered I do have some terminology flaws but I try to keep my answers simple for people like me, and I do not know all of the terminology.

But remember the computer is not controlling the pixel, the controller is controlling the pixels.

One loose wire or arch caused by a phantom wire can cause big issues. I had similar things happen to me caused by terminal lugs that were not righty tighty but that’s another thread years gone by.

Have you tried switching ports to test? As simple as just moving one of the other terminal lugs to that one to see if it still happens. You do not have to touch the prop.

If the above test shows that the changes string works properly , 

If it’s pixel nodes can you cut that pixel off to see if it carries over to the next?

JR

[TheDucks]

Right

The computer controls the Pixel Controller.

The controller controls the pixels.

The usual suspects:

Avoiding tight coupling of the CAT5 to the power cord for very many feet (no zip tie CAT5 to power cords for any distance.

Termination ON the end of longer network cable chains.  (helps reduce the above sensitivity

And as atated previously, Poor connections

[toddm1919]

It’s been raining all day, so the only thing I have was do was to switch the strings.  It did not follow the string.

I still ran my show tonight and noticed that there were other channels on the same network that came on when they were not supposed to.

This makes me think that it might be in the network cabling.

Ducks, what do you mean by Termination?  I just have the network cable plugged into the last controller.  Is there something else I could do?

[dibblejr]

Just now, toddm1919 said:

It’s been raining all day, so the only thing I have was do was to switch the strings.  It did not follow the string.

I still ran my show tonight and noticed that there were other channels on the same network that came on when they were not supposed to.

This makes me think that it might be in the network cabling.

Ducks, what do you mean by Termination?  I just have the network cable plugged into the last controller.  Is there something else I could do?

Sounds like you may have some residual sequence in your sequence or something. Even the dimmest fade can do that, but not for long.

Im still guessing a green terminal lug has something shorting it out.

What power supply and pixel voltage?

JR

[TheDucks]

15 minutes ago, toddm1919 said:

It’s been raining all day, so the only thing I have was do was to switch the strings.  It did not follow the string.

I still ran my show tonight and noticed that there were other channels on the same network that came on when they were not supposed to.

This makes me think that it might be in the network cabling.

Ducks, what do you mean by Termination?  I just have the network cable plugged into the last controller.  Is there something else I could do?

120 ohm resistor on the blue pair (4-5) in that last empty jack.   (use an old CAT5 cable with 1 broken plug )

RS 485 really should be terminated at the ends. The dongle does it at the source.  New (gen2 Pixies have a jumper you can add IF IT IS LAST)

I make mine with1/2W resistor ONLY because the lead diameter will crimp direct in a RJ45 plug. there is no power dissipation to speak of

Edited December 1, 2019 by TheDucks

[toddm1919]

 

1 hour ago, dibblejr said:

Sounds like you may have some residual sequence in your sequence or something. Even the dimmest fade can do that, but not for long.

Im still guessing a green terminal lug has something shorting it out.

What power supply and pixel voltage?

JR

I am using the same sequences from last year that ran fine, so there should not be any errant sequencing.  

I am using Holiday Coro 12 V pixels and power supply.  

I will check the green plug when I can get at it.  (It’s on our garage roof)

[toddm1919]

1 hour ago, TheDucks said:

120 ohm resistor on the blue pair (4-5) in that last empty jack.   (use an old CAT5 cable with 1 broken plug )

RS 485 really should be terminated at the ends. The dongle does it at the source.  New (gen2 Pixies have a jumper you can add IF IT IS LAST)

I make mine with1/2W resistor ONLY because the lead diameter will crimp direct in a RJ45 plug. there is no power dissipation to speak of

Good to know.  Thanks.

[TheDucks]

1 hour ago, toddm1919 said:

 

I am using the same sequences from last year that ran fine, so there should not be any errant sequencing.  

I am using Holiday Coro 12 V pixels and power supply.  

I will check the green plug when I can get at it.  (It’s on our garage roof)

I have 5 of those supplies in use (and a spare) Mine ar mounted on the Outside of the case. They get pretty warm when lights stay on white

https://drive.google.com/file/d/1OrEZiskjJj2fhwuzkadZ22k9mZ9LZU30/view?usp=sharing

[toddm1919]

Thanks JR and Ducks for your help so far..

I have an update on the symptoms....
I have three networks...Reg (COM3), Aux A(COM4), and Aux B(COM5). All symptoms are on Aux A(COM4) network.
I have 13 controllers on that network. 6 CCR, 2 LOR1602, 3 CMB-24D 1 Pixie16 and 1Pixie4

Our show went live 12/1.  For the most part things ran fine during the 3 - 20 minute shows.  However on one occasion 5 channels on one of the CMB-24D controllers turned on white when it was not supposed to.  On another occasion, that CMB-24D had one channel turned on white and the other had 4 channels turn on white.  

