PDA

View Full Version : Hopefully This Is Easier to Fix



covewi
07-18-2017, 06:22 PM
Again, thanks for all the networking help to get my PI up and running wirelessly.

I hooked up the 5X6 P10 panel and got this for a Butterfly effect.
36877

And this for a marquee
36878

Both are really off. This is how I hooked up my panels for a 5X6 setup - I added another 2 rows the same way.
36879

And then there is this setup - which is correct?
36880

I see a totally different "arrow" setup.

I got both of the setup images from the forum matrix section.
The xLights matrix is set up like this.

36881

Also, nothing is happening on the top two rows - no power or not video input i bet. I switched a ribbon cable around and that didn't help.
I don't think its the power distort board because some of the panels that are working get power from the same board.

Did some wire switching for diagnostics - I took a ribbon cable from output #3 of Ron's Pi Matrix adapter (which connects to the top 2 rows that don't work) and hooked it to the 2 middle rows (that do work)
and there was no signal output.

So I am assuming that output #3 is not working - Does anyone know which of the IC's controls the output to #3?

But that wouldn't explain the mix up in the panel display.

MrSmoofy
07-18-2017, 07:05 PM
Not sure if I can help I'm just learning this myself but I would take pictures of the back of how you have the ribbon cables connected and the arrows of the panels. Next is in FPP that page is as if you are looking at it from the front so it's reversed when looking at the panels from the back.

covewi
07-18-2017, 08:00 PM
Here's the photos of the rear.
36884
36885
36886
36887

I switched around all of the IC's on the output board and this is what I got - entire extra row and one panel of the last row worked - could the way I have th panels oriented - front verses left cause something like this or is it an IC problems?

mrGrumpy
07-18-2017, 08:04 PM
How many panels do you have, and what is their arrangement?
Also, what sequencer are you using?

covewi
07-18-2017, 08:07 PM
5 panels wide X 6 panels high. Just running a "butterfly effect in xzlights.

This is how its set up in FPP.

36888

Mine be like this?
36889
36890

mrGrumpy
07-18-2017, 08:44 PM
If your assembly is like the picture except 5W6H, then your arrows are wrong.

Here is 5X6 for that picture and xlights.

36892

covewi
07-18-2017, 10:53 PM
So I would have to take apart the panels and re-orient them so the arrow on the panels match the arrows in your drawing - correct?
And just for clarification - am I looking at the "back" of the panels in your drawing?
So the top right corner (in the drawing) is actually the top left from the front?

36896

As always - thanks

mrGrumpy
07-18-2017, 11:00 PM
Send me a picture of your panel back. I'll send you the FPP layout

mrGrumpy
07-18-2017, 11:40 PM
5 panels wide X 6 panels high. Just running a "butterfly effect in xzlights.

This is how its set up in FPP.

36888

Mine be like this?
36889
36890

FYI All three of these are different.

mrGrumpy
07-19-2017, 01:51 AM
So I would have to take apart the panels and re-orient them so the arrow on the panels match the arrows in your drawing - correct?
And just for clarification - am I looking at the "back" of the panels in your drawing?
So the top right corner (in the drawing) is actually the top left from the front?

36896

As always - thanks

Top left does not refer to cabling.....it tells FPP how the data will be received, then FPP will send to P10's based on output number, panel number, and arrows

covewi
07-19-2017, 10:37 AM
36899

mrGrumpy
07-19-2017, 11:35 AM
Use the below layout.

FPP Arrow page is based as you looking at the panel from the front......your layout was looking from the back.
36900

covewi
07-19-2017, 08:31 PM
Well that fixed everything - but now what the heck is going on.
i checked all the ribbon cables and everything seems to be connected properly.

The bottom half of the 3rd row of panels and the top half of the 4th row of panels do not display GREEN.
I find it impossible that all those panels don't work in that manner and I placed them randomly to all line up and not work.

Output from Ron's Pi board? Setup in xLights or FPP? Both of the non working rows are connected to the Ron's Pi output #2.

36902
36903 text misaligned
36904
36905 all reds and blues work correctly

mrGrumpy
07-19-2017, 09:21 PM
I think it is a cable issue.

Try swapping #1 with #2 feed cable
Also, try swapping #1 row or #4 row cables with #2 & 3, one at a time

That should show you what cable is a issue

covewi
07-19-2017, 11:00 PM
I switched input cables 1 to 2 and 2 to 3 - one at a time - no change.
I then switched all the cables (between panels) on working rows and non working rows - no change.
I then hooked a single new panel directly to output #2 on the output board - same problem.
I then switched around randomly all 4 IC's on the output board - same problem.

Looks like it has to be the board itself - bad solder joint, loose joint.
Does this sound correct?

Thanks

Ruppro
07-19-2017, 11:32 PM
Yea, I would check your solder joints, reflow them even, post some high resolution close up pic of the board, front and back.

mrGrumpy
07-20-2017, 02:16 AM
I switched input cables 1 to 2 and 2 to 3 - one at a time - no change.
I then switched all the cables (between panels) on working rows and non working rows - no change.
I then hooked a single new panel directly to output #2 on the output board - same problem.
I then switched around randomly all 4 IC's on the output board - same problem.

Looks like it has to be the board itself - bad solder joint, loose joint.
Does this sound correct?

