Any help on a LED Video Wall

[pete10uk]

Hi all

 

I’m not that familiar with video walls, so any guidance would be appreciated.

 

I’ve been donated an old p10 3 led (RGB) wall and I’m trying to get it working. The wall is one I used to borrow from time to time so I know the equipment is all capable of working together, the reason I got the wall is the owner didn’t need it anymore and couldn’t get it working for some reason.

 

The wall had a linsen sender and each cabinet had a receiver card, I suspected that the sender had broken so I bought a Novostar mctrl300 sender and a novastar receiver card to test but all I get is gibberish on screen, sometimes the screen kind of reacts to movement in the source but nothing remotely like the source. This is the same issue with the original linsen cards and the new cards, I suspect that some of settings in the original cabinet were lost and these settings are not present in the Novostar software yet.

 

I have tried smart setup but it doesn’t complete successfully.

 

I’m stuck on what to try next.

 

I think it is the led module settings which are missing on the receiver card.

 

Any thoughts?

[TheLX1]

Do you know which IC the panels are using? You probably need to flash the receiver cards with the correct RCFG file (the file that tells the receiver card which IC is driving the LED's and how to talk to it)

 

Failing that, is there any manufacturers information or model numbers on the panels themselves?

[pete10uk]

Do you know which IC the panels are using? You probably need to flash the receiver cards with the correct RCFG file (the file that tells the receiver card which IC is driving the LED's and how to talk to it)

 

Failing that, is there any manufacturers information or model numbers on the panels themselves?

 

Hi thanks for the reply.

 

Yes I’ve been running the smart setup where it asks for the driver chip but the one on the board is not listed so tried the generic option and also tried the closest Marco (there’s one with just a couple of numbers different) to the one on the board with no joy, it also asks for the decode chip and that is listed. I have a few rcfg files from the person who previously had it and the headline settings look the same.

 

I had it earlier where there was a distinct line horizontally across the screen every 6 or so lines and if I moved the mouse to a certain part of the laptop screen I got movement but it affected the entire row. There was definitely some movement.

[TheLX1]

In that case, I think your earlier hypothesis is probably correct and that the panels have lost their settings and are currently unsupported by the Novastar software.

 

It may be worth taking a look on rcg-rcfg.com, as they have quite an extensive library of RCFG files that are freely available and may have a user created Novastar file that is compatible with your panel IC's

 

I must admit, this is the first time I've heard of somebody trying to convert a Linsn panel to Novastar!

 

 

 

 

 

 

 

[pete10uk]

In that case, I think your earlier hypothesis is probably correct and that the panels have lost their settings and are currently unsupported by the Novastar software.

 

It may be worth taking a look on rcg-rcfg.com, as they have quite an extensive library of RCFG files that are freely available and may have a user created Novastar file that is compatible with your panel IC's

 

I must admit, this is the first time I've heard of somebody trying to convert a Linsn panel to Novastar!

 

Thanks again for your reply.

 

I suspect your right with the Novastar compatibility, the receiver card has HUB75E connectors where as the Linsn had HUB75 I assume these are not backwards compatible.

 

I'm just a little confused with the Novastar and Linsn software, both are very similar and have a smart setup option, the receiver chip is not listed but there is a common selection and the decode chip is listed, I don't really get anything meaningful out of either, both ask for the chip information and then just move on to a top left image onscreen asking do you see a blank screen or do you see the desktop, I see scramble and there's not an option for that. Not sure if you have to enter some more information before running the smart setting part.

 

All the videos I've watched don't show any scrambling. Is there a blueroom style forum dedicated to LED out there?

 

Not sure which way to read your last comment, are you saying why would you convert from linen to Novastar, linen is much better, or are you saying that LED models are normally made for one particular control manufacturer.

 

My reasoning is that I have a new wall on it's way to me which is based on Novastar, so I have purchased a spare sender and receiver for this new wall as a backup, I thought as I'm not having much luck with the linsn on the old wall I'd try this setup to familiarise myself with it, I'd not accounted for them not being compatible.

 

As there are hundreds of sellers of various panels out there who don't specify much information that I'd be able to get the old wall back in to life for some playing around using something I know is not broken.

 

The old wall came with 3 senders, a spare receiver, a spare module and spare cabling, so I've pretty much ruled out that the original wall has just blown beyond repair, but that said even though I have some config files that look to be named as though they should work, uploading them doesn't seem to work. Either LED is beyond me or this old wall has serious issues but as I've tried from scratch with a new sender, new cabling, new module I'm just confused now.

 

I might have now blown some of the spares with the HUB75 and 75E mismatch! :-(

 

Thanks again for your time

 

Just for reference the old wall modules look very similar to this.

 

https://www.alibaba.com/product-detail/Full-Color-Outdoor-Waterproof-Programmable-P10_60297917461.html?spm=a2700.7724857.normal_offer.d_image.37377bcbirfvOv

[pmiller056]

HUB75E suggests that it is a 1/32 scan panel. A HUB75E panel uses what is one of the Gnd pins on HUB75 connector as the 'E' pin. No - I'm not an expert but I have played with some small ones and a Rpi for curiosity.... This appears to be software configuration/setup problems. The chips used in these panels are fairly generic. It's how they're connected on the panel that really matters. Is there a part number on the rear of the panel that leads to a datasheet with a bit of google-fu? The only way I've heard of to damage these panels is if the multiplexing stops leaving a static line of extremely brightly lit LEDs - the lit LED's are only designed to be lit on a low duty cycle as the multiplexing works.

