P5 Panels

[TheDucks]

DO NOT tie the AC side together on any that are NOT fed from the SAME cord. (AKA Unplug the cord and those go dead)

 

My comment earlier is about All Negatives for those supplies using the same controller .  This eliminates possible odd ground loop paths.

To reinforce K6CCC: The fuse protects against the Maximum SOURCE going down a path that is under that CAPACITY. It has nothing to really do about how little of that is being used.

 

Here is a good Wiki page on Powering in General (not just as the title implies) https://www.doityourselfchristmas.com/wiki/index.php?title=Power_Injection

[rick gurnee]

As I said "I will have four AC plugs coming out of my display." So the two power supplies that have the hot leads connected will be on the same plug.

Also, I want to correct myself.  The Holiday coro controllers that I have do have a 5amp fuse on each individual port.  I do not see a fuse from the power supply to the controller board, but there might be one on the board.  As I said, I will be using 15 amp inline fuses for every pair of panels.  Steve, just out of curiosity, if you can easily look at the back of your display do you see fuses?  I didn't see any in the pictures you sent, but I could have missed them.

[TheDucks]

5A protecting 18ga is fine. With 5V, don't go wild with fuses. Each drops voltage.

[rick gurnee]

I put together a small 3x2 display to test and am running into a problem.  These are P5 outdoor panels with a 1/8 scan. I am using a raspberry pi 4 with a Hansen Electronics Hat.  I installed FPP 7.5 and configured the channel output as follows: Layout 3x2, single panel 64x32 1/8, wiring pinout standard, gpio slowdown set to 5.  After rebooting the pi, I ran the test pattern and got the attached output.  As you can see, I appear to have two problems (probably related) The panels are not ordered correctly and there appears to be two images in each panel. I tried changing the gpio slowdown, but that made things worse.  I thought maybe the wiring pinout is wrong, but I don't want to change something that could do damage to my panels.

https://drive.google.com/file/d/1pRYeTxKaI9rz_SsRmVuaCE5VES5TER0A/view?usp=drive_link

 

[k6ccc]

How are the panels physically wired?  By that, I mean, what output of the Hat goes to which panel, and how are the panels daisy chained?  A screen capture of the panel setup page in FPP would likely be useful.

 

[rick gurnee]

The first output of the hat goes to the top right panel which is daisy chained to the other two panels along the top row.  The second output goes to the bottom right panel which is daisy chained to the other two panels along the bottom row. I am not using the third output.

By the way, I was able to successfully set up a color light card to light the panels correctly.  I am now adding a pi to the color light card and will see if I can set that up.  After I do that, I will put the pi with the hat back on it and take a screenshot of the FPP setup page.

[rick gurnee]

As I said, I was able to get the color light card working fine.  I configured the second pi and everything works fine with the pi and color light card.  I then reloaded FPP onto my first pi (with the Hansen hat) and reconfigured it.  I still get the same issue when I select "test pattern".  Here is my channel output setup page.

[stevehoyt]

@rick gurnee

Can't help but ask....How the P5 grid coming?

[rick gurnee]

Thanks for asking. I'm glad there's 9 months to go before Christmas, because at the rate I'm going, I'm going to need every one of those days!

I started off by taking a scrap piece of ¼ “ plywood and building a test 3x2 display. I had some issues configuring FPP because most of the videos I watched skipped over several important issues. However, after reading the FPP manual it all made sense. I also resolved some issues by making every IP address static. I now have a test display talking to LOR using universes 101-172. My final display will be 6x7 using universes 173-676.

My brother 3D printed me a template so I could drill the 12 4mm holes needed to mount each panel to the back of the plywood. I transferred that to a piece of acrylic which seams to be working pretty well. My mistake was trying to drill the holes too small. Once I made them bigger so I could adjust the panels, the mounting got a lot easier.

I think I am close to starting to build my 6x7 display. Drilling over 500 holes and cutting 126 cutouts is going to take some time. I've done some research on configuring the color light card and think I might use two of them. I can configure it with one, but with 6 panels on each port I will have to compromise on some of the settings. Either lower the refresh rate, reduce the brightness or change the CPU clock rate. With a second card, I wouldn't need to do this. However, I made some of those changes on my test display and I couldn't see any difference.

You were right about how long it take to prepare an LOR sequence when you have a large panel. One 3 minute song is taking as much as 15 minutes to prepare. My laptop isn't that fast, but before I had the panel in my Preview it was taking about a minute to prepare.

 

[stevehoyt]

@rick gurnee

I know the feeling. Like when I was pushing 2600 pixels in by hand for my "project to reuse some pixels. But drilling 500 holes... Slow and steady as they say.

You must have a nice shop to take on this project. 

I am sure it is going to be a show stopper. 

FYI regarding the "compile time. I ended up making all my sequences no longer than 5 minutes. I also made good use of the sequencer button to turn off compiling every time I made a change. 

I am looking forward to seeing build and finished product pics when it is up and running.