Thanks
If switching feed cables doesn't change anything, then Ron's board is good....you have proven that
You have changed around cables.....so.

I think you may have a bad panel i think it is panel output 2 panel 1. (2-1)
Try using row 1 and row 3 feed cables to . . . Move #2 cable from 1-2 to panel 2-2
2-1 out to 1-2 in and 2-1 out to 2-3 in
You are electrically moving 2-1 to 1-2 and 1-2 to 2-1
Keep moving start down the line til you fine a pad panel.

covewi
07-20-2017, 09:23 AM
I will start the process tonight.
However, I did take a working panel from another row and attached it to output #2 (replaced the row 2 number 1 panel) and that single panel produced the same problem.

covewi
07-20-2017, 05:45 PM
mrGrumpy - yesterday when I switched the #1 and #2 cables - I actually switched the ribbon cables themselves to see if they were the problem - and there was no change.
Today I ran output #1 to the 3rd and 4th row of panels and output #2 to the 1st and 2nd row of panels - the problem still exists.
Expect the dead green bars are in the first and second row of panels.

36922

So it has to be the output #2 - agreed? I will be checking my soldier joints or for shorts.

mrGrumpy
07-20-2017, 06:44 PM
mrGrumpy - yesterday when I switched the #1 and #2 cables - I actually switched the ribbon cables themselves to see if they were the problem - and there was no change.
Today I ran output #1 to the 3rd and 4th row of panels and output #2 to the 1st and 2nd row of panels - the problem still exists.
Expect the dead green bars are in the first and second row of panels.

36922

So it has to be the output #2 - agreed? I will be checking my soldier joints or for shorts.

Once you moved the cable of output #2 to panel row 1....and found the issue moved to row 1......did you try swapping the cable from output #2 to Panel Row 1.....keeping the same connection from output 2 to panel row 1. Remove cable from output 1 and connect it from output 2 to panel row 1.

If no, try swapping. to see if Pi to P10 cables are good.

If yes, swapping cable keeps the issue at row 1, and you are using FPP test, then yes, I agree, we are at the Pi add on board.
- - - We have proven all panels and cables are good.

If you want another set of eyes, post a good picture of both sides or your Ron's board.

covewi
07-20-2017, 09:27 PM
Swapped the cables - the problem still exists. Looks like its the board.
Could a faulty board cause the flickering you see in this movie? Why aren't the whites whites? Faulty panels (please don't tell me that)?

This has been an unbelievable journey.
36923

36924

https://www.youtube.com/watch?v=C5dFoiXwHuY&feature=youtu.be
Yes thats a craps table - I rebuilt a 14 footer around 4 years ago.

mrGrumpy
07-20-2017, 10:46 PM
Pictures aren't the best, but I did mark a couple I think need to be looked at....Arrows especially.

Also, your link to video does not work.

36925

tucsonlights
07-20-2017, 11:28 PM
I would reboot the board and restart the fpp. I had the same issue 3 years ago and that was the problem.

mrGrumpy
07-20-2017, 11:37 PM
I would reboot the board and restart the fpp. I had the same issue 3 years ago and that was the problem.

Assume this was done when he was changing cables.....if not, good idea covewi

covewi
07-20-2017, 11:47 PM
Try this link.

https://www.youtube.com/watch?v=C5dFoiXwHuY

Could that be that the panels are too bright?

Thanks will recheck the soldier joints.

tucsonlights
07-20-2017, 11:59 PM
Could be power issue. Unplug half your panels and turn on your pi and retest to see if panels turn white

mrGrumpy
07-21-2017, 12:07 AM
Try this link.

https://www.youtube.com/watch?v=C5dFoiXwHuY

Could that be that the panels are too bright?

Thanks will recheck the soldier joints.

Maybe getting the other issue fixed will resolve this one too. Maybe lacking green
Or rev 1.9 will allow brightness setting. Try 50%

FLASHING: Change your pixel count to match your available.
30 panels X 1536 (521 X3) = 46,080

36928

covewi
07-21-2017, 08:20 PM
It appears that the brightness was the problem. On testing patterns I can turn it down. Guess I will have to remember to turn down the brightness in xLights before saving a sequence.

Ron P responded to my email with suggestions for solder bridges that might be causing the lack of green. I checked and ( I guess cleared) what I could but some of the tracings are under sockets.

Well as you can see the green now shows up.

Otherwise check out the butterfly effect. I put text on all three sections of panels that represent the three outputs from the board - the text is messed up.

Suggestions?
3692936930
36931


Misaligned pixel rows - I believe its the same rows that were not green before.
36932
36933

covewi
07-21-2017, 08:23 PM
Additional photos

36934

The white below now has one weird green stripe - which is half of a panel.

36935

The only time the two bars show up i when this test pattern is run.

36938

Would this be the correct setup for my matrix of 5 wide by 6 high.
What else could be causing the misalignment?

36939

mrGrumpy
07-21-2017, 09:20 PM
I still think you have soldering issues.
Go to the dollar tree or 99 cent store and buy a pair of +3.00 reading glasses. Use to examine your board for cold solder joints or solder bridges.

