Last year I rebuilt my HK Cit 1 preamp and I was just checking the voltages on all the
tubes if the fn reference to the voltage chart in the manual. All of the voltages measere
within 5v of the chart except one 12at7. This one tube measures fine on the A section
( pins 1,2,3) but draws 10% more current on the B section ( pins 6,7,8 )..I have tried
this tube in several positions in the preamp (it takes 5 12at7's) and in every position the
excess current draw follows where I place the tube. Should I be concerned about this tube?
I know that that the voltage chart says +-20%, but I"ve got every thing elae down to under 3%
Thanks,
Bob
tubes if the fn reference to the voltage chart in the manual. All of the voltages measere
within 5v of the chart except one 12at7. This one tube measures fine on the A section
( pins 1,2,3) but draws 10% more current on the B section ( pins 6,7,8 )..I have tried
this tube in several positions in the preamp (it takes 5 12at7's) and in every position the
excess current draw follows where I place the tube. Should I be concerned about this tube?
I know that that the voltage chart says +-20%, but I"ve got every thing elae down to under 3%
Thanks,
Bob
Last edited:
I agree with petertub: ±20% (if you think about it) means –20% → +20%, which is quite a span! 40% from bottom to top!
Its nice when you measure stuff and it is almost spot on the diagram's hopeful quiescent numbers. However, it also is the case (especially if you do this for many, many years) that you find a LOT of stuff which isn't terribly close to the average spec.
And as PETER says… “a well built amplifier should not be materially affected by this”.
Sorry, paraphrased and emphasized. But still its right.
As an ironic aside, I've noted that using fixed bias (such as the now very popular LED bias, string-of-diodes bias and even more exotic contraptions) tends to spotlight tube variations much more than either unbypassed or bypassed cathode-resistor biasing.
The reason is simple enough: the cathode resistor quiescent voltage depends on the current flowing thru. If a section of a tube is particularly 'hot', then a higher cathode voltage will turn up in your measurements. If it is weak, then the cathode voltage is lower. These make the grid more (less) negative, which in turn suppresses (enhances) electron flow respectively. So bias self-adjusts.
GoatGuy
Its nice when you measure stuff and it is almost spot on the diagram's hopeful quiescent numbers. However, it also is the case (especially if you do this for many, many years) that you find a LOT of stuff which isn't terribly close to the average spec.
And as PETER says… “a well built amplifier should not be materially affected by this”.
Sorry, paraphrased and emphasized. But still its right.
As an ironic aside, I've noted that using fixed bias (such as the now very popular LED bias, string-of-diodes bias and even more exotic contraptions) tends to spotlight tube variations much more than either unbypassed or bypassed cathode-resistor biasing.
The reason is simple enough: the cathode resistor quiescent voltage depends on the current flowing thru. If a section of a tube is particularly 'hot', then a higher cathode voltage will turn up in your measurements. If it is weak, then the cathode voltage is lower. These make the grid more (less) negative, which in turn suppresses (enhances) electron flow respectively. So bias self-adjusts.
GoatGuy
I use LED bias when it is practical so I built a jig to test tubes at the plate voltage I expect to use and the number of LEDS that provides the expected operating point. I then buy 10 tubes or so and test them all in the jig. I use the two double triodes tubes with the same plate current for all four triodes.
As GoatGuy says, cathode resistor bias is better at maintaining bias. That could be why almost all classic designs from the valve era used it. Selecting valves for bias point is daft, unless you are happy to repeat the exercise every few years as they wear out - no guarantee that a matched pair will remain matched.
- Status
- Not open for further replies.