DS90UB941AS-Q1: Asymmetric Splitting function

If the posted "LCD1 porch timing" can not work with LCD2, please give us a porch timing for us to take request to supplier

Hi! 

Thank you for your question!

I have taken a look at the information you have provided. In regards to timing, the superframe must include two identical size images. To do this, the 941 crops the smaller image and puts the extra pixels into the blanking of the smaller displays.

Can you please list out the minimum and maximum of each hsync, hbp, hfp, vsync, vbp, vfp for each display? 

Respectfully,

William

the current hsync value can not match it, we have to ask supplyer to ajust a porch value  for us. so I should suggest a porch value for the supplier.

the image include current timing porch max,min value.

if adjust LCD1 to this one can it work with LCD2?

William Yi said:

I have taken a look at the information you have provided. In regards to timing, the superframe must include two identical size images. To do this, the 941 crops the smaller image and puts the extra pixels into the blanking of the smaller displays.

so the smaller display need a bigger hbp,hfp, vbp, vfp value than the bigger display for croping ?

how about the hsync and vsync , if the two display should have save vsync value?  

Hi!

1. Yes, the smaller display needs a greater blanking state (hbp, hfp, vbp, vfp, vsync width, and hsync width) to support the extra pixels. Please see this diagram from the datasheet.
   
2. There are no timing requirements that require the same vsync width or hsync width values for our devices. The frame rate must be the same for the superframe and all output images. 
   
   
3. Your Hactive and Vactive for the LCD2 display are now 1268 and 360.
   
   Using the timing table that you have just provided for LCD1:
   Horizontal Maximum total + blanking: 1268+352+432+872 = 2924. This 2924 > 1268.
   Vertical Maximum total + blanking: 360+8+8+80 = 456. This 456 > 360.
   We would be able to support the display that you have now provided. 
   
   Are you sure this is the correct parameters for the LCD2 display? These parameters are different than the ones that you had initially provided 

Respectfully.

William

William Yi said:

There are no timing requirements that require the same vsync width or hsync width values for our devices. The frame rate must be the same for the superframe and all output images. 

If the splitted Vsync of two display can be different , how to set the value of vsync for two displays? there is not vsync setting register for port0 and port1? but there are hsync setting registers( . IMG_HSYNC_CTL1_IMG_HSYNC_CTL1_P1 )

William Yi said:

Are you sure this is the correct parameters for the LCD2 display? These parameters are different than the ones that you had initially provided 

the new provided timing table is set for changing LCD1's orgin one to the new one.

we are assessing orgin LCD2 timing display on 941port0 and the new provieded one on 941 port1.

I list them in this thread. can these two timing work together in 941 two port?

Hi,

Thank you for clearing up that confusion about your displays

In your case, your Hactive and Vactive for the LCD2 display are now 1920 and 720.

Using the timing table that you have just provided for LCD1:
Horizontal Maximum total + blanking: 1268+33+695+12= 2924. This 2008 >= 1920.
Vertical Maximum total + blanking: 360+21+376+3 = 456. This 760 is >= to 720.
We would be able to support the display that you have now provided.

In the 941AS after splitting, note that each video of each display would be half the frequency of the superframe. In addition, any horizontal blanking on the superframe image is divided by two for each individual image. Vertical blanking (VFP, VSYNC, VBP) are all the same number of lines between the two split images.

If you needed to, you can adjust the horizontal blanking to be divided into HFP, HSYNC, HBP. Those are the IMG_HSYNC_CTL registers. While the total number of pixels stay the same, you can use those registers to help you split into HFP, HSYNC, HBP to the way that you need it. 

Respectfully,

William

William Yi said:

In the 941AS after splitting, note that each video of each display would be half the frequency of the superframe. In addition, any horizontal blanking on the superframe image is divided by two for each individual image. Vertical blanking (VFP, VSYNC, VBP) are all the same number of lines between the two split images.

If you needed to, you can adjust the horizontal blanking to be divided into HFP, HSYNC, HBP. Those are the IMG_HSYNC_CTL registers. While the total number of pixels stay the same, you can use those registers to help you split into HFP, HSYNC, HBP to the way that you need it. 

thanks, I think in "Asymmetric Splitting" mode, HBP, HFP, VFP and VBP are also affected by croping reg( CROP_START_X/Y and CROP_STOP_X/Y) 

so after croping only the VSYNC of two display must be the orgin one as origin superframe? Is this true?

William Yi said:

Using the timing table that you have just provided for LCD1:
Horizontal Maximum total + blanking: 1268+33+695+12= 2924. This 2008 >= 1920.
Vertical Maximum total + blanking: 360+21+376+3 = 456. This 760 is >= to 720.
We would be able to support the display that you have now provided.

about htotal and vtotal comparing,  why we should compare htotal of LCD1 to hactvie of LCD2? I think in the working state, htotal and vtotal of two display must be exactly the same. is this opinion right?