Suggest you upgrade back to V1.9. You can use the LED Screen to adjust brightness.....beast to adjust at FPP

Ruppro
07-21-2017, 09:34 PM
I still think you have soldering issues.
Go to the dollar tree or 99 cent store and buy a pair of +3.00 reading glasses. Use to examine your board for cold solder joints or solder bridges.

Suggest you upgrade back to V1.9. You can use the LED Screen to adjust brightness.....beast to adjust at FPP

+1 on the Soldering issues, take your iron (hot) and touch each joint just long enough for the solder to melt and flow again. Add a little solder if necessary. (top pin on the 3rd row from the right). Make sure you remove the ICs first

covewi
07-21-2017, 09:41 PM
mrGrumpy - how old do you think these eyes are? ( ha ha)
I am using a lighted magnifier. But will check everything again. I guess since there was improvement the solder issue needs to be explored again.

I thought ver 1.9 doesn't support wifi and I need that.

Ruppro - thanks for pin pointing a possible problem area - I am on it.

Ruppro
07-21-2017, 09:47 PM
I am running 1.9 and I have wifi too, believe mrGrumpy said it "Did Support"

Also mrGrumpy had it pointed out in his response earlier

mrGrumpy
07-21-2017, 09:57 PM
mrGrumpy - how old do you think these eyes are? ( ha ha)
I am using a lighted magnifier. But will check everything again. I guess since there was improvement the solder issue needs to be explored again.

I thought ver 1.9 doesn't support wifi and I need that.



Ruppro - thanks for pin pointing a possible problem area - I am on it.

Yes, ver 1.9 is good....you read it wrong.
And old or young eyes, those +3's are super, and beat most magnifiers....just need some good light.
Although I can solder without, I still use them for the best view possible while soldering.

covewi
07-23-2017, 09:27 AM
Did all solder points and think it helped but problems still exist. Ron P gave a few suggestions and I will try again.

Notice in the far left edge of the photos - you can see interference at the start of the panel that is connected to each of three outputs.
And in the full matrix photo notice the "rows" of interference.

36952
36953
36954

mrGrumpy
07-23-2017, 11:22 AM
Did all solder points and think it helped but problems still exist. Ron P gave a few suggestions and I will try again.

Notice in the far left edge of the photos - you can see interference at the start of the panel that is connected to each of three outputs.
And in the full matrix photo notice the "rows" of

Follow Ron's directions since you have his ear. The soldering redo seems to have fix a lot. Maybe another cold joint exists. . . I also feel your input cables may be longer than needed.

covewi
07-24-2017, 10:11 PM
I hate to give up but I am near the end of what I can check, recheck and triple recheck.
Every solder joint has been gone over and I tried shorter output cables without any difference.
A white screen
36972
A RGB screen - missing green
36973
A green screen
36974
Text
36975
The interference on the left side of the photo is on the last panel (the 10th panel) of each two rows.
36977


Is it a chance that I just plain messed something up on the board, a faulty board?
Anything else I can check?

Does anyone have a working Ron's Pi Matrix Adapter that they would be willing to sell?

covewi

mrGrumpy
07-24-2017, 10:51 PM
Ok covewi, I have a spare I can send you. If it works, you reimburse for shipping, and order me a new kit from Renard-Shop. If it doesn't help; you send it back, and reimburse me for shipping to you.

Are you powering at the matrix board or Pi. If you are not, try powering at only Pi

mrGrumpy
07-25-2017, 08:51 PM
Covewi....how about your voltage, boyelroy11; I learned today that they (P10's) can have ghosting with the voltage too high.

http://doityourselfchristmas.com/forums/showthread.php?47322-Ron-s-Pi-Matrix-Board&p=467486#post467486

covewi
07-26-2017, 11:30 AM
Yes I turned down the brightness in the FPP test display section - which does not correct the problems - but will be updated FPP to V1.9 tonight.

This is getting really weird - this photo is of a green test - but remember I have no green output in other tests and now theres some sort of interference.
36981
This is a white fill test - but green shows up.
36982
Heres an RGB sequence - with no green.
36983

I am wondering if there is a short under one of the IC or cable sockets - don't know how to check that though.

mrGrumpy
07-26-2017, 11:59 AM
Yes I turned down the brightness in the FPP test display section - which does not correct the problems - but will be updated FPP to V1.9 tonight.

This is getting really weird - this photo is of a green test - but remember I have no green output in other tests and now theres some sort of interference.
36981
This is a white fill test - but green shows up.
36982
Heres an RGB sequence - with no green.
36983

I am wondering if there is a short under one of the IC or cable sockets - don't know how to check that though.

Maybe your solder bridge caused over voltage on IC(s) Or a hidden (under socket) bridge still exists.
Just for info; look at these two posts where just changing the output from port 1 to port 2 fixed this issue. Here we knew the IC's got overheated, but only port 1 seem to have been damaged.
Since IC cost & freight alone is more than 1/2 a new board, replacing your Matrix is likely the best bet for your next step.

Upgrading to v1.9 won't fix anything, but you are up to date.

Bad View: http://doityourselfchristmas.com/forums/showthread.php?47322-Ron-s-Pi-Matrix-Board&p=467480#post467480
Good View: http://doityourselfchristmas.com/forums/showthread.php?47322-Ron-s-Pi-Matrix-Board&p=467483#post467483

covewi
07-27-2017, 05:22 PM
I don't know if this matters - but I finally found my multimeter and started testing voltage points. I am using this Distro Board https://www.pixelcontroller.com/store/index.php?id_product=46&controller=product and I checked to be sure the PS is putting out 5.0 volts on the dot.

