Planck's constant from an LED

In this lab, we can use the principles on which an led operates to experimentally determine planck's constant.
We know the following relationships:
λ=h/mv
E=hf=hc/λ

We used 4 different color LEDs: Red, yellow, green, and blue. We ran the experiment the same exact way as the hydrogen and color spectra lab.

Color	Distance D (meters)	Voltage (Volts)	Wavelength (nm)
Yellow	0.67	1.88	592.43
Green	0.59	2.52	529.04
Red	0.735	1.82	641.93
Blue	0.525	2.64	475.63

Using the fact that E=hc/λ=q*V, we get that h = q*V*λ/c

Color	Wavelength (nm)	Voltage (Volts)	h
Yellow	592.43	1.88	5.94E-34
Green	529.04	2.52	7.11E-34
Red	641.93	1.82	6.23E-34
Blue	475.63	2.64	6.70E-34

From these results, only blue seems to give a value that was within scientifically acceptable bounds of the theoretical value at 1.12% error. Red also came in at a relatively low error of 5.98%. Both Yellow and Green had relatively high error. Coincidentally, their spectra also contained various colors which made it hard to pinpoint the maxima that corresponded to the color of the LED itself. This is likely due to the fact that the green LED had impurities and was contained additional wavelengths. The particular shade of yellow also had an orange tinge to it. Overall, it appears that the colors on the ends of the visible spectrum were much easier to identify.