bilalibrir commented 2 years ago

I'm having a problem using an RGB (16*32) 1/4 Scan panel with this Library with no succes , the RGB panel has the snake pattern. are you working to implement this feature or not? if not please can you do it and thanks in advance. I tried to read the code but didn't understand most of it. thankyou.

board707 commented 2 years ago

Hi, thanks for interesting. The current version of the library supports only straight-line pattern matrices. I have experience with matrices with other types of patterns, but there are too many types and it is difficult to understand them without having a matrix in hand.

bilalibrir commented 2 years ago

I'll help from my side and i will try my best, i have the hardware and. i can already see the the pattern i believe it's the same in the attachment. It only uses the A,B address wires just 2 wires. Thank you 😊😊. P10-32x16-RGB-1-4-scan

board707 commented 2 years ago

Hi @bilalibrir As I found out, I have a similar 1/4scan matrix, only the order of its segments is 0-1-3-2(according to the diagram above) instead of 0-1-2-3 like yours. So I'll try, but it will take a few days.

bilalibrir commented 2 years ago

Good luck 🤞🤞. And thank you for your help and time. What can i do to help from my side. If you need to test code or anything, i'm happy to help.

bilalibrir commented 2 years ago

hi again, i think this may help. thanks to Ruud he modified RGBmatrixPanel to work with 32*16 1/4S matrixs. see the attached link . https://forum.arduino.cc/t/p10-32x16-led-panel-1-4-scan-now-works-with-adafruit_gfx/483621

bilalibrir commented 2 years ago

helllo, update, after trying a different library "PxMatrix" with an arduino Uno and my panels this is what i find:

1- my panel uses STRAIGHT Mux pattern and need's all A,B,C,D pins to be connected (it's not a BINARY mapping of outputs A,B,C,D). you have to set pin A low to send data for the first Row and pin B after that, then pin C, then pin D.

2- for the Scan pattern my panels uses ZAGGIZ pattern , is this library the have a differant set of patterns : ( display.setScanPattern(x) where x={LINE, ZIGZAG,ZZAGG, ZAGGIZ, WZAGZIG, VZAG, ZAGZIG, WZAGZIG2}.)

here is a link to the library https://github.com/bilalibrir/PxMatrix.

I think that we can learn alot form this library. thank you for your time.

board707 commented 2 years ago

I know this library. As I understand, the PxMatrix library worked with your matrix, so my help is no longer needed?

bilalibrir commented 2 years ago

hello @board707 , I still need your help, because I want to use an STM32 on my project. i'm trying on my side but i don't have the skills and i don't understand the math behind, i'm still learning but i fell like my head is going to explode. so your help is really appreciated.

this is what i did yesterday, i modified the DMD_RGB.h (dev-V2) , line 653-654: ` template class DMD_RGB<RGB32x16plainS8, COL_DEPTH> : public DMD_RGB_BASE2 { public: DMD_RGB(uint8_t mux_list, byte _pin_nOE, byte _pin_SCLK, uint8_t pinlist, byte panelsWide, byte panelsHigh, bool d_buf = false) : DMD_RGB_BASE2(4, mux_list, _pin_nOE, _pin_SCLK, pinlist, //bilal panelsWide, panelsHigh, d_buf, COL_DEPTH, 4, 32, 16) // bilal {}

}; `

and the generate_muxmask() , in DMD_RGB.cpp (dev-V2) : ` void DMD_RGB_BASE::generate_muxmask() {

define set_mux_ch_by_mask(x) ((uint32_t) x)

define clr_mux_ch_by_mask(x) (((uint32_t)x) << 16)

/*
for (uint8_t i = 0; i < nRows; i++)
{
    mux_mask2[i] = 0;
    for (uint8_t j = 0; (1 << j) < nRows; j++)
    {
        if (i & (1 << j))
        {
            mux_mask2[i] |= set_mux_ch_by_mask(mux_ch_masks[j]);
        }
        else
        {
            mux_mask2[i] |= clr_mux_ch_by_mask(mux_ch_masks[j]);
        }
    }
}
mux_mask2[nRows] = mux_mask2[0];
 */
for (uint8_t i = 0; i < nRows; i++)     
{
    mux_mask2[i] = 0;
    for (uint8_t j = 0;  j < nRows; j++) 
    {
        if (i ==  j)
        {
            mux_mask2[i] |= clr_mux_ch_by_mask(mux_ch_masks[j]);    //low
        }
        else
        {
            mux_mask2[i] |= set_mux_ch_by_mask(mux_ch_masks[j]);    //high
        }
    }
}
mux_mask2[nRows] = mux_mask2[0];

} ` to output a STRAIGHT muxmask, and i think that it works but i have nothing showing on the panel, no output on the RGB's outputs.

thank you for your time and help.

board707 commented 2 years ago

Hi your fix of the generate_muxmask() seems to be correct However, you do not need to edit existing templates, it is better to create a new one.

Besides that, I found a few minor bugs that need to be fixed before running your matrix. So, please wait a bit, I will post the corrected code and we will continue

board707 commented 2 years ago

Hi, I just uploaded the corrected code to dev-V2, please update the library from github.

