Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

PIC18F and LVDS Display

Status
Not open for further replies.

dieplz

New Member
I am trying to use a PIC18F to drive a LVDS display (**broken link removed**). I have tried to program the PIC to output the values for HSYNC, VSYNC, and Data Enable as described in the data sheet, and from what I can tell on a scope, have all these timings correct. I am then feeding these signals to a DS90C365 LVDS transmitter that goes to the display.

As for the data inputs on the DS90C365, I am tying all of one color high while the other colors are tied low (for this application I only need to get a single color on the screen).

Unfortunately, the display is not turning on or showing anything. Is there anything else that I could be missing?
 
I don't see how you are going to drive this LCD with a PIC18F!
Where is the clock coming from? 23mhz-28.33mhz
I can see the H and V sync is easy.
How are you going to drive video, later on, when you need to update the data on R(0-7), G(0-7), and B(0-7) every 25.18mhz?
I think this takes an ARM computer with DMA. (or a 650mhz PIC)
If you can find a way I areally want to know.
 
I am using a 25MHZ clock. For now, I want to simply prove out a theoretical concept such as the one that was done here: https://webx.dk/oz2cpu/pic/pictft.htm

Ideally I'd like to just be able to create various different color patterns for the time being before I invest more time or money into the project, so I am not currently concerned with updating the data and have just tied all the data lines low except for one color register trying to get a single color to display in order to prove out my concept.

Is this possible you think?
 
Eh, perhaps I mispoke. Eventually I would like to get it working on the Sharp display, but it has been more picky, as it is actually housed in another unit that has an unknown LVDS receiver and I am unable to open it up. So for the time being, to get my timing right and such, I am trying to use the following display, but it is not working. https://www.electro-tech-online.com/custompdfs/2011/08/AND050VL-LEDPDF.pdf The backlight is controlled by a separate board and I can turn it on or off as I wish, but I am unable to get any image on the screen still. Also, for some reason if I toggle the B4 data bit (and only this bit), it activates the display for a split second. This has behavior been very confusing me to.
 
Now you have two displays that are very different!
The Sharp display has 8x3 lines for RGB.
The AND display has a R pair, G pair, B pair and clk pair of wires.
The DS90C365 is a little hard to make work. Can you get to the 63lvdf64 on the LCD? Is the clk exiting the 63lv64 and at the same frequency you are sending 25mhz? I now think you are not receiving Hsync, Vsync and Eenb.
Is Denb input set low?
The DS90C365 is picky about capacitors on VCC! The grounds must be good.
 
Ron, thanks for all your help so far.

Unfortunately, I can not get to the 63lv64 on the LCD, but I have hooked up my DS90C365 to a DS90CF386 and have successfully seen all signals I input to the DS90C365 on the output of the DS90CF386.

Denb input, I have that set to be a pulse that goes high approximately 1.7us after my HSync goes high and comes down a few us before the HSync comes down.
 
Inside the DS90C365 things are happening 8x faster than the 25mhz. Data is converted from parallel to serial. The serial is fast! (about 200mhz) You need to treat this part as a 200mhz part. (it is) There needs to be good caps on the VCC. I don't have the data sheet up now but 1uf, .01uf and 100pf in parallel. The smaller caps need to be connected with short leads. I see people using white breadboards with no caps. This will not work.

The LVDS pairs whould be using 100 ohm wire, probably twisted. Same length on all the LVDS. (200mhz)
 
Ron, ok, I have the power supply decoupled as best as I can with caps, and I believe that the serialization of the parallel data is working, because like I said before, I can see all my data reconstructed perfectly on the receiver end. I am just having a hard time getting anything even somewhat coherent on the display (the AND050 one for now), but for some reason, the display toggles when I toggle the B4 bit, which seems quite odd to me. Any other input?
 
correction; the sr is 7 bit long not 8.
New idea.
I think you have the clk- and clk+ exchanged.
B4 is (16) and if clock is inverted then it is (19) which is Vsync. So it may act like B4 & Vsync are exchanged. If this is true then Hsync and Enable are mixed up with B? and B?.
Please double check the (-) and (+) wires on clock and maybe data
 
Checked the polarities.... when I swapped them I no longer received anything nor was I able to activate the display at all.
 
OK...I don't know what to do now.
I can see why B4 could get mixed up with sync.
I have used these parts at 600mhz speed (80mhz clock) and great care was taken that all the + and - wires had the same length. That the impediance was correct. And the supplies were good. With out seeing your setup I have no more ideas.
 
Ron,

I was able to get it working this morning and boy do I feel stupid. I was using one setup at my computer to program my pic and look at signals on my scope. When I would connect them to the display, however, I had to go elsewhere as the display is currently mounted. Anyway, the supply I was using at the display was faulty and the on screen indication of its output was not the same as the actual output and my chips were not activating (5V was actually 5.8V, thus not allowing the PIC to turn on, though I am very surprised it didn't damage the PIC).

After adjusting the supply, however, everything worked just as planned. Sorry to waste your time with such a silly mistake. I think from now on I won't rely on power supplies and will use more of my own regulation.
 
Last edited:
Status
Not open for further replies.

Latest threads

Back
Top