-
Notifications
You must be signed in to change notification settings - Fork 190
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
P5 RVB 64x32 1/8 scan with ICN2038S controller... and mess! #500
Comments
Every panel contains at least two types of chips, the ICN2038 is only one from it. Knowing the second chip type is required to handle the matrix properly. Please try to read the headings in the other chips on the panel.
Is the panel works with this config? |
Hi, thanks for your answer. and checking the chips, I see that the ICN2038 is in fact a SM16208S and the other chip is a MW245B. thkx! fabrice |
No, the MW245B is not "the other chip" of the interest, it is just a buffer. Your panel should contains yet another driver. But it is not so important already. According to screenshot of the huidu software configuration screen, the panel has a standard drivers, supported by the library. All what you need - correct transform of the pixel coordinates. I'm sorry, but I can't offer you more useful advice, since I have no experience working with complex matrices in this library |
The best bet is to simply make a script that draws a pixel to y coordinate zero, x coordinate 0, 1, 2 etc, and then repeat the same for Y coordinate one, X coordinate 0, one, 2 etc and etc. From this you might be able to reverse engineer the pixel mapping approach to the panel, there's so many permutations of these panels now I don't bother even coding for them anymore, as long as it works electrically, it's a waste of time. |
Thank you very much for your answers, If I manage to understand the panel permutations, where in the code could I implement that specificity? thank you, |
In most cases the mapping is not more than two-three lines of code. |
@veryslow |
Thank you very much for your answers, I run this code (you know it!) with a pause 1000ms tu understand what happens with the pixels inrelation with the coordinates... led01.mp4
|
and here it is at normal speed (with a 10ms pause between each iteration). 20230916_105242_720p.mp4 |
You do it in incorrect order, the first (outer) loop should be by Y direction and the inner loop by X. According to the video your problem is not (or not only) with geometry. You have a problem with multiplexing - either with your ABC connections or with multiplexor chip. I have to ask you again to find the ID of all the chips on the matrix. (Excuse my English, I don't have a translator on the phone) |
Thank you for uploading the new video. But the conclusion still the same - you have a multiplexing problem. Look to the picture - the panel is filled in by two lines with two lines gaps between it. It looks like the C pin is not worked or not connected. |
Thank you! I did this mistake while soldering and soldering wires again and again! If I want a complete horizontal line, I have to define a 128 wide panel forr 16 high... 20230916_165916_edit.mp4 |
Perfect ! |
Yes! Can yout tell me where int the library I can tune this mapping? With an adaptated code, I get a full-filled screen, with an organisation of pixels like in the sreen capture I posted before... 20230916_172613_edit.mp420230916_172754_edit.mp4
|
This is a question to the author, @mrfaptastic |
Hi, after a lot of tests, I finally switched to hzeller library for Raspberry pi... here is the code for mapping in hzeller's library (If it can help anyone with thi s kinf of panels)
I'm still interested in the way I could implement this code in ESP32-HUB75-MatrixPanel-DMA library! Thanks for all your answers, and I come back in this issue if I find a way to do it with ESP32. |
Addressing the matrix as different size is not seems correct for me. Does this transform supports a chaining for more than one panel? If you ready to change your main board to rp2040 or stm32, I could help you with your panels. |
Yes, no problem with chaining after remapping! |
Perfect |
Look at the Chained Panels example (Virtual Matrix Panel) for examples. These panels seem to act like thinner chained panels within....can use the Virtual Matrix Panel class to either use an existing mapping that might work or add another option. |
Hi,
I'm fabrice and trying to deal with panels that I received from eagerled.
They are P5 RVB 64x32 1/8 scan with ICN2038S controller, and seem to act as double 64x16 panels...
I did not notice that they were 1/8 scan and I've been digging for 2 weeks with an ESP32 WROOM trying every combinations in the library, changing pins, adjusting variables without any success yet.
The panels are functionnal, I have a HD-D16 asynchronous controller from HUIDU that works well and gives me technical datas on the panel structure (see screen captures below). Nevertheless, I cannot manage to show any line without messy blocks...
I've been following this feed, with no success...
I'm an amateur programmer with some skills but the library is a bit to complex for me to manage and modify it for my needs.
I also wonder if my pin configuration is ok, considering some pin limits in the ESP32 (I tried different ones based on default pins configurations from library), the actuel one is :
#define A_PIN 23
#define B_PIN 19
#define C_PIN 5
#define D_PIN -1
#define E_PIN -1
#define R1_PIN 25
#define R2_PIN 14
#define G1_PIN 26
#define G2_PIN 12
#define B1_PIN 27
#define B2_PIN 13
#define CLK_PIN 16
#define LAT_PIN 4
#define OE_PIN 15
here are pictures frome the huidu configuration software that may help and 4 pictures of the panel itself.
I plan to create A 6 panels linear display to make scrolling texts updated by wifi for a theatre company (begat.org).
Thanks by advance of any help, I've been almost pulling all my hair!
The text was updated successfully, but these errors were encountered: