[Drake] 6JB6A tubes

[Garey Barrell]

Larry -

There are a couple of problems caused by using "mis-matched" final tubes. 

Since they are operating in parallel, if one draws more power than the 
other, the dissipation of that tube will be higher than the other(s).  
If each tube is rated at 10W dissipation and the total dissipation is 
20W, one may be dissipating only 5W, leaving the other to dissipate the 
other 15W.   Obvious.

The not so obvious problem is the 3rd order distortion produced by the 
amplifier.  The distortion products go up rapidly as a tube is 
overloaded.  So if your TR-4 has three tubes, but one isn't "pulling 
it's share", the other two must be driven harder, into distortion, to 
produce the output power you expect to see.

So it's not catastrophic, especially since typical differences in good, 
new tubes are fairly small.  But if you put in two new tubes with one 
that has seen better days, the two good ones will have to work harder to 
produce the expected output, generating more distortion and failing sooner.

Drake, and now RF Parts, offer matched pairs and matched trios of 
6JB6s.  This same situation afflicts ALL transmitters using parallel 
tubes, not just Drake or sweep tube amplifiers.

73, Garey - K4OAH
Atlanta

Drake C-Line Service Manual
<http://www.k4oah.com>



Larry Knapp wrote:

>Yes, I agree.... interesting thread.  As an owner of three Drake transmitters (two using 6JB6's) I
>would be interested to hear from the 'experts' on the issue of 'matched pairs' on transmitters
>using three 6JB6's, like a TR4 and it's variants.  Seems to me that a 'matched pair', then a
>single tube, not matched, might be an issue; or not?  Yes, I understand measuring of cathode
>currents with three tubes and matching them.  However, for the majority of us, what's the harm in
>using three tubes that may or may not be 'matched'; or if two are 'matched' and the third is not?
>Just curious.  I'm not sure if any of my two Drakes (T4XB and TR4CW/rit) use any kind of 'matched'
>tubes...seriously doubt it.   73, Larry KC8JX
>
>
>  
>