BOI AudioWorks
  Top » Catalog » Article   
Categories
Popular Pre Tubes
 12AT7 / ECC81
 12AU7 / ECC82
 12AV7 / 5965
 12AX7 / ECC83 / 5751
 12AY7/6072
 12BH7
 2C51 / 5670
 6CG7 / 6FQ7
 6DJ8 / ECC88 / 6922
 6SL7 / VT229 / 5691
 6SN7 / VT231 / 5692
 EF86 / 6267
 Other Preamp Tubes

Popular Power Tubes
 6550
 6BM8 / 6GW8 / 14GW8
 6L6 / 5881 / KT66
 6V6
 7591A / 7868
 EL34 / 6CA7 / KT77
 EL84 / 6BQ5 / SV83
 KT88 / KT90
 2A3 / 300B / 811 / 572
 Other Power Tubes

Other Tubes
 Rectifier Tubes
 Sub-miniatures
 Tubes By Number->

Parts and Stuff
 Capacitors->
 Sockets
 Transformers
 Amps and Accessories
 DIY Parts
 Specials! Specials!
Manufacturers
Various refers to several manufacturers
Quick Find
 

Advanced Search
Information
FAQ
Shipping & Returns
Privacy Notice
Power Tube Matching
12AX7 Guide
Languages
English Japanese
Power Tube Matching
Tubes have several important charactersitics, but they all vary quite a bit from one tube to the next. This is hard to avoid because vacuum tubes are hand made to this date.

Balanced Operation
On a power amplifier, whether for Audio or Guitar application, the final stage is often configured in push-pull arrangement. In this configuration, the power tubes work by helping each other like a seesaw. If the power tube characteristics do not match, then the circuitry has to work like an unbalanced seesaw.

Many amp circuit claims to compensate for this by having bias adjustment. This is, however, comparable to forcing an unbalanced seesaw. You may make it seem like a seesaw is balanced, but as soon as you start moving, things are unbalanced. On a high output amplifier, or on a Class-A amplifier with high plate current, one of the power tubes will end up working a lot harder than the other ones by flowing more plate current through. This can affect sound quality and tube longevity.

On Tube Variation
The plate current varies quite a bit from a tube to tube, even in the same lot. The table shows the actual measurement from a lot of 25 Sovtek 6550WE tubes in triode configuration with -48V bias and 400V applied to plate. You can see that the plate current is centered around 55mA which is close to the value derived from data sheet. The variation is quite a bit as you can see, and the difference between minimum and maximum is twice as big!

Suppose you used the minimum and maximum plate current tubes from this lot. You can adjust the bias so they both idle at the same plate current. But tubes with this much difference will not have the same transfer curve. So once the circuit starts driving the output, you will end up with asymmetric operation. If you continue high power operation, or if you use such set of tubes in an amp with no bias adjustment, one of the tubes will end up working much harder. This can lead to shorter tube life and lower sound quality.

Conductance Matching
So far, the discussion is about bias matching, which is a static thing. What about dynamic matching? Does such a thing exist? Well, yes. The transfer function of a tube is called conductance. It shows how a tube trnaslates input voltage into current flow through its vacuum. The conductance values can vary quite a bit, and within certain range, a larger number isn't any better than smaller number. But it is important for the matched tubes to have similar conductance. This minimizes distortions that are caused by assymetry in the push-pull configuration.

So, matched power tube sets are a good thing. Even if your amp is single ended, If you purchase a matched set and keep the spare, your amp will operate and sound the same when you are forced to use your spare.

About Factory Matching
Now, vendors can get factory matched sets which are matched to a rather wide tolerance. This is because manufacturers and big name brands have to process a large number of tubes, and they can only have so many different buckets of values. For example, we took a quad matched set of KT-66 from a famous brand and measured the plate current. They varied as much as 20%. Normal factory matched sets are only within 25% tolerance or more.
As of January 2005, we match all power tubes ourselves.

Our Matching
On tubes that we match ourselves, we first burn-in the tubes at a fairly high plate and screen voltages. We don't believe in overstressing tubes during burn-in, but we do apply max voltages to look for failures including any abnormally high grid current. We reject any that fails, then test to match tubes.
On certain consistently made tubes like SED EL34, we can usually match the tubes to within a couple of mA of Ip and less than 5% variation in transconductance. Unfortuantely, the same can't be true for all tubes by all manufacturers. Our standard is to match within 5% static Ip and 10% conductance. On tube types where variation makes it more difficult to match as close, we target 10% static Ip and 10% conductance matching.

Number Game
Because of the wide tolerances in power tubes, matching has become an important value added service. Since tubes are matched to within a tolerance, the measured value written on the box should vary. A lot of vendors employ a bucket system where tubes that fall into certain range are grouped as the same. Some vendors simply write the same numbers on the box to make it appear as if the tubes are perfectly matched. We try to represent what we ship the best we can.

Continue

All Contents Copyright © 2003-2008 BOI AudioWorks

Your IP Address is: 38.103.63.16