One of the questions I wanted to answer with my Garduino was how much light is lost through the plastic covering of the Greenhouse.  What I have found is it is very difficult to find a light sensor that has the dynamic range from darkness to full sun, but I finally stumbled on the MAX44009, which has a range from 0.045 to 180,000 Lux.  The downside of this device is no one is making a breakout board for it, so I was on my own.

Homemade breakout board for the MAX44009

Yes, this beautiful homemade PCB does work.  I made it by the laser toner transfer method, but my home printer, a Brother HL-4040CN did not work; the toner did not stick to the copper board no matter how long I put the iron on it.  I ended up creating a PDF file with Gerber2PDF and printing the image on some HP printers at work.  Unfortunately,  in my rush to not be caught screwing around with something that wasn’t work related, I forgot to make a mirror image of the PCB, so everything is backwards on the PCB.  Even with the HP toner, the toner transfer was not very good, but, considering the fact that the MAX44009 chip is 2mm x 2mm and the pads for the 6 pins are 0.36mm x 0.48 mm, the toner transfer was accurate enough to work.  Adding to the fact that this was the first time I used my hot air re-work station to do something constructive, you could have knocked me off my chair with a feather when the final assembly actually returned valid data.

It wasn’t a complete Eureka moment, though, because while the data sheet says the I2C address can be selected (with the AD pin) to be 148 or 150, that does not seem to be true.  I found this important utility, I2CScanner which runs through every I2C address, and shows which addresses a device is responding to.  This really saved me, because it showed that the real address of the MAX44009 was 203 (I forgot to note the address when AD is low, so I’ll document that when I make the improved version of the PCB).  Once I had the right address, interfacing with the MAX44009 is very straightforward, since it uses the I2C Wire Arduino  library:

#define MAX_ADDR 203

float getLux()
{
int luxHigh=readI2CAddr(0x03);
int luxLow=readI2CAddr(0x04);
int exponent=(luxHigh&0xf0)>>4;
int mant=(luxHigh&0x0f)<<4|luxLow;
return (float)(pow(2,exponent)*mant)*0.045;

}

int readI2CAddr(int addr)
{

Wire.beginTransmission(MAX_ADDR);
Wire.write(addr);
Wire.endTransmission();

Wire.requestFrom(MAX_ADDR,1);

int n=0;
for(n=0;n<25;n++)
{
if(Wire.available())
break;

delay(100);
}

if (n==25)
return 0;
else
return Wire.read();
}

I was expecting to have to handled the situation where I would have to programatically change the sensitivity of the device so that I would have to set longer integration times in low light conditions, and shorter ones in bright conditions, but the device handles that all by itself.  The device can be configured in a manual mode so that the user has control of exactly what integration time will be used for each reading, but in my case, the automatic setting behaved the way I wished.

I’ve created an Eagle library that contains the land pattern for the MAX44009.

I really like the I2C bus, but its major disadvantage is the max length which is reportedly pretty short.  I currently have the MAX44009 sitting off of about 2.5 M of 6 conductor satin cable, connected to the same 3.3V bi-directional voltage translator circuit my BMP085 is connected to, and it is working well.  I expect when I have to connect another MAX44009 to measure the light loss from inside the Greenhouse, though, the length will become a problem.  I will be researching some I2C bus lengthing techniques and will report on them in the future.

The waterproof enclosure for the MAX44009 consists of a plastic food container with a clear plastic cover, and the bottom cut out for ventilation.

This is the PDF for the Break Out Board that I used for the Toner Transfer method to create the PCB.