LCD_flex 2X16 (JHD162 A)

[kamillas]

I made a circuit with a 2x16 LCD display, but in the simulation it works well, ie, it displays the word "temp", but in achieving it displays square! I have not resistors connected, and I do not think that's his problem, I connected this way,
/ / D0 enable
/ / D1 rs
/ / D2 rw
/ / D4 D4
/ / D5 D5
/ / D6 D6
/ / D7 D7

As the driver of LCD indicates, knowing that I have not connected DB0 DB1 DB2 DB3.
Where is the problem please !!!

[Ttelmah]

First, have you verified you chip is actually running, and running at the speed you expect?.
Does A0 go high, and low again, and is it high for 0.5 second.
If not:

1) is 4000000 > 4MHz - is HS the right fuse?.
2) Is the MCLR pin pulled high?.
3) Is power connected to the PIC?.
It is almost impossible to tell on your pictures (too blurred), but I cannot see a connection to pin 1 of the chip, in which case MCLR is floating, and the chip won't run. Also I can't see the loading capacitors on the crystal, or a decoupling capacitor right by the PIC supply connections. I can't even see the supply connections to the PIC. There looks to possibly be one connection to pin 12, with a long red wire, but no visible connection to pins 11 31, or 32. Red wire to pin 12, is 'puzzling' since this should be the 0v connection.
I'd suspect you have taken the design from Proteus, and have not realised that Proteus _automatically_ makes the supply connections for you. If there is a rail called '5v', or Vdd in the design, it will connect this for you to the Vdd pins on the PIC, and not show these connections in the drawings. Similarly if there is a 0v or Vss line, this is connected to Vss automatically. Also Proteus treats the MCLR input as a TTL input (which defaults to floating 'high'), while in fact it is a CMOS input, with no such default.
Two examples of why people don't like Proteus.....

As a comment, don't keep initialising the LCD in the loop. This wants to be done _once_.

Best Wishes

[kamillas]

[kamillas]

I find in another forum the solution to this problem, but where is the solution (mode 4 bit...)? In the driver lcd_flex,




http://forums.futura-sciences.com/electronique/474012-affichage-lcd-un-p18f4550.html

[kamillas]

Although I used the driver flex_lcd, applied here in the library, and I've connected to pins correctly, and I change with another display lcd2x16, the program works. The microcontroller executes the instruction, the evidence I've tested all the pins connected with lcd, are identical with the simulation, it shows why the tiles!

[PCM programmer]

[kamillas]

[PCM programmer]

Post a schematic of your connections to the LCD. Here's an example:
http://www.extremeelectronics.co.in/pictutorials/images/pic_lcd_interface_schematic.gif

[Mike Walne]

You're making it very difficult for us, again. This thread's going nowhere fast.

We need ACCURATE & CONCISE information including a schematic.

You're not telling us what your V0 voltage is.

You need to be VERY VERY careful with LCD modules.

I got caught out with the 1602 series parts, even etched a PCB wrong.

They are NOT all the SAME.

They may LOOk the same, BUT have DIFFERENT pinouts & V0 requirements. Start by verifying the part No.

You have to be VERY careful, double check, then check AGAIN, AGAIN, and AGAIN......

On a positive note:-

I applaud your use of white boards. I think they are great. Your layout, however, is ATROCIOUS. You need to keep ALL connections SHORT, NEAT & TIDY, just as you should on a PCB.

AND you MUST pay particular attention to decoupling caps and crystal circuits (they go in first, everything else fits around them).

I only learn ONE thing from your .jpg; the appearance of your LCD. NOTHING else is of value.

Mike

[kamillas]

[PCM programmer]

Do some experiments. Edit the lcd_init() function in flex_lcd.c and
increase the delay times by a lot.

1. For example, increase first delay of 15 ms so it is 100 ms. Yes, a lot.

2. Then, in the for() loop, increase the delay of 5 ms so it is 25 ms.

3. There is one more delay near the end of the function for 5 ms. Make it 25 ms.

If that doesn't fix the problem, then edit all the delay_cycles() statements
and change them to delay_us(10).

These are big delays, much more than should be needed, but as a test
it's worth doing, to see if it fixes the problem.

[masterat]

Try to insert 20ms delay time before initLcd may be done.
Some LCD working slower than pic MCU.

[ezflyr]

Hi,

Is that really how your crystal is wired? Because as shown it's wrong!

Can you blink an LED at a predictable rate?

John

[Mike Walne]

This topic, or similar, surfaces at regular intervals.

Do a search for either 'square' or 'squares'.

You'll get loads of results, including an odd one less than a week old!

Mike

[temtronic]

Like others trying to help. here's a few things I've learned during the past 20 years of using PICs.

1) white boards are fine to use,I've got 16F877s running 20MHz no problem for the past 20 years....18F4550s at 48MHz!

2) buy some 10 conductor, 22ga solid wire cable.Cut off the sheath, and use the 10 wires for neat and tidy interconnections.You'll have all 10 colours !

3) when wiring try to match wire colour to PIC pin( PIC pin 37 =purple, pin 36 = blue,etc.) Always use Black for ground, Red for V+.

4) If possible for the LCD wiring ,use PIC pins = to LCD pins( ie, RB7 -> LCD db7, RB6 -> LCD db6) This is easy to remember, easy to follow.

5) for my testbed, I use 2 whiteboards. the lower one has the 40 pin PIC(16F877, 18F4550) mounted in the far left postion.It's easy to remove for programming. A 16x2 LCD is mounted in the far right postion,this allows easy,clean wiring of LCD pins to PIC. The upper whiteboard has 'peripherals' like MAX232,RTC,sensors,LEDs,opamps,etc. This way it's very easy to rewire the project to use different PIC pins for various functions.

6) all wiring should be neat and tidy.Always add 100mfd filter caps to each whiteboard on the +5, gnd bus lines.