All inputs to the Distro Board read 5.0 volts - but - all the outputs are only sending 3.80 volts to the P10 panels.
Is this normal and will it make a difference in the P10 functioning?

mrGrumpy
07-27-2017, 05:59 PM
I don't know if this matters - but I finally found my multimeter and started testing voltage points. I am using this Distro Board https://www.pixelcontroller.com/store/index.php?id_product=46&controller=product and I checked to be sure the PS is putting out 5.0 volts on the dot.

All inputs to the Distro Board read 5.0 volts - but - all the outputs are only sending 3.80 volts to the P10 panels.
Is this normal and will it make a difference in the P10 functioning?


Your missing a lot of information. First, do you have any of the P10'ls turned on

Input to Distribution Board. Is that at the PS or Distribution Board.
Outputs are 3.8. Again. Is this at the Distribution Board or at the P10

How long are your 5V power cables to the Distribution Boards.

With P10 sleeping, what is the power at the P10
Unplug 2 of the 3 distribution boards. What is the power at the P10
Plug one back in - what is the power at the same P10

Can't find my spare P10's to test mine, but I believe they should be nearing 5V

Oh, by-the-way: I use the same Power Supply for the P10 as fro the Pi.....all 5V

covewi
07-27-2017, 09:17 PM
Input to Distribution Board = 5.0 v
Output from the PS = 5.0V
Outputs at the distro board = 4.95 V
How long are your 5V power cables to the Distribution Boards = the longest is around 18 inches

With P10 sleeping, what is the power at the P10 = 4.95 V ( I had several loose GRN wires coming out of the distro board connectors - I was getting inconsistent readings)

Just another learning curve thing for me. My new PI output boards shipped today and I ordered a new soldering iron form Amazon.

mrGrumpy
07-27-2017, 09:46 PM
Input to Distribution Board = 5.0 v
Output from the PS = 5.0V
Outputs at the distro board = 4.95 V
How long are your 5V power cables to the Distribution Boards = the longest is around 18 inches

With P10 sleeping, what is the power at the P10 = 4.95 V ( I had several loose GRN wires coming out of the distro board connectors - I was getting inconsistent readings)

Just another learning curve thing for me. My new PI output boards shipped today and I ordered a new soldering iron form Amazon.

From previously posted picture, your power to distribution board looks very long. Keep that lead as short as possible, leaving some flexibility for you to install and trouble shoot

covewi
08-01-2017, 08:01 PM
Success!!!
I soldered a new Pi Matrix board and everything works.
Blue
37028
Red
37029
RGB
Trust mw the green works now - I just forgot to take a photo
37030
37031
Butterfly - notice the green works
37032

covewi
08-01-2017, 08:05 PM
Well - heres a little new problem - the text is out of sync.
37033
37034

I am almost there - but what moment of joy to see the entire panel light up. Thank you everyone for your persistent help.

Can we solve the text problem - please?

Covewi

mrGrumpy
08-01-2017, 08:39 PM
What did you use to generate the text. I would start there.

covewi
08-01-2017, 10:45 PM
xLights - It appears that the edges of the lettering "bleed"
37037
37038

mrGrumpy
08-01-2017, 11:10 PM
See if you can download and extract this fseq file.
Load it into FPP and try my text test

Done in Vixen, based on 1 - 6H x 5w panel, start channel is 1

37041

covewi
08-02-2017, 06:00 PM
This is a test file from mrGrumpy and the weird out of sync text still shows up - the upside down does not matter thats the way the file showed up.

37044

The next photo is also some text and the weirdness shows up everywhere.
37045

A video showing whats happening.
https://youtu.be/j_F3qpRQK3I


I read that the same thing is happening to jCook as posted 07-31-2017, 09:49 PM
http://doityourselfchristmas.com/forums/showthread.php?47485-P10-panel-columns-shifted

The red circles are the three rows that show interference. Its the last row of each section. In my case its panels 6-10 of group of three.
Also the light blue color indicates that at the very end of the 10 panel there is also some interference. This interference only shows up in those rows - not the first 5 panels of each group of three.

37046

mrGrumpy
08-02-2017, 07:16 PM
This is a test file from mrGrumpy and the weird out of sync text still shows up - the upside down does not matter thats the way the file showed up.


The next photo is also some text and the weirdness shows up everywhere.


A video showing whats happening.
https://youtu.be/j_F3qpRQK3I


I read that the same thing is happening to jCook as posted 07-31-2017, 09:49 PM
http://doityourselfchristmas.com/forums/showthread.php?47485-P10-panel-columns-shifted

The red circles are the three rows that show interference. Its the last row of each section. In my case its panels 6-10 of group of three.
Also the light blue color indicates that at the very end of the 10 panel there is also some interference. This interference only shows up in those rows - not the first 5 panels of each group of three.


Just an FYI: my test file was built in Vixen, which is BOTTOM LEFT input. Your page is TOP LEFT. Upside down was expected. The test does shows a different sequencer does similar.