Then replace the DMD_RGB.h in your copy of the library with file from the archive attached to the message. In this file, I have added a new template for your matrix.

However, all of the above only fixes the problem with mux pattern. To change the date pattern, we first need to understand how the pixels on your matrix are connected. To investigate this, I will ask you to run test code bilalibrir_test.ino from the archive and make a video how it work in your matrix. Thans for your help. bilalibrir_test.zip

bilalibrir commented 2 years ago

Helllo @board707, I will try this code Tonight, and update you with everything. Thank you for your help.

bilalibrir commented 2 years ago

hello @board707 , you are a genius, you made it look so easy.

https://user-images.githubusercontent.com/86444589/152657496-ec7d5fa5-2cd9-4314-8173-554103113e36.mp4

thank you for your time and help.

board707 commented 2 years ago

Hi, my congratulations. It looks like you have the same matrix as mine. Wait a little more, I will send you the corrected code for it.

board707 commented 2 years ago

Tried to add support for your data pattern, check it. Replace file DMD_RGB.h in the library with one from attachment. bilalibrir_2.zip

bilalibrir commented 2 years ago

hello, it didn't work for me, i will upload a new video.

bilalibrir commented 2 years ago

hi, this is the new video with bilalibrir_2.zip code: https://user-images.githubusercontent.com/86444589/152661399-a73c2560-7063-49d3-bb43-34a522de4425.mp4

thank you for your time and help.

board707 commented 2 years ago

try this bilalibrir_3.zip

bilalibrir commented 2 years ago

Morning @board707 , it works perfectly the pattern is good. from top to bottom, from left to right. perfect. i tried the test example, it works to perfectly. great i will try tonight (DISPLAYS_ACROSS 21) and (DISPLAYS_ACROSS 41). and report back. and maybe a 2*2 zigzag.

Thank you for your time and help, I really appreciate it.

board707 commented 2 years ago

i will try tonight (DISPLAYS_ACROSS 2_1) and (DISPLAYS_ACROSS 4_1). and report back. and maybe a 2*2 zigzag.

Sorry, I'm afraid this fix will work only with one matrix, connecting several panels with non-standard pattern is a separate complex task. I will try to finish the code if you are ready to test many times.

I really appreciate it.

Thanks ^) you can add a star to the repo if you like it

bilalibrir commented 2 years ago

Hi again, I am always ready to help, for as many times it takes. 1000 time's No problem. With an open heart. The only thing is than i have to work during the day, so i'm free every evening starting from 8 PM . And i don't work on the Friday's. Otherwise I'm all ready to help. Again thank you for your help and thank you for your time.

board707 commented 2 years ago

bilalibrir, Thank you for your help.

i have to work during the day, so i'm free every evening starting from 8 PM .

8 PM in what time zone? ^) I am in Moscow, Russia at GMT+3

bilalibrir commented 2 years ago

Sorry, i'm from Algeria +1GMT.

bilalibrir commented 2 years ago

hello, hope you are doing fine @board707, do you have any updates?

board707 commented 2 years ago

hello, thanks, I am fine Have you tried connecting multiple matrices? What is the result?

bilalibrir commented 2 years ago

hello, yes I did, it did not work, i will upload a video.

bilalibrir commented 2 years ago

https://user-images.githubusercontent.com/86444589/152704442-eba992c8-eadc-456b-93a5-8fb84562d88e.mp4

https://user-images.githubusercontent.com/86444589/152704496-61d8c98e-baf8-4be0-a226-e993975a597a.mp4

board707 commented 2 years ago

Thanks for video. For now I don't have a solution I have to think....

bilalibrir commented 2 years ago

sorry for taking so long to upload the videos. i can't upload more then 10Mb / file, so i had to use a video convertor. take your time and thanks for your help.

board707 commented 2 years ago

Hi, Here is a new try I'm not able to check this file with compiler right now, sorry for possible typos bilalibrir_4.zip

bilalibrir commented 2 years ago

Hello , hope you are fine, I think that I found a bug in the RGB text example , i will make a video to show you, i'm not sure if it's just with my panel or for every one, you have to adjust the time to more then 200ms so you can see it. And i will test the new code the soonest when i get home. Thank you. Stay safe.

bilalibrir commented 2 years ago

i tested the code, it compil's with no problem. it's very very close. i did a test for 21 , 12 and 2*2 .

this is 2*1: https://user-images.githubusercontent.com/86444589/152866102-59a87c38-800b-40b3-867e-349c1448d73e.mp4

bilalibrir commented 2 years ago

this is 2*1:

https://user-images.githubusercontent.com/86444589/152870040-d9aacfc5-2243-4b1a-bdf8-4d58f5922d69.mp4

and i don't have a long data cable to do a 2*2 in the correct way. sorry .

thank you for your time. and help.

