Tubes Asylum Questions about tubes and gear that glows. FAQ

---

[ Follow Ups ] Thread: [ Display  All  Email ] [ Tubes Asylum ]

---

[ Alert Moderator ]

---

Re: Non-match power tubes

Posted by Steve O (A) on July 22, 2002 at 13:14:39

In Reply to: Non-match power tubes posted by steven oda on July 21, 2002 at 22:48:42:

First, my understanding of tube matching basics:

Good tube testers measure transconductance (gm), (emission testers need not apply). gm is a measure of a tubes ability to convert a voltage signal at the control grid (g1) into a current signal at the plate under a specific set of conditions (plate and screen voltage and plate current). It is usually taken to be a constant but actually can vary as static plate current varies. For this discussion assume it's constant. It is expressed as a ratio, say Amps/Volt. This is also unit for conductance so mho(old term) or sieman (new term) is used. For tubes, a good unit of measure might be mA/V (umho is 1000 X smaller). If you were to plot various mA on Y axis against V that produced it on X for a given set of operating parameters, you would end up with a straight line with slope of mA/V (disregarding real life gm dependence on plate current). The slope is gm. The current at any given grid voltage will be the bias or idle current under those same conditions. Note that the idle current isn’t necessarily related to the gm. Unfortunately, the slope and only the slope is what most decent tube testers are measuring. A slope w/o some reference point is pretty meaningless since a line of a specific slope could be located anywhere on the plot. Properly locating the line on the plot requires a reference point of some sort. This reference point can be provided by knowing the specific current produced by a specific grid voltage under specific conditions. In other words, a POINT and a SLOPE associated with the point unambiguously defines a SPECIFIC line. In effect, defining a SPECIFIC line is what a good job of tube matching does. You need both the gm (slope) and the reference point (reference bias current). This is just another form of the equation for a line : Y=mX+b.

How does all of the above relate to runaway tubes? If one matched only for gm then idle current match is pure chance. Say you have two 6BQ5s both at 4000 umho (made up value). Say also that at a plate V & g2 of 375 and g1 at -15 (approx ST35 conditions), one 6BQ5 draws 20ma and the other draws 40 mA. Say also that both share a common cathode resistor. If you measure the current thru the resistor you’ll find 60mA or average of 30 per tube. 30mA x 375 = 11.3 W which appears OK. However, we know that one of the tubes is actually drawing 40mA = 15 W which is pushing it. This tube will run hotter and depending on tube characteristics g1 may become hot enough to begin emitting electrons itself. When this happens the grid voltage will be reduced causing increased current flow thru the “hot tube” and diverting current from the cool tube. This causes the hot tube to run hotter yet, increasing g1 emission , increasing plate current and so on till something gives…usually hot the tube. Conventional cathode bias will tend to have a somewhat limiting effect but can still run away. Fixed bias where the two tubes aren’t individually adjustable can also run away. Bottom line here is that although gain was matched, bias current wasn’t and potential runaway can occur. Even if it doesn’t, there is unbalanced current in the OPT which can induce premature core saturation and distortion if we're dealing with a PP amp.

Now imagine that the tubes were matched for bias at 30mA each but not gm. At idle both tubes draw 30 ma and the dissipation is 11.3 W/tube. Since gain isn’t matched, there will probably increased distortion but at least the tubes are operating under design conditions. If corrective feedback is used, even the distortion may be reduced to tolerable levels. No doubt, this is what you experienced in your example of mismatched gm sounding lousy.

The best match would include bias point to prevent tube runaway and gm to produce an inherently low distortion output stage. As I mentioned previously, very few tube testers will do bias matching and yet a lot of tube sellers are pedaling “matched” pairs and quads where the match was only for gm…a disaster waiting to happen. The really knowledgeable sellers match for gm AND bias at approx. real world operating conditions using either a fixture designed to measure both or by using a conventional tube tester for the gm meas. and an actual amp for the bias meas. (not optimum but a whole lot better than what most do). Another means around the bias match issue is individually adjustable bias for each output tube along with adjustable drive to handle minor gm mismatches. I like this route + the fully matched set of tubes.

Note that a stock ST35 has four output tubes tied to a common cathode bias resistor which places severe demands on the bias match of the output tubes. Since the amp employs corrective feedback, the gm match is a lesser concern. Although Dyna isn’t adamant about it, a stock ST35 MUST use a FULLY matched quad of output tubes. You're taking big chances by replacing only the bad one.

As far as examples of the runaway happening to me: you bet it's happened. Interestingly, on a ST35 (same basic situation as original poster), an ST70, a MK3 and an MC225. This was in my naive days when I thought a good tube tester told me everything I needed to know. Then I set up a fixture to measure plate current vs g1 voltage and was astounded at how wide the variation in idle current was given a very close gm match on a tester. I now either buy large quantities of tubes and match for myself using primarily bias point secondary gm or buy small quantities from reputable dealers who match the same way. Some of the AA sponsors do it "my way' so check em out. Hope this helps. And...critiques and error correction of my explaination welcome: not easy to explain all this in text only.

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

VH Audio

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]

---

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

---