I think I've gone as far as I can helping. You could have tweaked something in the Pi with the faulted Matrix Board, or Ron's Matrix board can't handle.

Try FPP or Ron for further assistance.

covewi
08-02-2017, 11:17 PM
Did some diagnostics - a single panel works correctly - but as soon as I try a 2 X1 panel matrix things go wrong.
With the 2 X 1 setup the first half of the first panel works correctly - then the next half and the second panel all have interference when showing text.'

You can also see some interference with any pattern.

tucsonlights
08-02-2017, 11:34 PM
Did some diagnostics - a single panel works correctly - but as soon as I try a 2 X1 panel matrix things go wrong.
With the 2 X 1 setup the first half of the first panel works correctly - then the next half and the second panel all have interference when showing text.'

You can also see some interference with any pattern.
Did you make your own risdon cables or did you buy them. I would swich out the ribbon cables and see if that is the issue. I had a similar issue.

covewi
08-03-2017, 08:58 AM
I am using the ribbon cables between panels that came with the panels.
The longer ones from the Pi came from Boscoyo - I have made every possible switch with the cables to rule them out with no luck.

I am now wondering about the panels themselves.

Covewi

mrGrumpy
08-03-2017, 08:53 PM
I am using the ribbon cables between panels that came with the panels.
The longer ones from the Pi came from Boscoyo - I have made every possible switch with the cables to rule them out with no luck.

I am now wondering about the panels themselves.

Covewi

Her is Vixen 2W1H Text test file in fseq - good luck
37062

I bought 2 spare panels this year from DIYLEDEXPRESS, and tested them with this test file. They run fine for me. Clean edges, and no noise.

covewi
08-04-2017, 01:12 PM
I probably know the answer to this question but I wanted to make sure I am not missing something that may have been causing my problems.

I have 5 X 6 (30 panel) matrix. After creating that model in xlights do I need to do anything with the Unicast or MultiCast? I am assuming no.
Can I run a sequence from my laptop using xlights directly to the matrix?

Could this make a difference? Is FPP not playing the xlights file correctly?

tucsonlights
08-04-2017, 01:44 PM
If you want to run the fpp from your pc you would put the fpp in bridge mode.

covewi
08-06-2017, 11:23 PM
First - would it make any difference if I ran FPP from my PC vs. running it from the uploaded file on FPP?

Now this is really weird - I made a second layer in xLights behind the text - as you can see the text looks fine - but as soon as the background stops the text gets weird.

https://youtu.be/VmWk0IF4X2Y

What the heck?

mrGrumpy
08-06-2017, 11:33 PM
First - would it make any difference if I ran FPP from my PC vs. running it from the uploaded file on FPP?

Now this is really weird - I made a second layer in xLights behind the text - as you can see the text looks fine - but as soon as the background stops the text gets weird.

https://youtu.be/VmWk0IF4X2Y

What the heck?

Do you still have the link to the xLights Wednesday Conference- share this video with them if you don't have answer by then - 7:00 eastern start time.

covewi
08-10-2017, 07:42 PM
Well today is a very disappointing day. A board member sent me a know working Raspberry Pi and a Ron' Pi Matrix board.
Guess what the problems still exist.
Lets see what I have tested.


changed ribbon cables
changed ribbon cables to diferent rows of panels
different Pi
different Pi matrix board
tried many different matrix sizes
a single P10 panel performs correctly - no text problems - but not if more panels are used
connected all the grounds together on the PS
grounded the PI and output board to the PS



I am really at a loss. I am really wondering if somehow its the panels themselves.

All thoughts and suggestions welcomed - I should have been an easy project

tucsonlights
08-10-2017, 07:45 PM
I wonder if the scan rate is wrong on the panels. Where are you located I can send you some panels to see if it's your panels

Sent from my SM-G955U using Tapatalk

JCook
08-10-2017, 08:55 PM
i could send you a couple of my spare panels and you can try them with your setup to see if they work. As long as you send them back. I got mine from Ray Wu and work fine.

mrGrumpy
08-10-2017, 09:02 PM
Well today is a very disappointing day. A board member sent me a know working Raspberry Pi and a Ron' Pi Matrix board.
Guess what the problems still exist.
Lets see what I have tested.


changed ribbon cables
changed ribbon cables to diferent rows of panels
different Pi
different Pi matrix board
tried many different matrix sizes
a single P10 panel performs correctly - no text problems - but not if more panels are used
connected all the grounds together on the PS
grounded the PI and output board to the PS



I am really at a loss. I am really wondering if somehow its the panels themselves.

All thoughts and suggestions welcomed - I should have been an easy project

Tucsonlights seems to have a great offer to nail down the P10 are the issue, but it may be only a couple panels with bad outputs.
Good Luck

covewi
08-10-2017, 10:30 PM
We will see soon. Hopefully the panel swap will solve the mystery.

covewi
08-13-2017, 12:37 PM
No further update on testing new panels yet - hopefully this week.
I believe the problem is the panels outputs somehow. Unless you have other ideas.

