AVR Freaks

Whats that?

I do not suggest sinking the LED backlight through the AVR pins.  I use a MOSFET or NPN transistor as a buffer.

Jim

I see what you did there when translating your native language to English - the same mis-translation can easily be done from Swedish. The English word to use is not "tension" but "Voltage".

Besides, if your power supply drops then it takes a special regulator to raise the voltage, and even those can not cope if the input to them drops too low.

I am troubled by the fact that you supply the back-light LED on the LCD module from an I/O pin on the AVR.

What does the data sheet for your LCD module say about current consumption of the back-light? (Hint: I have seen numbers as high as about 250 mA.)

What does the ATmega88 data sheet say about maximum current you may source from an I/O pin? (Hint: 40 mA.)

What does the ATmega88 data sheet say about maximum current for supply (VCC) and GND pins? (Hint: 200  mA.)

If you want to be able to switch the back-light on/off by firmware in the AVR, then let the I/O pin control a BJT or FET that in turn switches the supply to the LED back-light on/off. Ask the electronics gurus here about a good circuit for this.

Chances are you fried PC2 of your ATmega88 already..

Why i need AVCC?

 that powers half of the chip.

R e a d   t h e   d a t a   s h e e t. [1]

Spoiler: Some parts of the ATmega88 internal circuitry draws its supply voltage from AVCC rather than VCC. It is spelled out in the data sheet which parts that is.

I am not sure your problem with the LED is associated with your wiring up of the LCD back-light. I was merely pointing out that the wire-up of the back-light

i) may result in a current draw larger than the allowed absolute maximums (for the specific pin, or for the ATmega chip as a whole), and

ii) that this might damage parts  or all of your ATmega88, and

iii) that this might already have happened.

If the supply going in to the 7805 is dropping then the 7805 cant do anything about that. It is not constructed in such a way that it can raise that voltage. A 7805 can only lower a voltage higher than 5 Volts down to 5 Volts. (In fact, the supply going into the 7805 has to be substantially higher that 5 Volts in order for it to do what it is designed to. 6 Volts or so, if I recall correctly.) Again, you need to read the data sheet - for the 7805 in this case.

Your electronic design seems to be much about "shooting from the hip" and possibly based on assumptions rather than facts. Slow down. Read data sheets. Do a design. Post in the "General Electronics" sub-forum here for constructive criticism.

Test my hypothesis about the ATmega88 being fried. Pull the '88 out of it's socket. Insert a fresh one. Do not connect the LCD back-light! Program the '88 and test if things are working.

Yes: ATmega88 was probably fried. No: Can't rule out that ATmega88 was fried, but there is at least one other problem somewhere.

Further investigations then could be e.g.

• Remove peripherals parts, e.g. everything but the LED and see if it lights up brighter, remove everything but the LCD and see if it starts working etc..
• Measure the current through the LED with your DMM. Is it what the LED needs in order to light up?
• Measure current consumption for the whole solution e.g. with a DMM. Again, don't connect the LCD back-light! If the current consumption is high (>100 mA?) then remove parts external to the '88, one by one, and see if the current drops substantially. Report findings here.
• How about, after reading about it in the data sheet, connecting AVCC? ;-)

And of course:

• Don't think (i.e. assume) that it's not the power supply. Measure the voltage.

Fault seeking is about gathering facts. Guessing will not get you far.

[1] Sometimes I wish we could get blinking text in posts here at AVRfreaks ;-)

Also: If the thing that was working yesterday was wired up exactly as in your diagram above with AVCC not connected, then PC0..3 must have been supplied by stray/leak currents internal to the ATmega88. This situation actually makes it more likely that some parts of the '88 is damaged ("fried"). When you now have connected AVCC those pins get the supply from where they should get it, but it does not rule out that some parts of the '88 actually are damaged.

Keep this in memory. If you use the '88 for other purposes in the future and it does not behave like you think it should (after reading the documentation ;-) then the hypothesis that it is damaged is one of the first to test. In fact this holds tru even with the current wire -up if it starts to misbehave in the future.

And my firm advice to design a better supply for the back-light still holds. I repeat: Maximum current for a single I/O pin on the '88: 40 mA. Maximum for '88 as a whole: 200 mA. Most back-lights will draw more than 20 mA, some will exceed 200 mA by a good margin.

You have been warned. Twice. Happy chip-frying! ;-)

Ha ha - another good example that just because something appears to "work" does not mean that it is correct!

#ProvenProductSyndrome

Joao,

Firstly, is the AB9V battery like this one?

If yes, these are very weak batteries and perhaps your LCD backlight has drained it flat.

Secondly, you should use a resistor between your LEDA and PC2 connections. This will limit/control the current to the LCD backlight. Without a resistor, your mega88 will attempt to supply a large current and cause a failure mode which might include damaging your mega88 and/or pulling too much current from the battery so that it drops from 9 volts to something much lower. That in turn will cause the 7805 to stop supplying 5 volts... a much lower voltage will be produced and then your mega88 will possibly stop working and your LCD will almost certainly go blank.

Hope that helps you.

Ross