Tips on trouble shooting a 3V LCD using a NT7603 Controller?

JimB · Posted: Sat Apr 22, 2006 2:21 am

I am trying to get a Densitron, DV3002B2C16SF-3, 2 x 16 display up and running. This 3V display uses a NOVATEK NT7503 controller, which they have assured me has the same interface as the HD44780, used in the CCS displays.
I have my own board setup to use port D, same as the development board. The interface to this LCD is the same 14 pin, as the CCS interface. I have setup a test program that works on the 5V dev board from CCS. However, when I run it on mine, with the 3V LCD, it appears to hang in the lcd_init() routine. (Watched the stack in the debugger) Everything else on my own board is functioning correctly. This display is the last thing to get up and running.
It appears from what I find here, that timing could be the problem. I assume that this means the times called for in the lcd.c driver file.
I could use some suggestions on how to approach solving this problem.

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

JimB · Posted: Sun Apr 23, 2006 2:12 am

After looking around the forum I ran across this same change to the 30ms delay solving a problem. I have an RS232 readout to help me verify things. I changed the delay to 30ms and at first there was nothing noticable, but as I varied the contrast voltage to the LCD, the RS232 output ran when the pot was near one end or the other, but stopped when near the center of the range. Don't understand that quite, but will do the additional increase in the AM.

Guest · Posted: Mon Apr 24, 2006 9:07 pm

I changed the delay in the lcd_init() and kept increasing it until it was 60ms. But no luck. I still have the problem that when the contrast voltage is varied the program will hang, except when the voltage is below 1.8 v and then again over 3.0 v. Inbetween the routine hangs. I have an RS232 port and its output is my indicator whether or not the program is looping through.
I have sent a query to Densitron, but haven't heard from them yet.

Jim

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

Are you sure you're really adjusting the contrast voltage ?
Is it possible that you're adjusting the Vdd voltage on the LCD ?

JimB · Posted: Tue Apr 25, 2006 2:04 am

Is my face red. I've checked that board so many times.......I had to look twice tonight.

That is exactly the problem. Had to go back to the board layout but found that you hit it right on. Had pins 2 and 3 swapped.

Will make the mod tomorrow and give it a whirl.

Thanks again,
Jim

JimB · Posted: Thu Apr 27, 2006 2:46 am

Made the changes to the circuitry to put the Vdd and contrast on the proper pins. Changed delay to 30ms in lcd_init. I have two of these displays to work with. I got one to display the first 8 characters of the first line, but nothing on the last 8 nor the second line. This one worked through a couple of on-off power to the board cycles. And then nothing anymore nor anything ever on the other display.
Do you think that there are more timing issues or could I have damaged them when reversing the to pins. I wouldn't think so since one ultimately had some display output. Also my RS232 output is now running, so I must be getting through the lcd_init successfully.
There were many warnings about ESD damage, but I live near the beach and it is usually humid enough to not create static problems.
Any more ideas to try or timing issues to look into?

JimB

drh · Posted: Thu Apr 27, 2006 7:27 am

If you are using a 3V display, what voltage is your board? If it is 5V, do you have any level translation between to two devices?
_________________
David

JimB · Posted: Thu Apr 27, 2006 10:13 am

The entire board is operating from a 3.3V Lithium cell. According to the spec, the display will function down to 3.1V and the internal chip will work all the way down to 1.8V, or so. I was surprised to see that the contrast voltage was near 0V when I did get the display to show something in the first 8 characters, but even the CCS display module, when connected to 5V, the contrast setting is just a little over 1V.

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

The latest NT7603 data sheet is available here:
http://www.andilcd.de/de/download/controller/andilcd_conchar_sheet_de_nt7603.pdf

On page 21, it shows the Application Circuit diagram. It shows that V5
is the contrast pin.

On page 6, right below the section on Absolute Maximum Ratings,
it says that Vdd must always be greater than or equal to V5.

But you connected V5 directly to the +3.3v supply, and you put
a trimpot on Vdd, and at one point you turned it down to 0v.
So it's likely that the LCD units are damaged. I would buy new LCDs.

JimB · Posted: Thu Apr 27, 2006 1:46 pm

I don't have the data sheeet for the DV3002.... here at my day job, but it is designed to operate from a 3.3V battery (3.1V min for the LCD). I believe it may be a custom design just for this job. The NT7603 is buried inside the display module with the resulting interface, being the same 14 pin interface as the one provided by CCS.

I have never connected this display to 5V, just the battery operated board. They have been kept in ESD bags. I did have the contrast and Vdd swapped and one other pin open. So the interface to the controller is internal. The data sheet they sent me for the NT7603, shows that it operates down to 2.8V.

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

You're assuming there is some external buffering of the signals on
the NT7603 controller before they go to the LCD's connector.
I strongly doubt this.

This page shows a photograph of the DV3002 LCD.
http://news.thomasnet.com/images/large/008/8243.jpg
The NT7603 data sheet says the chip is available in COG form.
(Chip on Glass).

I have a graphic LCD constructed in a similar way, with a flexible
flat cable like that one. If I peel back the black tape at the base
of the LCD, I can see a thin sliver of shiny material, which is the
controller I.C. It's embedded in black epoxy. There is no
additional circuitry. The COG chip is all there is.