AVC Repair - AI Part 3

Prompt: I grounded A2. Immediate reduction in static noise. I set generator output to 10 millivolt (1 millivolt into antenna). I tuned away from 1400 KHz and zeroed S Meter again. Now, as I tune from slightly below 1400 KHz to slightly above, I see a very small indication on S Meter and slight positive movement on AVC voltage (on scope) as I pass 1400 KHz. Double checked - it's positive, not negative. I still need to increase signal level to 530 millivolt to get the AVC to 0.

The following has been reformatted using HTML

Grounding A2 was exactly the right move. The immediate drop in static noise proves that your antenna coil primary is now properly referenced to the chassis, allowing the signal to magnetically couple to the secondary.

Now, let's look at your voltage readings, because they actually contain fantastic news hidden inside a confusing paradox.

### The Math of the Voltage Swing

You noted that the baseline is sitting at **$+1.4\text{ V}$**. When you hit the radio with a massive signal ($530\text{ mV}$ from the generator, which through your attenuator is $530\,\mu\text{V}$ΓÇöa crushing S9+20dB signal), the AVC line drops from $+1.4\text{ V}$ to $0\text{ V}$.

**Going from $+1.4\text{ V}$ down to $0\text{ V}$ is a negative movement of $1.4\text{ Volts}$.**

This proves definitively that your 6H6 AVC rectifier diode is in perfect working order! It successfully generated $-1.4\text{ V}$ of DC bias. The problem is that instead of pushing the tubes into the negative region to control their gain, that $-1.4\text{ V}$ was entirely "used up" just fighting the $+1.4\text{ V}$ DC leak to get the bus back to zero.

### The "Slight Positive Bump" on a Weak Signal

When you tuned across 1400 kHz with the weak $10\,\mu\text{V}$ signal, you saw the line go slightly *more* positive. This is a classic symptom of a vacuum tube being forced to run with a positive grid. When a grid is positive, it draws current from the cathode. When you peaked the tuned circuit, the AC signal briefly altered the grid current draw, and the AM modulation (your 1 kHz tone) likely caused a slight asymmetrical pumping effect that added a few millivolts of positive charge to the line.

### The Mandate: We Must Find the Leak

You cannot align the RF or Mixer stages while this leak exists. A positive grid acts like a heavy resistor strapped across your coilsΓÇöit kills their "Q," broadens tuning, and destroys sensitivity.

Leave the signal generator off for a moment. We need to find exactly where that $+1.4\text{ VDC}$ is coming from.

Leave your scope or DMM clipped to the AVC line so you can see the $+1.4\text{ V}$ baseline. With the radio powered on and warmed up, **pull these tubes out of their sockets one at a time**, watching the meter:

1. **V1** (1st RF) 2. **V2** (2nd RF) 3. **V3** (Mixer) 4. **V5** (1st IF) 5. **V6** (2nd IF)

If you pull a tube and the $+1.4\text{ V}$ instantly vanishes, you have found a gassy tube (grid emission). If you pull all of them and the voltage is *still* there, the leak is coming from the AVC amplifier/rectifier stage (likely a leaky coupling capacitor or a shorted tube in that section).

Go pull those tubes one by one, and let me know exactly when that voltage drops.

Page: /Avc_test_and_repair/Gemini_903.shtml

Last modified: Monday, 20 Jul 2026