Round-off Errors Voc
Brought to you by:
dclugston
Menu
▾
▴
#1 Round-off Errors Voc
I create an Isc-Voc curve by varying irradiance in batch mode. (The Isc-Voc curve is a good 1st approximation to the series-resistance corrected I-V curve.)
Then I copy the tabular data output from PC1D and make a local ideality factor plot (an m-V curve) in Matlab.
It appears that 4 significant digits for Voc is not enough - the spread in the m-V curve is too big.
4 significant digits is enough for Isc (it changes much faster). However, Voc increases logarithmically with irradiance and thus changes much more slowly.
As the m-V curve involves the derivative dV/dI, the difference
An m-V curve from Isc-Voc
Continued:
The m-V curve involves calculation of the derivative dV/dI. The difference dV between two consecutive Voc values may be too small and thus suffer from round-off errors, as "base Voc" is displayed with 4 significant digits.
One solution is of course to decrease the number of data points in the Isc-Voc curve.
Anyway, is there a way to tweak the config files of PC1D so that Voc and Isc are edisplayed with 6 significant digits?