Tube DIY Asylum Do It Yourself (DIY) paradise for tube and SET project builders.

---

[ Follow Ups ] Thread: [ Display  All  Email ] [ Tube DIY Asylum ]

---

[ Alert Moderator ]

---

Tube Distortion and Bias ... a perspective

Posted by Naz (A) on July 7, 2006 at 08:45:29

I’m forever trying to find correlations between measurements and sound and recent thread on NFB got me thinking about distortion YET again. To post a response I needed to search and review my notes and measurements on some amps I’ve played with over the past few years.

In doing so it reminded me about my first observation of distortion cancellation between the driver and OP stage of an SE amp. I’d like to share my findings and invite discussion, especially from those that have had similar experiences and particularly from the techos amongst us who might wish to comment.

I’ll pre-empt the following by stating that I’ve read with interest the theories of the benefits of 2nd harmonic distortion and the evils of odd order harmonics, particularly high order. At first all seemed logical, then I doubted them and now … well, I partly believe them but I remain unsure about the threshold of our perception of high order harmonic distortion.

Equipment used for my measurements were an HP8903B Signal Generator, which was later replaced with a Stanford Research DS360 Ultra low Distortion Function Generator with a typical THD of less than 109dB at 1V from .001Hz to 5kHz. Pico ADC-216 Spectrum Analyser software was used to monitor the results. This 16bit version has a specified dynamic range of 96dB with a typical figure closer to 110dB. Also used was a Tektronix 468 100mHz Digital Storage CRO.

I’ll begin with some measurements I made a few years ago on a pair of SE1 300B monoblocs using RCA 6SN7 drivers and JJ 300Bs biased at a high 80mA. Load was 4 ohm. OP power was 3W under test. Freq was 1kHz sine wave.

% Distortion at grid of OP tube -
2ndH, 3rd, 4th, 5th
2.78, 0.1, 0.006, 0.001
% Distortion at speaker terminals -
0.27, 0.54, 0.103, 0.026

Same as above with 6N8P in driver position in lieu of 6SN7
Grid
2.09, 0.17, 0.035, 0.004,
Spkr Term
1.00, 0.60, 0.072, 0.030

Observations
1 – Distortion cancellation between driver and OP stages is evident,
presumably due to the effects of opposing transfer curves.
THD is mainly comprised of 2ndH and (more than expected) 3rdH.
2 - Different tubes have slightly different distortion characteristics.
3 - OP distortion can be reduced by increasing out of phase
distortion at the driver stage. However, this could also change the
relative harmonic spectra of the distortion.

Similar results were noted for an early version of my previous single tube SE 845s, which were the forerunners of my current amps. Driver tube was CCS 6N8P, Primary load was 8k, HT was 960V, Idle current was 80mA, Power was 10W.

% Distortion at grid of OP tube -
2ndH , 3rdH, 4thH 5thH
0.11, 0.004, 0.003, 0.001
% Distortion at speaker terminals -
0.40, 0.015, 0.002, 0.001

Same as above with CV1988 in driver position in lieu of 6N8P Grid
0.035, 0.002, 0.001, 0.001
Spkr Term
0.55, 0.027, 0.001, 0.001,

Observation – Similar results to the SE 1 amps, albeit with lower overall distortion. Note the significantly lower distortion of the CCS loaded driver in this example.

Question.
Can you engineer lower total amp OP distortion by adjusting bias on the driver stage? Well, apparently you can. I accidentally found this to be the case by substituting the 6SN7 driver with an ECC33. It sounded much better than the 6SN7 but intuitively, it shouldn’t have because as expected, due to its lower gm the direct swap had the ECC33 biased at such low idle current (<2mA) that it was running in its non-linear region. However when I “fixed” the situation by increasing the current it sounded worse than the 6SN7.

In order to properly compare different tubes and bias currents I made the cathode resistance adjustable. I discovered that contrary to the norm, biasing the driver tubes to lower currents than normal sounded better … sweeter. Low and behold this was also entirely consistent with my distortion measurements.

Further, I found that the more current I ran through the OP tube for lower local stage distortion, the more current I needed to run through the driver tube for optimal linearity. In other words, make the OP stage more linear and less compensation (read distortion)is required in the driver stage. BTW, there are limits and there is an optimal driver current (which depends on the current through the OP tube), below which distortion rapidly increases.

Conclusions?
1) I can now understand the disagreements that continue to rage between inmates regarding tube bias. In fact I recall one of my own with Jeff (drlowmu). I have always advocated higher currents and whilst my belief remains intact for an individual stage, I can see that this will not always be the case depending on the preceding or following stage.
2) The same principle outlined in 1) applies to tubes. Low distortion tubes are generally better but depending on the preceding or following stage this will not always hold true.
3) It’s surprising how easily our ears can pick minor differences in distortion when we have a reference.
4) I’m convinced that SE amps do not sound good because of their high 2ndH distortion. It is possible to engineer an SE amp with very little overall distortion including 2ndH (although it should remain dominant), without the use of Global NFB for correction … and still have that SET magic.
5) Lower distortion is generally perceived as a sweeter sound with less grunge that, I might add, you sometimes didn’t even know existed until it’s removed.

Obviously we cannot perfectly match (in opposition) the distortion characteristics of the OP stage in the driver so this method of cancellation seems suspect mainly due to the possibility of increasing high order distortion. From the results I can only surmise that this is not a problem provided that neither stage contains high order distortion beyond very low levels ... very easy to accomplish with triodes, particularly when CCS loaded.

I apologise for the extremely lengthy post. However I felt the need to balance it with sufficient detail to make sense.

I’ve been meaning to put this forward for discussion for some time so comment is most welcome. Please remember that this is only my perception, based on my experiences. BTW, I’m also not claiming anything new … just trying to put a few things together because I think we so often isolate them to the point that we cannot draw meaningful conclusions.

In this case what I measured is what I heard and not necessarily in that order. In other words, sometimes I heard things that just happened to be supported by later measurements.

Naz

This post is made possible by the generous support of people like you and our sponsors:

Signature Sound

[ Signature Sound Lounge ]

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---