Express Harmonics and Feedback

[gearhead]

SB: Jump off bridge? Maybe throw it off a bridge, but I already started my Liverpool. Now if that sucker is a bummer, then all three of us will jump together, LOL.

For the Plate resistor quandry:

What I'm talking about is doing two things and then comparing.

1) Measured at the TOP of each of the plate resistors (which gives you B+5, B+4, B+3, which matches whats coming off the power filter caps -after- the dropping string). V1a and V1b share the same B+5 feed. Then measured the voltage of all four plate resistors at the bottom, which is connected to each corresponding plate socket pin. Subtract and you get a per-triode difference between the B+ feeding the plate resistor and the plate.

2) Measure voltage directly across each plate resistor.

Comparing 1) and 2), they aren't the same values.

For instance, in the most extreme case,

Top of V2 plate resistor measures as 295V, which does match the B+ at the filter cap.

Bottom of V2 plate resistor measures as 250V

Difference is therefore 45V

Measure directly across that V2 Plate resistor, one probe on one lead of resistor and other probe on other lead, and I get 25V !! Consistently. 20V is disappearing.

This is eerily similar to my initial problem of extremely low volume.

Glen:

Sometimes I think it is a solder issue. When it hit the "zone" yesterday, I suspect either the bad joint or the bad component was still hot from being reheated and conducting like it should. When it cooled down, bam, back to where it was. Or I could be making this up. Cause and effect is so tenuous at times. Hahaha,

Thanks. Really appreciate your thoughts and will try some of your suggestions.

[jjman]

Low volume can be caused by ultrasonic oscillation. The plate resistor voltage drop change sounds like it could be from one measurement-scenario being during oscillation and one w/o oscillation

[gearhead]

It's consistent and highly repeatable. There may be osciallations, but don't think they are turning on and off to give different readings. I did take both the plate and B+ voltages wrt ground, and the across-resistor obvioulsy not wrt ground, so may be mixing apples and oranges.

At this point, I'm going to just start replacing and/or resoldering all suspect items, starting from the input jack forward.

[Firestorm]

I can't think of anything that would account for the plate voltage discrepancy (you're measuring with no signal applied, right?) except a screwy meter. Do you know what the meter's input impedance is when measuring DC volts?

Before you tear too much stuff out, look real hard at the bright switch and volume pot wiring (one place that's hard to see in the pics you've posted). Maybe check the resistance from the wiper of the treble pot to the V1B grid as you rotate the volume pot (amp off, of course). This is the place in the amp where all signal attentuation occurs, so if something is wrong you could lose a lot of signal here. Also if the series resistance is overly high here (a bad joint, etc.), that resistance could work with the tube's Miller capacitance to roll off a lot of high frequencies. The fact that you were getting strange behavior when rotating the volume pot makes that seem the first place to start looking.

[gearhead]

I ended up completely disconnecting the bright switch to eliminate that as a possible leak source. Did the volume pot measurement, but got pretty wonky readings. Near one end of rotation it was zero (short), but all the way over to near the other end it hovered in the 150-200k range.

Believe that that pot section is in parallel with other parts of the circuit, so it's (x*y)/(x+y). Pot, isolated, does work correctly.

WRT plate discrepency, don't believe it's the meter. I can't find the manual; it is a sacrifical cheap one, but it measures about the same as my best one for voltages in that range. I'd be leery if it was a 5% error, but not a near 50% one. It must be reference point specific, or there is one huge current drain.

My turrets in the PI seciton had so much solder due to solder reflow, and replacement, that they were beginning to look like Hershey Kisses. LOL. Am in the proces of cleaning that up (new wires, new socket, solder removal).

Am probably going to yank the entire pot assembly and check it out. Have done -numerous- continuity checks, so there is no hard short. If it's there, it's some kind of solder-dielectric semi-capacitive issue.

[jjman]

If the meter is "using" some current when measuring across the resistor it would explain the different drop reading. I would imagine that analog meters would "use" much more current than digital but I have no guesses as to those figures. You're not using an analog right?

[gearhead]

No, it's digital. Am perplexed to think a meter's internal impedence would drop 20V of a 45V value. Who knows though!!

Keep the ideas coming! Thanks,

Dave