Here is a panel matrix - the text is fine
37166

Here is a two panel matrix - the text gets weird about halfway through the first panel and into the second
37167

Heres a three panel matrix - the same problem
37168

Here is made the second panel of the 3 panel matrix the first panel in a 2 panel matrix - just to show the each panel work fines as a single panel.
37169

Has anyone else purchased P10 panels in the last group buy and are you having any sorts of panel problems?

mrGrumpy
08-13-2017, 01:16 PM
Have you tried 3 panels in a row using 3 outputs. Using the 3W X 1H setup

See Pic for setup
37170

covewi
08-13-2017, 05:37 PM
That works and shows that all three Pi Matrix board outputs work correctly - even with my output board.
37175

I made a 6 X 1 matrix ( 2 panels on each output) and again the second half of the first panel and the entire 2nd panel on each output has the problem.
Ian only showing the last 2 panels of output three so I could get close enough for a good photo.
37176

Its the boards! Resolution?

mrGrumpy
08-13-2017, 06:12 PM
Yes, I think we are one step closer to prove the P10 panel output has issues.

covewi
08-15-2017, 09:00 PM
Another board member came up with a post from the xLights Facebook page. I don't have Facebook but it dealt with P10 panels having flickering problems and they determined it was a voltage issue.

It appears they turned the PS voltage down to 4.3 V and running the FPP brightness at 100% got rid of the problem. Others talked about voltages run at 4.7 and never at 5.0.

I tried changing the voltages from 5V to 4.5 and 4.3 - the lowest the PS it will output. I also turned the FPP brightness all the way up to 100%.
The panels still flicker, there are extra pixels that cause the ghosting effect and at those voltages / brightness the text is almost not viewable.

I should have a few borrowed panels later this week to run a comparison. Its very strange when 2 or more panels are connected the second half of the first panel and all the other panels run funky.

covewi
08-16-2017, 06:18 PM
The mystery is over. I received 4 borrowed panels from a fellow board member. They will remain unnamed unless they would like to come forward.

Any way - using my Pi, Ron's Pi Matrix board (the one built), PS, cables and xLight sequence - the borrowed boards work perfectly with voltages down to 4.3 V and the brightness ranging from 20% - 100%.

The final decision the P10 PANELS ARE FAULTY WHEN MORE THAN ONE IS CONNECTED IN SERIES. There is no denying it.
As you probably know with the help of other board members I have swapped out a PS, PI's, Ron's Pi Matrix board, cables and sequences from xLights and Vixen.

A single panel works fine (displaying text and effects without interference or ghosting) but as soon as multiple panels are connected the problems show up at all PS voltages and FPP brightness levels. They only logical explanation is there is a problem with the OUTPUT of the panels.

Perfect Text at 100 % and 4.5 volts
37187

Further proof

https://www.youtube.com/watch?v=n_WNECjUPU0

And now for the bad panels - nothing has changed except for the panels.

https://youtu.be/0eMEPLEQgJw

mrGrumpy
08-16-2017, 06:38 PM
Bruce, I think you won the battle, but lost the war....Wishing you luck getting those panels replaced.

DrNeutron
08-16-2017, 06:45 PM
The mystery is over. I received 4 borrowed panels from a fellow board member. They will remain unnamed unless they would like to come forward.

Any way - using my Pi, Ron's Pi Matrix board (the one built), PS, cables and xLight sequence - the borrowed boards work perfectly with voltages down to 4.3 V and the brightness ranging from 20% - 100%.

The final decision the P10 PANELS ARE FAULTY WHEN MORE THAN ONE IS CONNECTED IN SERIES. There is no denying it.
As you probably know with the help of other board members I have swapped out a PS, PI's, Ron's Pi Matrix board, cables and sequences from xLights and Vixen.

A single panel works fine (displaying text and effects without interference or ghosting) but as soon as multiple panels are connected the problems show up at all PS voltages and FPP brightness levels. They only logical explanation is there is a problem with the OUTPUT of the panels.

Perfect Text at 100 % and 4.5 volts
37187

Further proof

https://www.youtube.com/watch?v=n_WNECjUPU0

And now for the bad panels - nothing has changed except for the panels.

https://youtu.be/0eMEPLEQgJw

Well a guy with a few seasons of MacGyver on the shelf was bound to figure it out!
Good stuff!:rolleyes:

covewi
08-16-2017, 06:56 PM
An email, with links to my posts, was sent to the vendor along with what I feel is evidence of working through the problem and supplying proof of faulty panels.

I will certainly keep the forum updated as to this vendor willingness to make good on a refund.
Nice reference - how did you know I was a fan?

37188

DrNeutron
08-16-2017, 09:36 PM
Hehe check your pic above.....Seen the MacGyver box set in background.
The vendor I used for my panels a couple years ago is JS Industrial : https://www.aliexpress.com/store/1090165?spm=2114.10010108.0.0.4b24ae2boxpfgd
Actually RonP suggested them when I bought his 16 universe controller to drive the small 3X2 matrix I built with them. It is now on its 3rd year and going strong with no issues (fully enclosed with plexi front).

covewi
08-16-2017, 11:18 PM
DrNeutron - Too funny. I did not even think to look in the shelf.

covewi
08-17-2017, 11:39 AM
I don't think its too much to hear from a vendor that sold me 35 faulty P10 panels considering this is what's posted on their website.

"Returns - All pre-sales are final except in cases of defective product. If you have received a defective product please contact us."

Contacting them is easy. Hearing back is a different story.
I will be purchasing my replacement panels from another forum member this week.

