Digital Guitar Tuner

Eric Tai | Daniel Tsui

ECE 476: Spring 2005


Introduction | High Level Design | Program/Hardware Design | Results | Conclusions | Appendix




††††††††††† The code ran with a timebase of 0.1ms, and for the most part, was indicating frequencies correctly.Our tuner was calibrated and tested with two guitars, as well as the signal generator in the lab.For the acoustic guitar, five of the six strings were being tuned accurately, with the exception being the G3 string.For some reason, the G3 stringís first harmonicís amplitude was so strong that it was overwhelming the amplitude at the fundamental frequency.We attributed this to strange behavior on the guitar, and continued testing on the electric guitar.


††††††††††† With the electric guitar, we had the same issues, where upper order harmonics were still dominating the signal input.After some consultation with other guitar players, we realized that the same sort of behavior is also apparent in commercially-made tuners.While the signal will start with the dominating upper harmonics, the fundamental frequency will slowly take over later in time.


††††††††††† The benefits of testing using the signal generator were that specific frequencies and amplitudes could be calibrated easily.We could feed a specific frequency into the tuner and not have to worry about fluctuations in amplitude or frequency associated with plucking a guitar string.However, we could not simulate upper harmonics, or the low pass filter.


††††††††††† When testing the guitar tuner against the function generator, where the input frequencies are known, we noticed a lot of fluctuation in the LEDís that were due to the natural error from the generator.While the generator can be accurate to 1 Hz, it fluctuated a lot in the tenths and hundredths of Hz.A jump in 0.3 Hz would cause the guitar tuner to be out of tune.This corresponded to a %0.36 error.


††††††††††† Safety enforcement was not a big factor in this project, as there were no lethal or dangerous parts used.In addition, there was no RF intereference caused by the tuner.The tuner also generated no noise.In terms of usability, it was not a very complicated design.The audio jack allowed for the sound input, and one of six pushbuttons were pressed.There was no complicated calibration, since the frequency checking was done internally.


Picture: Using the oscilloscope for tuning purposes.