Although, this directly may not give you AM 1.5 spectrum, the attached file is probably a good start.

Thanks for asking a good question. I am looking to build a decent one myself.

Thank you Ananth Saradhi...

Hi,

For that you need Keitheley source meter and simulator with AM 1.5 spectrum, which consists of one sun condition (100 mW/cm2).

For best understanding, check the New Port-Oriel solar simulator, which is very standard one to characterize the I-V measurements of any solar cell in lab scale.

_All the best......

_Thanks,

Chandrasekhar. 

very difficult to match the 5800K blackbody-like spectrum, you need at least a powerful lamp (e.g., Xenon with feedback system for stabilisation), dedicated filters (to match the solar spectrum) and lens to get an uniform beam.  Larger is the beam needed more it will be difficult.

Another possibility is to use directly the sunlight which can be reflected by a flat mirror onto a concentrator and then focused on your PV device. Have a look on this website for instance:

http://www.dlr.de/sf/en/desktopdefault.aspx/tabid-8558/14717_read-28267/

Besides the hardware issues that you are very probably aware of (light uniformity, temporal instability of irradiance, spectral quality), you need to take care of the stability (or let's say light soaking) effects when measuring organic devices. c-Si devices are usually stabilized under continuous light before IV characterization in laboratories. The stabilization is well defined for c-Si, but for organic devices, they may show partially reversible light soaking effects. I would suggest to start with a reference calibration of your device through a PV lab. They will tell you if your device can be stabilized for reliable measurement. In your region, I would recommend Solar Energy Research Institute Of Singapore.