I would like to thank all the forum members who reached out to help - especially those who trusted me enough to send their own hardware for testing with my panels. This really shows how a community of like minded people can help each other.

Today is day number 1 after requesting to be contacted by DIYLEDEXPRESS regarding their faulty panels. Please feel free to contact me by PM regarding any issues you may have with this online vendor. My matrix is now going to cost me twice as much as I expected and this is unacceptable.

Does this forum have any rules regarding vendors who do not live up to reasonable expectations.

Covewi

covewi
08-17-2017, 03:20 PM
I just received an email from DIYLEDEXPRESS. I should know more later tonight or tomorrow.

Covewi

mtosh
08-17-2017, 03:47 PM
I have read about tons of issues with the latest p10's. I did everything to get them working. Apparently the new panels have a different refresh rate. I had to move from hat to the bbb and then still no luck till switching to the beta version of fpp. Several are having success with this. I'll take your panels if the vendor doesn't want back.


Sent from my iPhone using Tapatalk

covewi
08-25-2017, 11:16 AM
I'll take your panels if the vendor doesn't want back. Sent from my iPhone using Tapatalk

mTosh - would you be interested in me sending you a few panels to test with your BBB setup?

Covewi

covewi
08-25-2017, 06:10 PM
I have read about tons of issues with the latest p10's. I did everything to get them working. Apparently the new panels have a different refresh rate. I had to move from hat to the bbb and then still no luck till switching to the beta version of fpp. Several are having success with this.


Does any one have more information on the topic of scan rates, FPP and BBB?

It looks like I will be stuck with 35 panels and this might be a fix - however is it just a work around? I don't want to purchase a BBB and output board and be stuck with that too - if the panels are faulty/

Thanks

tucsonlights
08-25-2017, 07:52 PM
Send me some panels and I will test them with my bbb and octoscroller and see if they will work.

covewi
08-25-2017, 10:16 PM
Is this only for text display or with it play all effects within xLights including video?

"An "octoscroller" is simply a message display that happens to use a BeagleBoneBlack microcomputer"

tucsonlights
08-25-2017, 10:25 PM
Is this only for text display or with it play all effects within xLights including video?

"An "octoscroller" is simply a message display that happens to use a BeagleBoneBlack microcomputer"I play video and text as well as xlights effects on the p10s with the BBB and octo

Sent from my SM-G955U using Tapatalk

covewi
08-26-2017, 07:19 PM
I just purchased 34 P10 panels from https://www.crockettfantasyoflights.com.

Their a pleasure to deal with. I received a thank you email within 30 minutes of the purchase and couldn't be happier with the support they provided in helping to get my matrix up and running.

covewi
09-08-2017, 02:27 PM
For anyone that has had problems with "ghosting" on their P10 panels - regardless if its a scan rate or software problem - it appears there is an easy work around.
CrocketFantasyLights suggested making a black layer underneath the text - which worked.

What also works is to make a layer under the text and use the "off" effect. I also noticed this weekend in Las Vegas on all the giant digital billboards - none of them show text without some sort of background color or image. Keeping the pixels busy behind the text probably makes for a sharper text edge.

Its also more appealing to the eye - its time I start to think and look for guidelines for the production of digital projects.

37337
37338
37339

Gilrock
09-08-2017, 02:39 PM
That doesn't make sense to me if you are using xLights cause I can tell you the output already has black pixels surrounding the text so adding a black layer or off effect does not change the data being sent to the display. The only way that can help is if you have another effect painting to those same pixels that the text is overlaying.

mrGrumpy
09-20-2017, 02:52 PM
I think I found Covewi's P10 issue.....I think it's the Pi 3

Looking at this problem, again, I think I found the issue. With the Pi 3 attached to Ron's Matrix, I get noise around the TEXT. If I change just the Pi to a 2B, Zero or Zero W, I have no noise.
This is typical for 2 recently new P3's

This is repeatable too.

.

covewi
09-20-2017, 06:23 PM
mrGrumpy - I cant wait to try this fix. I have a few of the "faulty" panels left over and will build a Tune To sign.

Thanks - I hope this helps out others too.

covewi
10-09-2017, 03:57 PM
After nearly 3 months of work my 5 X 6 P10 panel matrix is up an running. Thanks to everyone for their help and ideas. I learned a lot and can't wait for the holidays.
Thought I would post a few photos and ask another question.

Which do you suggest? Lexan or Plexigalss (http://www.plasticgenius.com/2011/08/plexiglass-vs-lexan-matchup.html)
Also what thickness would work best for this size enclosure?

37641
37642
37643

tucsonlights
10-09-2017, 04:36 PM
After nearly 3 months of work my 5 X 6 P10 panel matrix is up an running. Thanks to everyone for their help and ideas. I learned a lot and can't wait for the holidays.
Thought I would post a few photos and ask another question.

Which do you suggest? Lexan or Plexigalss (http://www.plasticgenius.com/2011/08/plexiglass-vs-lexan-matchup.html)
Also what thickness would work best for this size enclosure?

37641
37642
37643I use 1/4 inch lexan it stands up well to IF light and adds nice rigidity to the enclosure.


