LCD problem

[Night Rider]

Hi guys!
I have a problem and i can't understand what kind of problem is that. I made a power supply for my bench. I made two channels 0-30V and one channel -30V-0V. Also i put 0.1ohm resistors to measure the current of all three channels. Then i decided to display the voltages and the currents via 2x20 LCD controlled by a PIC16F877. I transformed the voltages and the currents with op amplifiers to bring them at 5V maximum at the six analog pins of the uC. I wrote the program, i burned the pic, i turned the switch on and the problem is that the LCD displays nothing except if i put my fingers on the pins DB4, DB5, DB6, DB7. Only then the LCD displays and working as i programmed the pic. I drive the LCD with four pins data and not with the eight. 4Mhz clock for the pic. It is the first time i saw this kind of problem at pic-LCD's. Is the problem to the connector? Cold soldering perhaps?
Please help!

 

[ericgibbs]

Do you have the LCD's 'contrast' pin connected.?

 

[Night Rider]

Yes, it's all ok. I played with the contrast and the light.

 

[Pommie]

Disconnect everything except the pic, the LCD and a 5V supply. Then try to get it to display "Hello World!". If it doesn't work then post your code and a schematic and we should be able to help you get it working. Then you can go on to other things.

Mike.

 

[Night Rider]

Ok guys, thanks. I'll search it more and i'll tell you the results.

 

[ericgibbs]

Yes, it's all ok. I played with the contrast and the light.

I am assuming that the DB7 thru DB4 are the pins that are wired and that DB3 thru DB0 are floating.

You say you can change the contrast of the LCD pixels using the pot OK.

When you hold your fingers on the pins it adds capacitance and the LCD pulse edges are degraded, that is, the rise/fall times are increased.

Can you increase the LCD 'enable line' time periods, [within the program] as a test,?

 

[Night Rider]

Ok, i made a lot of projects with this type of program and are working fine. I'll test this. I played a little with the voltage and current at the contrast and at the supply and worked. I'll try to make it perfect.
Thanks!

 

[Night Rider]

I built the circuit on a raster and it works perfect. So, the problem i supposed is at the pcb. But i can't believe i'm telling this, because there is no shortcircuit, there is no opencircuit after a detailed test. Also i observed that when i'm placing far the crystal with the two capacitors, the oscilator circuit, it failed for the most times(at the raster). But at the pcb the oscilator circuit is only 2cm far from the uC. There is nothing else connected to the uC. Here is a foto of the pcb. The two wires are for 5V LCD supply. Rather i'll make another pcb to test it. Thanks all for your thoughts.

 

[Night Rider]

Here are the fotos

 

[mvs sarma]

please put a 0.1uF MLC cap across Vdd and Vss close to the PIC chip. I didn't fine one. perhaps this might help to normalize

 

[blueroomelectronics]

It doesn't appear all the VDD & VSS pins are connected. You need to connect them all.

 

[mvs sarma]

Congrats Bill,You got it,


Pins 11 and 12 not used .

 

[blueroomelectronics]

Congrats Bill,You got it,


Pins 11 and 12 not used .

Actually both answers yours and mine are essential. Was there any large filter cap on the board? That'll help too.

 

[mvs sarma]

Actually both answers yours and mine are essential. Was there any large filter cap on the board? That'll help too.

Many a time, people are pushing projects without any filter components. Few examples are very accurate LC meter from www.electronics-diy.com and few projects from www.josepino.com like LCD clock cum calender. Most of these don't contain needed Vdd line filters.

 

[blueroomelectronics]

The 0.1uf caps are important enough you could buy IC sockets with 0.1uf caps built in for TTL ICs.
I wish simulators would complain if you didn't have them.

 

[Night Rider]

Pins 11 and 12 are internally connected to other Vdd and Vss. The pic is working without them. 0.1uF didn't solve the problem. I'll make another pcb and i'll tell you the results. I think i tired you. Sorry, but something strange is happening.

 

[Nigel Goodwin]

Pins 11 and 12 are internally connected to other Vdd and Vss. The pic is working without them. 0.1uF didn't solve the problem. I'll make another pcb and i'll tell you the results. I think i tired you. Sorry, but something strange is happening.

They don't add extra pins for no reason, connect ALL Vdd and Vss pins, and always add decoupling capacitors to all circuits.

You might have a problem elsewhere?, but not connecting required pins isn't a good start.

 

[Night Rider]

I'll tell you what i made. I cut the lines from pic to lcd. The rs, rw, e, and the 4 data lines. I connect this signals to lcd with wires. Without capacitor and without connected the other Vdd and Vss. The lcd worked perfect. Nigel perhaps you have right about the Vdd and Vss, i'll try this to future applications, but until now in all my projects i didn't connect the extra supply and all projects are working fine. Final, i suppose that the high frequency and the small distance between the signal lines makes the pcb problematic.

 

[tunedwolf]

There is so much burnt flux on the board I would say it's the likeliest culprit. It might even be conductive, if not an actual short. The other possibility is that you have removed the board from the etchant too early, and it is actually still conductive in small spots. Try using a little less heat that way you will not scorch the laminate, and water soluble, flux cored solder. Thoroughly scrub your boards after soldering in warm water to remove all traces of flux, dry the board completely, then very carefully inspect all joints for shorts, splashes, and cold joints. Make sure discreet components like diodes, capacitors transistors etc are oriented the right way round. Only then power the board up and measure your supplies and if everything looks fine on the volt meter, then power down and fit your chips etc. Best way to avoid letting out the magic smoke...

rgds

 

[Night Rider]

Thanks to all guys. You are particularly instructive!