In this blog post I compare distortion data between the stock KT90 tube versus a new pair of Electro Harmonix KT90 Tubes.
I originally tested the Minda amplifier here. I deceded to repeat my measurements using the old tubes just for consistency. My test involves measuring distortion at these output voltages:
I test the following types of distortion
- Harmonic (H2,H3, H4)
- Intermodulation (IMD)
- Gedlee Metric (Gm)
I will start each measurement with the stock tube and then directly compare the new tube's performance in the next graph.
We can see that the EH tube has no change on harmonic distortion at the 0.06V test voltage.
At the 0.25v test we see the EH tube doing slightly better through the midrange with H2.
We can see that the EH tube does is doing better particularly with H2 at 0.05% compared to 0.12% with the stock tube.
Again, we see H2 at 0.27% with the EH tube compared to 0.53% with the stock tube. H3 and H4 are virtually unchanged between tubes.
I then measured IMD at the various test voltages starting with the stock tube. We see IMD at -60dB for the stock tube for the 10kHz region. Through the midrange we see some noise creeping up to -50dB.
With the EH tubes we see the same result but with a little bit more noise artifacts through the midrange.
With the 0.25v test voltage we see IMD at -59dB for the stock tube with the EH tube at -70dB (10kHz region)
At the 1.00v test voltage we see things unchanged in the 10kHz region but we do see the EH tube perform better in the midrange (1kHz region) and the upper treble region (20kHz).
We see the EH tube doing about +3dB better across the spectrum.
Gedlee Metric (Gm)
Below is a comparison of the Gm distortion between the stock tube and the EH tube. We can see that the EH tube does better in the upper treble, reaching a Gm value of 0.0002 for the lowest test voltage.
Below is a graph showing the stock tube performance. The graph shows the Gm distortion at the four test voltages as an overlay.
Below is the Gm distortion for the EH tube. The EH tube does better than the stock tube across the spectrum.
I decided to measure the square wave performance at the 0.25v test voltage. I did not detect any change in the square wave between the 0.25v and 1.00v test voltage.
It comes as no surprise that the EH tubes outperformed the stock chines tubes. However my measurements quantifies the improvement in a meaningful way. It appears as though the output tube does little to improve the noise floor of the amplifier at the 0.06v, but we see significant improvements in the 0.25v to 1.00v output voltages.