Sent from my SM-G955U using Tapatalk

mrGrumpy
10-09-2017, 05:48 PM
Extruded Acrylic is the best and least pricy. Lexan is strong, but more prone to scratching and yellowing - does resist them rock throwers best. Acrylic does well with rocks, but nowhere like Lexan. Acrylic is much hard to scratch and no yellowing over time. I have 3 smaller P10, second season and all - all with acrylic.....Biggest P10 is 12 panels; I use 1/8", but 1/4 would be best for your size.

covewi
10-13-2017, 11:23 AM
Has anyone used or know about this product.

http://www.homedepot.com/p/OPTIX-48-in-x-96-in-x-1-4-in-Clear-Acrylic-Sheet-MC-102S/205351368

Its expensive but its more online especially when I figure in shipping and cutting costs. Hopefully this will be an acceptable choice.

covewi
11-14-2017, 08:34 PM
I couldn't see spending $200 + shipping for 1/4" X 65" X 39" for plexiglass to cover the matrix right now. I saw online where people are making storm windows and screen porch windows out of 16 mil clear vinyl. Its very strong, clear and tough.

I am using these metal frame making strips - https://www.homedepot.com/p/Saint-Gobain-ADFORS-5-16-in-x-84-in-Brown-Aluminum-Screen-Frame-Piece-FSP8492-U/100397094 and https://www.growerssupply.com/farm/supplies/prod1;gs_greenhouse_films;pg105129_107760.html

I did find that my local Joann fabric has the same stuff - http://www.joann.com/disco-vinyl-fabric-clear/1532449.html

The corners will be connected using these - https://www.homedepot.com/p/Saint-Gobain-ADFORS-Brown-5-16-in-Screen-Frame-Corners-4-Pack-FSP8571-U/100397097

The plan is to screw down the finished frame to the 1 X 1 that hold the P10 panels and then use silicone sealant to seal any gaps. This will also save adding 35 pounds to the overall weight of the matrix.

I will post a photos by the weekend - I have to get this done in time for the next Thursdays opening night.

covewi
11-18-2017, 04:24 PM
Made the screen frame today - hopefully I get it covered in the next few days with the clear vinyl.

38317
38318
38319

covewi
11-20-2017, 10:42 PM
The clear vinyl frame was installed today and look like theres no interference. I feel that it will be a good alternative to plexiglass - at least for this first display season.

https://youtu.be/xW3VYgd8KeA

covewi
11-23-2017, 10:37 AM
I finally got the protective vinyl secured. 38408

covewi
12-15-2017, 09:33 PM
Some things never change - always a little problem. Does this happen to anyone else?

The FPP runs perfectly from day to day but then every once in a while the show starts and just freezes at the first sequence. Reboot, restarts and power down doesn't help.
So I take out the USB drive and reload all the sequences only and everything works again.

Cover

kzaas
06-29-2018, 01:25 PM
Follow

covewi
08-28-2018, 07:07 PM
Wow - its already been another year and thought I would get a head start on setting up the Pi P10 matrix.
I updated FPP to the latest version v1.10-27-g259bc38 and have no issues with the software as long as the Pi is hardwired.

Heres an image of the wifi setup and believe its okay.
Once I disconnect the ethernet cable I cant connect to FPP.

Covewi

39982

mrGrumpy
08-28-2018, 08:27 PM
I'm heard from several that FPP does not like special characters in SSID. Appears the dash may be your issue.

covewi
08-28-2018, 10:21 PM
No go. Any problems reported with wifi not working on version 10?

I only get a 192.16.8.1.XX IP address showing up on the start screen no 10.0.0.X wifi IP address like all other wifi devices on the network.

mrGrumpy
08-28-2018, 11:37 PM
Are you on a 5g network. FPP only likes 2.4g

covewi
08-29-2018, 08:16 PM
Everything seems to be working now. Thanks for the tip

Upgraded to latest version of FPP - what happened to the up and down arrows for setting the orientation of the P10's?

39991

mrGrumpy
08-29-2018, 08:35 PM
I have them on my screen - also, your channel count is off.

What Browser are you using? FPP does not like IE. EDGE, FireFox, Safari and Chrome all work.

What image dis you start with? I send you something

covewi
08-29-2018, 08:47 PM
39993

I think the count is now correct - but no arrows with showing while using Chrome

covewi
11-22-2018, 01:25 PM
Updated to version 2.1 of FPP and finally got everything working - sort of.

Everything seems to be shifted over.

And shouldnt I have more channels that 1536?

This is a P10 matrix (32X16 panels) and is 5WX6H
40571

ezellner
11-23-2018, 09:06 AM
Updated to version 2.1 of FPP and finally got everything working - sort of.

Everything seems to be shifted over.

And shouldnt I have more channels that 1536?

This is a P10 matrix (32X16 panels) and is 5WX6H
40571

It looks like the ending zero is dropped (should be 15360). May just be a display issue where they only allocated 4 places instead of five or six.

mrGrumpy
11-23-2018, 11:02 AM
It looks like the ending zero is dropped (should be 15360). May just be a display issue where they only allocated 4 places instead of five or six.

Count should be 46080. 32 X 16 X30 X 3

ezellner
11-23-2018, 12:15 PM
Count should be 46080. 32 X 16 X30 X 3

Right! Not enough coffee...

covewi
11-24-2018, 09:43 AM
Thank you for your help - everything is working. I need to start much earlier next year.