Long Tail Pair Puzzle

[clyde]

I built a mule to test preamp circuits, using a LTP for EL84 outputs.  The output is nastily distorted with hard playing, much like parasitic oscillation or having the plate leads too close to each other.
The wiring seems kosher but the voltage readings don't make sense.  Please see attached.  I have subbed in another 12AX7 with an outboard socket and hookup, no difference.  Thanks for any and all suggestions. (31v on the top grid)

[sluckey]

You cannot accurately measure the grid voltage on a LTP PI because your meter will load the high impedance grid circuit and cause the voltage reading to always be lower than it actually is. You can accurately measure the plate, cathode, and the junction of those four resistors. The voltage at that junction is the same voltage that's actually on the grids.

The top triode seems to be working normally, but the bottom triode is passing no current as evidenced by no voltage drop across the 100K plate resistor. A likely cause would be an open cathode circuit due to a bad connection or maybe a missing jumper between pins 3 and 8.

[clyde]

Thank you Steve, yes the tube doesn't seem to be conducting.  The jumper is there and I measure 1k2 from the bridge and across the resistor.  I did sub in another tube in a different socket (jumpers) same result.  Would lead dress be an issue as the resistors are almost touching?

[sluckey]

Are you absolutely sure the circuit is wired as you have drawn? Are you using 6.3VAC for the heaters? If so, is there a good connection between pins 4 and 5?

A hi-rez pic would be very useful.

[clyde]

When I remove the tube and clip in an outboard socket and tube, same results, yes 6.5v on heaters.

[clyde]

Another pic.

[sluckey]

I see plenty of opportunities for poor connections using those terminal strips.

[clyde]

True that.  I made the exposed leads long enough so they'd be clamped by both screws and everything measures as it should ohmage-wise.  Using the different tube/socket combination eliminates those two variables, and that's how I troubleshoot.  Maybe I'll just solder up the PI as it's config won't change but I'd still like to know what's causing this.   

[clyde]

Bad cap.  I replaced the .1 ceramic cap to gd. with a film cap, all's good.  Any explanation as to why this wouldn't allow the triode to conduct?

[tubeswell]

I see plenty of opportunities for poor connections using those terminal strips.

Poor connections AND unwanted signal noise coupling.


(The only place I’d use a screw terminal strip in a guitar amp is on the mains side of the power transformer to connect the neutral mains wire to the PT primary)

[sluckey]

Bad cap.  I replaced the .1 ceramic cap to gd. with a film cap, all's good.  Any explanation as to why this wouldn't allow the triode to conduct?

How did the cap fail? Shorted, open, some high resistance in between? If the cap were shorted you would have zero volts on the grid and about +50V on the cathode. That's enough to bias the triode way beyond cutoff.

[clyde]

The mystery continues as a cursory examination using the cap function on my Fluke shows normal, resistance shows normal, doesn't pass DC, wrong colour perhaps? 

I see plenty of opportunities for poor connections using those terminal strips.

Poor connections AND unwanted signal noise coupling.


(The only place I’d use a screw terminal strip in a guitar amp is on the mains side of the power transformer to connect the neutral mains wire to the PT primary)

Not to worry, this will never leave my bench, strictly a test rig and it's not excessively noisy even.  Thanks everyone for you help and suggestions, very much appreciated.