On one occasion one channel on one of the LOR1602 went to 100% when it should have been 50%.  On one occasion one port on the Pixie4 glowed orange when it should have been off. Finally on one occasion a different port on the Pixie4 seemed to get stuck on during a sequence.  All of these happened at different times during the show and different songs except the CMB-24D and the glowing port on the Pixie4 happened during the same song.  I also saw the random blinking pixel happen a few times during the shows.

During yesterday’s show this is what happened during the same song as the last clip

 

I have changed out the Pixie4 board as well as replacing the Cat5 cable to and from the Pixie4 controller in an effort to solve this problem.

I know it is not in sequencing since I have not modified the sequences since last year when they were working fine and I have run one of songs again through the sequence editor with different random channels turning on.

I am looking for suggestions on what to look at next.  Should I be chasing the CAT5 Cabling/connections or the power?

 

Edited December 3, 2019 by toddm1919

[dibblejr]

4 hours ago, toddm1919 said:

Thanks JR and Ducks for your help so far..

I have an update on the symptoms....
I have three networks...Reg (COM3), Aux A(COM4), and Aux B(COM5). All symptoms are on Aux A(COM4) network.
I have 13 controllers on that network. 6 CCR, 2 LOR1602, 3 CMB-24D 1 Pixie16 and 1Pixie4

Our show went live 12/1.  For the most part things ran fine during the 3 - 20 minute shows.  However on one occasion 5 channels on one of the CMB-24D controllers turned on white when it was not supposed to.  On another occasion, that CMB-24D had one channel turned on white and the other had 4 channels turn on white.  

On one occasion one channel on one of the LOR1602 went to 100% when it should have been 50%.  On one occasion one port on the Pixie4 glowed orange when it should have been off. Finally on one occasion a different port on the Pixie4 seemed to get stuck on during a sequence.  All of these happened at different times during the show and different songs except the CMB-24D and the glowing port on the Pixie4 happened during the same song.  I also saw the random blinking pixel happen a few times during the shows.

During yesterday’s show this is what happened during the same song as the last clip

 

I have changed out the Pixie4 board as well as replacing the Cat5 cable to and from the Pixie4 controller in an effort to solve this problem.

I know it is not in sequencing since I have not modified the sequences since last year when they were working fine and I have run one of songs again through the sequence editor with different random channels turning on.

I am looking for suggestions on what to look at next.  Should I be chasing the CAT5 Cabling/connections or the power?

 

That's a pixie16 to many for that  network. You are exceeding the recommended count.

I will tell you that I have experienced things happen in a working sequence one year to a few channel problems the next and I am not alone. I just found out it wasn't only me this past Saturday. You should look at your channels and separate that pixie from the others.

JR

[toddm1919]

51 minutes ago, dibblejr said:

That's a pixie16 to many for that  network. You are exceeding the recommended count.

I will tell you that I have experienced things happen in a working sequence one year to a few channel problems the next and I am not alone. I just found out it wasn't only me this past Saturday. You should look at your channels and separate that pixie from the others.

JR

What is the recommended count?  I thought I was OK...I do only use 8 of the 16 ports on the Pixie16 and I only use 50 pixels per port.

[TheDucks]

28 minutes ago, toddm1919 said:

What is the recommended count?  I thought I was OK...I do only use 8 of the 16 ports on the Pixie16 and I only use 50 pixels per port.

50*8*3=1200 

Think about how many colors (the 3) change at any time. If you change all 400 Nodes in one bang, that is a lot of data.

Come to think. if you change ANY node, you send those bits to ALL the Lights (I don't know how much was sent on the LOR network) in the string, because that is how SPI works

[toddm1919]

42 minutes ago, TheDucks said:

50*8*3=1200 

Think about how many colors (the 3) change at any time. If you change all 400 Nodes in one bang, that is a lot of data.

Come to think. if you change ANY node, you send those bits to ALL the Lights (I don't know how much was sent on the LOR network) in the string, because that is how SPI works

I get how to calculate the count, but I thought I found here or from LOR the magic number of channels that could be controlled on a 500K ELOR network.  I can’t seem to find that number anymore.

[k6ccc]

41 minutes ago, toddm1919 said:

I get how to calculate the count, but I thought I found here or from LOR the magic number of channels that could be controlled on a 500K ELOR network.  I can’t seem to find that number anymore.

The number varies depending on how involved the channel commands are.  500K enhanced should be somewhere around 1200 pixels or 3600 channels.