Negative grid on Valve Jr.

[Randall]

I'm flummoxed by this simplest of amps. I have very low output. I didn't test drive the amp before I tried some Guitarologist mods the customer wanted. When finished, the amp sounded like poo, and I wasn't figuring out why, so I restored back to original with the exception of a 100K grid stop between the volume wiper and power tube grid. I believe this is working correctly.

Problem is, I have negative voltages on my grids. With the amp powered down I have correct resistances on all three cathodes. I am not seeing where the negative voltages are coming from in this little guy. Here are my voltages with volume full off/on:

V1 vol off ..........V1 vol on

1. 216v .............180V
2. 0v ..................0v
3. 1.7v ...............1.4v

6. 221v ..............169v
7. 0v ..................-2v
8. 1.5v ................1.5v

V2 vol off............ V2 vol on

2. grid 0.3v ........-17 to -21v
3 cathode 10v ....20v
7. plate 337v ......301v
8. grid 2 309v......254v

These measurements are with sine wave at input jack. Looking with the scope I see healthy gain at the power tube grid pin 2, but with a sloped clip at the top, and a heavily clipped and squat signal at the plate pin 7, smaller in amplitude than the grid. Seems to me it is getting biased somehow, but I am not seeing where or why.

Different tubes, same result. Pulling V2 still has -2v on V1 grid pin 7, but the -20v is gone on V2 grid pin 2.

Something simple and should be obvious I'll wager.

http://www.muzique.com/news/images/Epi1.gif

[Helmholtz]

You seem to be driving V2 into heavy grid conduction.
As soon as the grid voltage exceeds the cathode voltage, the grid act as a diode, clipping the positive signal.
This results in a negative net DC voltage at the grid.
Always expect DC levels to change with signal distortion.

Apart from that, meters do not necessarily read correct average DCV in the presence of large ACV/signal.

I would reduce signal until grid and plate signals are clean and measure output power.

[J M Fahey]

Agree and add:

V2 vol off............ V2 vol on

2. grid 0.3v ........-17 to -21v
3 cathode 10v ....20v
7. plate 337v ......301v
8. grid 2 309v......254v

These measurements are with sine wave at input jack. Looking with the scope I see healthy gain at the power tube grid pin 2, but with a sloped clip at the top, and a heavily clipped and squat signal at the plate pin 7, smaller in amplitude than the grid. Seems to me it is getting biased somehow, but I am not seeing where or why.

1) you are HEAVILY driving V2 grid.
As "helmholtz" says, you have a diode there.
Since cathode is grounded, diode anode (which is the Grid) rectifies signal and becomes negative, it also charges coupling capacitor with negative voltage, exactly what you are measuring.

2) other voltages exactly match that description:
a) cathode voltage is higher (20V vs 10V idle) BECAUSE V2 is being driven hard, so cathode resistor drops more voltage.
b) Screen voltage drops big time, AGAIN because V2 is being driven hard.

3) plate signal is being heavily clipped , again showing V2 is being driven balls to the wall.

4) but ... but .... then amp should be putting out maximum squarewave power and should be deafening inside the Lab if plugged into any reasonable cabinet!
Yup, and it is doing so.

So, WHERE is that power going? Plate amplitude is less than Grid amplitude!!!!

Into the SHORTED OT, where else?

Try another if you have a spare.

IF it were, say, a Twin Reverb I´d suspect you plugged into the wrong out jack, but here only suspect is OT.

Unless you have shorted wiring or speaker itself is shorted.

PS:V1 pin7 is also showing a similar negative rectification, but to a lesser degree since signal level is much lower.

[Randall]

OK. I was driving it too hard I guess. Now it is at 90 mVac at input jack. V1 grid 1 is 46 mVac. My Fluke is not true RMS, but it at least gives me an idea of where it is at.

This gives me about 60 vp-p at V2 grid, pretty clean looking, but with the top a little sloped/ clipped. Plate is about only 50 vac p-p, and a little more distorted, so no gain. This translates to about 1.4 vac p-p at the speaker dummy load and pretty clipped. V2 voltages are: grid = -8.8v, cath = 14.8v, plate = 323v.

So, can it be argued that I am now not driving it too hard, yet still have the problem?

[Helmholtz]

I think, you're still driving it too hard considering the negative grid voltage and agree with JMF that you probably have a shorted OT.
Means that V2 only sees the DCR of the primary (some hundred Ohms) as load instead of 5.2k

The typical "true" RMS meters are AC coupled and thus ignore the DC content (which does contribute to real RMS).
I haven't seen a handheld DMM yet that measures real RMS with ACV superimposed on DCV.

But what you want is not RMS but averaged DCV.

[Randall]

It doesn't pass the neon bulb test, so I guess I do have a bad OT. Thanks for the help gents. Man, I wasted a lot of time on this one. But, at least it makes sense to now.

[J M Fahey]

Sh*t happens.
Neither the first time, nor the last.

[Jazz P Bass]

Yup.
Been there.
Done that.

[Randall]

Put a new Hammond 125DSE in there. It is putting out almost 5 watts now with the original tubes.