[pete10uk]

HUB75E suggests that it is a 1/32 scan panel. A HUB75E panel uses what is one of the Gnd pins on HUB75 connector as the 'E' pin. No - I'm not an expert but I have played with some small ones and a Rpi for curiosity.... This appears to be software configuration/setup problems. The chips used in these panels are fairly generic. It's how they're connected on the panel that really matters. Is there a part number on the rear of the panel that leads to a datasheet with a bit of google-fu? The only way I've heard of to damage these panels is if the multiplexing stops leaving a static line of extremely brightly lit LEDs - the lit LED's are only designed to be lit on a low duty cycle as the multiplexing works.

 

Thanks for the reply.

 

The receiver card has a hub75E connector but the module is hub75.

 

The only references on this are YLP10-16x16-11-346 on the plastic frame which is how I found the item linked to above, and P10c1616-5TD5 on the circuit board which returns no results on Google, if I remove the 5TD5 you get this http://persian.led-billboard-display.com/quality-347453-high-resolution-led-display-modules-full-color-rgb.html which is very similar but the board is slightly different layout and the board colour is green where as mine is blue. The receiving chip s are Chipone ICN2026DF the Decode chip (at least this number is found in the software as a decode chip) is SM74HC138D and there are 2 more chips which I thought were also decode chips as the number is close to the other decode chip SM74HC245D and 4 other smaller chips 4953B TF342C

 

Is there any compatibility with Hub75 and the E if I disconnect the E pin? if you know from your playing.

 

Thanks

[TheLX1]

Not sure which way to read your last comment, are you saying why would you convert from linen to Novastar, linen is much better, or are you saying that LED models are normally made for one particular control manufacturer.

 

I was just saying this is quite an unusual thing to do, as modules are normally made for one particular control manufacturer. Personally I prefer the Novastar system over Linsn, its been a lot more reliable and rugged in my experience.

[timsabre]

I don't know if this will help but I have used similar panels with Arduino and can explain the "E" pin, the 74HC138 just has 8 outputs which are selected by 3 binary inputs (so 000 selects output 1, 001 selects output 2 etc). This is used to scan through the rows on the display, the controller outputs the column data for all the pixels in row 0 and outputs the code 000, then for all the pixels in row 1 and outputs 001, then the same for the other 8 rows. This would be a 1/8 scan. On the input connector these lines are called "A B C".

 

On most panels there are 2 of these chips and they are controlled by a 4th line "D", this alternately selects rows 0-7 and 8-16 giving a 1/16 scan.

On a few panels there is a 5th line "E" which gives you a 1/32 scan.

If you don't have the right ABCDE drive signals then the controller outputs get repeated on some rows and other rows don't light up at all. If you're just getting a single line with all the pixel data mixed up, sounds like it is not scanning any rows at all. The panels are pretty dumb, they are just some logic latches basically to convert the serial pixel data into LED outputs.

 

[pete10uk]

Not sure which way to read your last comment, are you saying why would you convert from linen to Novastar, linen is much better, or are you saying that LED models are normally made for one particular control manufacturer.

 

I was just saying this is quite an unusual thing to do, as modules are normally made for one particular control manufacturer. Personally I prefer the Novastar system over Linsn, its been a lot more reliable and rugged in my experience.

 

That’s what I had read and watched online, many people saying the Novastar kit is probably where most people will end up after trying others, so with the new wall coming my way I thought I would save the trouble and pay that bit extra.

 

While looking at the many sellers of led modules I’ve never noticed any saying one particular manufacturer. I’ve seen a few with all of them listed so my assumption was the modals are pretty standard so was expecting there not to be a problem as long as I get the settings correct. Obviously there would be exceptions and my assumptions could definitely be wrong.

 

Thanks

 

I don't know if this will help but I have used similar panels with Arduino and can explain the "E" pin, the 74HC138 just has 8 outputs which are selected by 3 binary inputs (so 000 selects output 1, 001 selects output 2 etc). This is used to scan through the rows on the display, the controller outputs the column data for all the pixels in row 0 and outputs the code 000, then for all the pixels in row 1 and outputs 001, then the same for the other 8 rows. This would be a 1/8 scan. On the input connector these lines are called "A B C".

 

On most panels there are 2 of these chips and they are controlled by a 4th line "D", this alternately selects rows 0-7 and 8-16 giving a 1/16 scan.

On a few panels there is a 5th line "E" which gives you a 1/32 scan.

If you don't have the right ABCDE drive signals then the controller outputs get repeated on some rows and other rows don't light up at all. If you're just getting a single line with all the pixel data mixed up, sounds like it is not scanning any rows at all. The panels are pretty dumb, they are just some logic latches basically to convert the serial pixel data into LED outputs.

 

Thanks for this, it makes a little sense to me. So what your saying is that the e is an additional data line to improve the scan rate. If that’s the case I should be seeing an image on the screen but potentially with doubled up lines or lines missed?

[timsabre]

This page has some info about the control signals:

https://learn.adafruit.com/32x16-32x32-rgb-led-matrix/new-wiring

 

The "E" line allows an extra 16 rows of LEDs to be controlled, but at the expense of a longer scan time - this makes the panel dimmer and means you have to scan faster to keep the same refresh rate. Your panel is a 16x16 so it won't have a use for the E line. If your controller is driving an "E" line then every 16 rows will have 2 sets of pixel data displayed on top of each other. If you are just getting a single line then it isn't scanning at all which sounds like a problem with the controller or the wiring.

 

 

[pete10uk]

Thanks all for the responses on this one.

 

Just for the benefit of anyone else who ends up here with the same issue, I have now resolved this problem. the LED modules from the factory have had their ribbon cable connection installed the wrong way around.

 

On close inspection I could see some spraining on the connecter where the key had been jammed in the wrong way around on the sender card. I tried this and it fixed the issue on both the Linsn and the Novastar kit.

 

What was confusing was the fix I.e the cable reversal was done on the receiver card, but the issue was on the modules.