Gameduino2 doesn't like to start

[warren]

I have two Gameduino2's and several Arduino platforms that I use with them, and sometimes each of them for some reason fail to start. No screen or even the backlight appears, until I try to power up over and over. Each of them eventually works, but not until after a lot of prodding. Any suggestions?
Warren Davis

[jamesbowman]

Two things to check.

(1) Put a formatted SD card in the slot - if there is no card the GD library can become confused

(2) Try alternate power connections. USB power is fine, but some cables/ports cannot deliver the current needed for the LCD backlight.

[BrewBot]

I have a similar problem that the unit only works when plugged into a laptop usb (arduino mega 2560)

When I use a 12v power supply input, the LCD is blank and dark

The voltage on the 5v input to ground is 4.6v when plugged into laptop and 5.06v when plugged into the 12v supply (not sure if the drop with laptop is because the backlight is on)

There is a white flash when the screen starts when laptop plugged, but no flash with 12v supply.

Previous threads suggest that the spi startup command is not activating, but I am unsure why

Any help would be greatly appreciated

[BrewBot]

Update:

1. An external 12V power supply appears to be causing over-heating issues with the board. The on-board (arduino mega 2560) regulator is dumping too much heat when powered by a 12V supply (12-5v @ 200ma = 1.4 watts of heat dissipated if its all dissipated as heat). I found that be powering it directly with a 5V 1000ma USB power supply through the laptop / programming port to be problem free. So now my external power supply is fed into the programming port instead of the DC power jack.

2. When directly plugging in the 5V (or 12V) power supply, it appears to take a couple of seconds to read stable voltage. This period of dynamic voltage change was messing with the gameduino's ability to startup (not sure if its arduino or gameduino related). What I found was that if you started the power supply, then a couple seconds later plugged in the arduino, the system would startup fine. So to simulate this I got a time delay relay and wired it in, so the system will start about 2 seconds after the power supply starts putting power out.

[Kitty]

Hmmm...just spent a day trying to resolve a similar issue. Starts fine when directly plugged into the Arduino board but won't if't is connected by positive and ground leads as short as 4cm. Have tried separate power supplies, starting one up first then the other, etc, etc....it will shortly be going in the bin.

[jamesbowman]

MicroSD cards are a frequent source of problems, two suggestions:

 * Run "helloworld" or "fizz" as a test application
 * Remove the microSD card form the slot

[Kitty]

I haven't been using a MicroSD card at all in the slot. However, I tried an old one and for a second it seemed as if it was going to work properly but then came up with a 'coprocessor error' (I think). Anyway, I then tried again with a brand new MicroSD card and it didn't start and it didn't get the error.

It's extraordinary how it seems to know when it's directly attached to the power and ground pins on the Arduino and how it immediately fails if for example I use short jumper wires. I'm certain there's no faulty connections anywhere. Same results with Fizz.

[jamesbowman]

Yeah that is weird. We test them using 10cm jumpers for all connections, so I'd fully expect it to run fine on jumpers.

Is this running with a 5V Arduino, USB powered?

[Kitty]

The boards I have tried are an Arduino Uno and an Elegoo MEGA 2560 clone.

The power supplies I have tried are a cheap 1000mA 9 volt wall plug, a USB lead from the PC and a very nice TTi bench power supply.

All the above with the same results.

Using a scope I can see a spike on startup and like someone else I wondered if this was the initialisation of the backlight. When does the backlight come on - when you first power up the display or when the Arduino hits some particular part of the code? I have a suspicion that it's the Arduino that's getting stuck. I have used an LED to see if the Arduino code is running properly and it isn't when the screen fails - it seems to hang.

[jamesbowman]

Right, this could make sense.

The backlight gets turned on inside GD.begin().

At that moment there's a big dI/dt, because the backlight takes a lot of current.

Just a hackish suggestion to isolate the problem: clip a big electrolytic across 5V/GND.

[Kitty]

I have just tried the Electrolytic idea and the scope shows a much cleaner power up but still no joy.

I can see from the code modified version of Fizz below that it is definitely getting stuck at the line "GD.begin(~GD_STORAGE);"

This is really mystifying.



#include <EEPROM.h>
#include <SPI.h>
#include <GD2.h>
int ledPin = 13;

void setup()
{

pinMode(ledPin, OUTPUT);

myLedFlasher(30);

GD.begin(~GD_STORAGE);

myLedFlasher(30);
}

void loop()
{
GD.Clear();
GD.Begin(POINTS);
for (int i = 0; i < 100; i++) {
GD.PointSize(GD.random(2 * GD.w));
GD.ColorRGB(GD.random(), GD.random(), GD.random());
GD.ColorA(GD.random());
GD.Vertex2f(GD.random(PIXELS(GD.w)), GD.random(PIXELS(GD.h)));
}
GD.swap();
}

void myLedFlasher(int t) {
for (int i = 0; i < t; i++) {
digitalWrite(ledPin, HIGH);
delay(50);
digitalWrite(ledPin, LOW);
delay(50);
}
}

[Kitty]

Thanks to your answer to another post I think I will swap over to a Due - for this board it's fine if the Gameduino attaches directly to the Arduino board (and I've just tested it and it works fine). I think I'm going to have to give up on the above issue.