board707 commented 2 years ago

Thank you for tests Please, comment the whole procedure transform_XY() at template of your matrix (near 760-770 line of the DMD_RGB.h) and show the video Matrix connecting of 2*1 is enough

bilalibrir commented 2 years ago

1*2:

https://user-images.githubusercontent.com/86444589/152873335-501f2da3-8c4f-4dd3-a75a-317a90a43b01.mp4

and 2*1:

https://user-images.githubusercontent.com/86444589/152873931-920629b5-d363-4c2f-ac20-cd72fdf0fb17.mp4

board707 commented 2 years ago

Thanks, I will think :)

bilalibrir commented 2 years ago

this is a test for 2*2, sorry for the short cable: https://user-images.githubusercontent.com/86444589/152876181-94223e98-7a35-4262-9152-669e8d80261e.mp4

take your time, thank you.

board707 commented 2 years ago

Good evening @bilalibrir Try this, please (i hope it will work with horizontal sets as 2x1, 3x1 ...) bilalibrir_6.zip

bilalibrir commented 2 years ago

hello, sorry for the delay i just get home and tested the new code it work's perfectly, for 11, 21, 31 and 41 all work's perfecty.

bilalibrir commented 2 years ago

hello again , about the issues i noticed in the RGB example this is the video, and i thinks it's not the library code it's just the code example.

https://user-images.githubusercontent.com/86444589/153070682-e7507a1d-ea21-4e1b-b60d-8eacdb370c50.mp4

https://user-images.githubusercontent.com/86444589/153070791-80323de9-6525-45d4-8c17-66c5069d2398.mp4

board707 commented 2 years ago

tested the new code it work's perfectly, for 1_1, 2_1, 3_1 and 4_1 all work's perfecty.

O, great. Please, try to test with vertical panel sets : 1x2

About RGB test issue - it is not a issue, this is test of dual buffering. Change

define ENABLE_DUAL_BUFFER false

to true and see/

Thanks, Dmitry

bilalibrir commented 2 years ago

it didn't work for the vertical panel setup, sorry i will make a video for that. for the RGB test code , the dual_buffer is already 'false' and you only see it when (case 0) is runing:

board707 commented 2 years ago

sorry for too much tests... I attached new one: bilalibrir_7.zip

the dual_buffer is already 'false' and you only see it when (case 0) is runing:

try to set it 'true'

bilalibrir commented 2 years ago

https://github.com/board707/DMD_STM32/files/8027859/bilalibrir_7.zip works fine for the 21. not for 22. this is the video for 2*2

https://user-images.githubusercontent.com/86444589/153085694-d4d5ca6b-67c9-4a23-9ea9-56008de83c84.mp4

: you don't need to say sorry , i have no problem if it take's a 1000 times, lol thank you for your time. iknow that it's really hard to make code without hardware next to you.

bilalibrir commented 2 years ago

i have tested the " #define ENABLE_DUAL_BUFFER true ", even worst the 6 first rows alwas off , and this is just for case0. everything else is okay for case1 and case2 ....

board707 commented 2 years ago

i have tested the " #define ENABLE_DUAL_BUFFER true ", even worst the 6 first rows alwas off

that's how it should be. Look at the code, the text scrolling, but the first 6 rows are erased. Without double buffering, you see artifacts in this place, with double buffering everything is clean

Thanks for your help, see you tomorrow

bilalibrir commented 2 years ago

thanks for your help, see you tomorrow

board707 commented 2 years ago

Next code bilalibrir_8.zip

bilalibrir commented 2 years ago

I just tested the 22 and 14 (vertical) configuration and it didn't work, sorry. but it works perfect for horizontal 41 configuration this video is for 22 configuration: https://user-images.githubusercontent.com/86444589/153275092-09d46945-8863-4ea8-946f-49d364b08086.mp4

for the RGB test code i find that if i modify the line 181:

` if (step[i] != 1) dmd.drawFilledBox(0, 0, 5, dmd.height() - 1, GRAPHICS_INVERSE);

to ` if (step[i] != 1) dmd.drawFilledBox(0, 0, -1, dmd.height() - 1, GRAPHICS_INVERSE); /// solved the problem

` solve the problem for the 6 first flickering rows. thank you.

board707 commented 2 years ago

@bilalibrir, thank you how this panels are connected? Do you make any changes in test script?

bilalibrir commented 2 years ago

for the last video it 2 x 2 configuration , the connection is 1 and 2 the two upper one's , 3 and 4 the for bottom one's. sorry for short cable between 2 and 3 panel. ( i jest made a new long cable) yipiiii. lol.
I did not change the test code. only this lines (#define DISPLAYS_ACROSS 2 #define DISPLAYS_DOWN 2 ).
is that really a bug in the RGB test code?? or just for me? i don't understand!!

https://user-images.githubusercontent.com/86444589/153275092-09d46945-8863-4ea8-946f-49d364b08086.mp4

board707 / DMD_STM32

issues with 1/4 scan for rgb panel #9

define set_mux_ch_by_mask(x) ((uint32_t) x)

define clr_mux_ch_by_mask(x) (((uint32_t)x) << 16)

define ENABLE_DUAL_BUFFER false