Bias Rite measurement discrepancy

[jetwolf]

Can anyone think of an explanation for this?

I just installed a quad of the GTEL34Ms in my Marshall 1959 homebuild last night. This amp has installed quality 1-ohm resistors between pins 1 and 8 and then the extra lead through pin 1 continued to ground for cathode current conversion. I first checked the amp with my Bias Rite, and thought my tubes looked quite a bit out of spec for matched tubes. But then I decided to doublecheck by taking the pin 8 (cathode) readings and they were pretty much right on.

BIAS RITE (measured twice)
Tube
A 36.6 A 38.0
B 31.3 B 32.4
C 40.4 C 42.2
D 35.3 D 36.9

FLUKE- PIN 8
Tube (measured twice)
A 35.9 A 37.4
B 35.8 B 37.2
C 36.1 C 37.5
D 35.9 D 37.4


As you can see, the Fluke measurements on Pin 8 are right on. The Bias Rite B and C tubes are off a bit more than I would like. One is about +5ma and one is -5ma.

Next, I picked Tube A to compare the Bias Rite measurements using the same tube and tube slot Slot A but different Bias Rite heads. The measurements were pretty much dead on.


SAME TUBE A/SAME TUBE SOCKET A (far left) (I had already raised the bias to bring it closer to the 30-31ma (489v at plate) range from my first readings, so ignore the fact that these are lower)

Bias Rite A 30.6
Bias Rite B 30.8
Bias Rite C 30.8
Bias Rite D 30.8
Fluke MM on Pin 8 30.6

All right dead on. Any ideas on what is causing the 5ma discrepancy on the two right tubes when measured by the Bias Rite? Which readings should I trust?