Bypass Systems Compared
Distortion in audio bypass switches.
©2019 By Jack Orman
It is often the case that once you begin an experiment, the results will lead you to a place that you did not anticipate, and the the answers are not as they first appear. Such are the results from my tests of the bypass switching systems.
I had done some initial tests with the QA401 audio analyzer and thought that I had a good idea of what whas happening with the jet bypass switching systems. The data that I obtained from the inital tests were accurate but did not tell the whole story.
I refined the settings on the analyzer and turned off all equipment in the test area that was not being used. I also switching off the overhead fluorescent lights to reduce interference. In the end, the lights made no difference and the residual noise that I encountered was coming from my HP laptop computer. Interestingly, the laptop was less nosiy when running from its power adapter than when running off its internal battery.
The best way to reduce noise it to increase the distance between the noise source and the circuit that is picking up the interference. By moving the test device and carefully routing the wires to pedals being tested, the hum and interference disappeared into the noise floor. All pedals were powered by the same brand-new 9v alkaline battery though their power jack. During testing, the audio plugs were pulled from one pedal, then the next was placed in the same position and the audio plugs inserted. Care was taken to make sure that all pedals under test were in the same exact position and distance from the computer.
Also, the AMZ relay bypass system was tested but since the results were no different than with a standard 3PDT, the graphs are not included here.
The first sample of the true bypass system was an AMZ Mosfet Booster pedal in a 125b box using a 3PDT switch. The measurements were the same as with the test osciallator going direct to the QA401, except for a tiny increase in 60Hz interference. The total harmonic distortion (THD) is 0.00045% which is extremely good and can be attributed to the residual from the test oscillator. (Click on the image to see the full size graph)
The second and third harmonics are below -120db and will be inaudible.
The test was repeated with an AMZ Mosfet module, which was mounted in a 1590A box. It is also true bypass using the common blue 3PDT switch just as with the previous pedal tested. The results are practically identical.
Minor differences in the 60Hz noise are likely due to the internal wiring of the pedals, but in either case, are so low as to be below the threshold of hearing.
The true bypass pedals also had no signal loss. There was 316mVrms going in, and 316mVrms coming out of the pedals.
For the test of the buffered jfet bypass switching system, a Boss SD-1 pedal was selected for testing. The pedal was new and never used.
There was a signal loss of about 0.8dbv due to the transistor buffers, as was expected. The interesting thing was the numerous low level harmonics that were seen above the 1kHz test frequency. More importantly, the THD had risen to 0.062% (from 0.00045% with true bypass).
The results from the SD-1 showed so many high distortion harmonics that it was decided to get another new SD-1 pedal and repeat the test run. The results were the same, with plenty of high frequency distortion peaks. A third new SD-1 was then taken from its box and tested with the same results.
As had been planned for this test series, a new Boss DS-1 pedal was inserted into the signal loop and tested. The results were significantly different than those obtained from the SD-1 pedals. There were far fewer distortion harmonics with only the second, third and fourth appearing at significant levels.
However, even without the higher harmonics, the THD was 0.0644%, which is much greater than the results achieved with true bypass switching, and similar to the distortion from the SD-1 bypass.
Since there was not another Boss DS-1 available to test, a new Daphon compressor pedal was tried since it has a similar jfet switching method. The results were almost the same as those obtained with the DS-1. Then aother new Daphon pedal was tested and the response was also very similar. It became clear at this point that the jfet switching method was not completely at fault for the numerous distortion harmonics of the Boss SD-1.
While thinking about the test results with the jfet switching pedals, I remembered a problem that has long been associated with the Boss SD-1, which is leakage of the distorted signal into the audio output, even when the pedal is in bypass mode. Clearly I was measuring this defect in the Boss layout and switching. I assume that layout is partially to blame since the Boss DS-1 is also a very high gain distortion pedal with jfet switching, but it has fewer harmonics in its output.
The bottom line is that the increased THD in the Boss SD-1 and DS-1 pedals are real, and far greater than that obtained with a true bypass pedal.
Jfet bypass switching is good enough for guitar pedals and amps, which characteristically have limted high frequency response, though I would not choose the system for high end audio. The loss of signal level from the buffers is small but could be noticeable, and will add up if numerous buffered pedals are used in series.
In comparison, the true bypass system has no loss and introduces no distortion, but still has the problem that long cables used with it will result in some high frequency roll off. A combination of a buffered signal and true bypass would likely result in the best of both worlds!Share this article with your Twitter followers: Tweet
AMZ-FX Home Page
Lab Notebook Main Page
Guitar Effects Blog
©2019 Jack Orman
All Rights Reserved
©2019 Jack Orman