Pixie 16 Puzzle

[Denny]

During the Try for Free sale, I ordered 5 Pixie 16 Boards.  One board has a steady red status led showing it is communicating and power is going to both sides of the board, but no pixels on all 16 ports are controlled, they all remain off.  Tonight, a second Pixie 16 that has been running flawlessly for three nights will not light half the pixels on 8 strings.  I am controlling 100 pixel strings on each circuit.  Ports 9-16 are now only controlling the first 50 pixels rather than the full 100.  I am using a 30 amp power supply for each side of the board.  I am at a loss as to what is going on.  Is it possible that this might be a bad batch of boards?  Has anyone else had problems with the boards purchased during that sale?  What would cause only half the strings of pixels to quit working all on one side of the board.  

I still have two more new boards so will swap out this board tomorrow, but am quickly running out of Pixie 16s that were intended for my Christmas show.

[dibblejr]

I’m going to be honest and throw this out but it sounds like you may not have built them correctly and or configured them correctly.

Not probable of a bad batch. As you said “they worked “ but hey have started failing.

You say “30 amps psu” how many watts are they?

May be as easy as cables.

How many networks?

Network speed?

5 x 16x 100  - way exceeds 2400 pixels on an LOR Network at 500k.

3 of them also exceeds the limit.

If you changed the Logical Resolution in HU from 50 to anything put it back to 50.

Best bet is to go through the settings. Connect one and test. 
 

repeat any add to the daisy chain only after each is went through in HU.

You may find a problem any where in the chain as you go.

JR

 

 

[Denny]

JR

Thanks for the response.  I am using a N4-G4-MP3 Director to run the show.  I am using Network 1 to run only a 48x100 matrix (4800 pixels) divided among three Pixie 16 boards (16 circuits x 100 pixels per circuit = 1600 pixels per board).  Each of the three Pixie 16 boards have two 30 amp, 360 watt power supplies, one for each side of the board so each board has 720 watts total available.  The problem board controls the first 16 rows and the problem circuits are 1 through 8 (I said 9-16 in my original post).  The first 50 pixels in all 8 problem circuits work fine, however, pixels 51-100 no longer work.  I checked the power output of the power supply for those 8 circuits and it is putting out 11.8 volts. 

Network speed is 500K.

I was aware of the limit of 100 pixels per port, but over all these years, I never paid attention to or was not aware of the 2400 pixel limit per network.  Since I am running 4800 pixels on a single network, this is probably (hopefully) the problem.  I will split the matrix into three 1600 pixel networks and hope that corrects the problem.

Thanks for the advice.

Edited October 25, 2022 by Denny

[Mr. P]

I don't think you have to split the network up and make three sperate networks. First off a matrix seldom lights everything at one time and if you just increase your network speed to 1000K enhanced you should be okay.

[Jimehc]

Refresh the Firmware and verify the Settings on Just the "Problem" Board....  If that does not Fix, Swap the Problem Ports (1-8) with known working ports ((9-16) and test if the problem follows the Ports or Pixel Strings....

You have three possible common factors - Bad Ports 1-8, Bad Pixels Strings 1-8 or Bad Power Supply 1-8

And since Pixel 1-50 seem to work and Pixel 51-100 do not - End Connectors???  But that will be self evident in test above...

The Pixie16 should have Firmware Version 1.07 and allows for 170 Max Pixels per Port....

Edited October 25, 2022 by Jimehc

[Vince4xmas]

170 pixels per port on a pixie16 is 2,720 pixels. Run at 30% intensity and no issues.

 

[Denny]

Just an update on my problem.  As JR suggested, I split the matrix into 3 networks which seems to have solved the original problem.  However, a new (?) problem now on the same two boards.  None of the pixels on Ports 13-16 on one board and 12-16 on another do not light at all.  Went back and checked that I didn't change anything other than networks in the prop definitions, checked the fuses on the affected pors, voltage from the power supplies, pigtail connections all seem OK.  Tomorrow when it is daylight, I will do as also suggested and swap the affected strings to known working ports and see what happens.  Depending on what that shows, I may replace the boards and see what happens.  If I still have the problem, then it has to be something I have overlooked in my troubleshooting.  If the problem goes away, then it must be the board.  Will keep you updated on progress,  Thanks for all the suggestions, things I didn't think of.

Edited October 26, 2022 by Denny

[k6ccc]

My guess is a config issue...  And when you find it, the next sound you hear will be your hand slapping your forehead as you you yell "Duh!" (or "Doh!" for Simpsons fans).

 

[Denny]

Jim,

Checked the pixel strings and they all test good. HOWEVER, when I ran the Hardware Utility again, it found the Pixie 16 but identified the one set to 1A as 1F and the one set as 3A it identifies as 3E.  Checked and rechecked the DIP switches on the boards and they were all set correctly.  Turned them all to off and then reset them again, and the hardware utility still finds the boards as 1F and 3E rather than 1A and 3A.  I checked two more Pixie 16 boards that are functioning properly and the Hardware Utility finds and identifies them correctly.  Now, I am at a loss!  These are brand new boards, never used.  I have two more new boards still in the boxes so guess I will change out these boards and see what happens.  

[k6ccc]

Flip the DIP switches back and forth a few times.  Could be a poor contact or two.  Also try changing one switch and see if the HU recognizes the change.  Might help identify if there is a bad switch.  Does seem odd for new boards however...

 

[TheDucks]

33 minutes ago, Denny said:

Jim,

Checked the pixel strings and they all test good. HOWEVER, when I ran the Hardware Utility again, it found the Pixie 16 but identified the one set to 1A as 1F and the one set as 3A it identifies as 3E.  Checked and rechecked the DIP switches on the boards and they were all set correctly.  Turned them all to off and then reset them again, and the hardware utility still finds the boards as 1F and 3E rather than 1A and 3A.  I checked two more Pixie 16 boards that are functioning properly and the Hardware Utility finds and identifies them correctly.  Now, I am at a loss!  These are brand new boards, never used.  I have two more new boards still in the boxes so guess I will change out these boards and see what happens.  

Are you testing with only a Single board attached?

It sounds like you have an ID overlap.  Your expected ID just seems weird to me. 1A-29,

FWIW This is one of the reasons I start all my ID assignments on natural boundaries.

(I leave 0x for dumb controllers)

1x,2x,3x,4x,5x,6x,  a Single board range is in the same 2nd digit 10-1F.  I leave gaps between my Pixie 4's 20-23, 28-2F.  ( have already replaced 1 with a Pixie8. Unpug the old board,  swap in the new w/Same base ID)

[Denny]

I  set all the DIP switches to off and used the Hardware Utility to change the ID number.  Success - so far.  Now to test string control.

[dibblejr]

10 minutes ago, Denny said:

I  set all the DIP switches to off and used the Hardware Utility to change the ID number.  Success - so far.  Now to test string control.

This is what I have always recommended to everyone. Eliminates user errors and to continue to open up an enclosure if needed to set unit IDs.

JR

[Denny]

Just an update on the matrix problem.  Spent most of the day communicating with the LOR Help Desk, taking screen shots of the hardware utility and trying different things they suggested.  Nothing worked, so, this evening, I changed out the controller with one configured just like the problem controller.  Everything works!  It must have been a defective Pixie 16; see what happens after a few nights running.