Fender SF Twin Reverb Hum

[Helmholtz]

But if those resistors were removed... wouldn't the caps still get the same 50% voltage?

Not necessarily.
Equilibrium cap voltage would depend on their DC leakage resistance which can't be expected to be exactly same.

[JAelec]

I re-tested the paralleling a 100uF capacitance on the filter caps today and although the noise is lowered, I found that it did not make that much difference as I remembered from last time. After using the tube matching pot + new tubes on a noisy Bassman 70 (see another one of my threads) I realised that this SF Twin reverb could be much quieter.

Rotating the bias balance pot inside the chassis, I got the old tubes to be dead quiet.

But I noticed that the pot had to be in fully clockwise position for the tubes to be matched/quiet:



I tested again with a brand new set of quartet-matched tubes and then the pot was approximately in the middle when matched/quiet:



My interpretation of this is:

The two pairs of old tubes are severely mismatched and by luck the pot is able to match the pairs in fully clockwise position. Is this correct?

So should I install new tube for this amp or is it ok to just match them with the bias balance... I have a feeling that if the tubes become any more mismatched in the same direction then the pot won't be able to compensate.

[g1]

The two pairs of old tubes are severely mismatched and by luck the pot is able to match the pairs in fully clockwise position. Is this correct?

So should I install new tube for this amp or is it ok to just match them with the bias balance... I have a feeling that if the tubes become any more mismatched in the same direction then the pot won't be able to compensate.

Yes, this is correct. You may be able to shuffle the tubes around to find a way that is matched more in the middle of the pot. Try a pair of tubes at a time and see if you can find pairs that are somewhat close.

[JAelec]

Yes, this is correct. You may be able to shuffle the tubes around to find a way that is matched more in the middle of the pot. Try a pair of tubes at a time and see if you can find pairs that are somewhat close.

That's a good idea, I'll try that... 4 new 6L6s are expensive.

Could I check the bias on the individual tubes somehow to see if one is off? Transformer method will only work on each pair I think

[g1]

You can do one at a time with transformer method. It will tell you which are running hot relative to the others.
Use the same socket every time.

[JAelec]

Do I understand correctly, pull all power tubes except the one I’m testing, lets say its V7, so only one power tube will be in the amp. Test single power tube w transformer method. Remove tube then place the next tube to be tested in socket V7.

Is this correct?

Will a single tube draw more, less or same current compared to when all tubes are mounted in the amp or is every tube its own entity so to speak?

[g1]

Yes, that is the method.
I would guess it will give a higher value when on it's own, but you can compare the numbers for matching purposes.

[JAelec]

I did some measurements today, with a single 6L6 at a time, going through all four of the old tubes, one by one installed in socket V7.

I did two "rounds" of voltage measurements. I had a dummy load connected.

As you can see from the first round of measurements there was 0V from plate to rectifier on tube #1. Not sure what this means, but in the second measurement round pin 3 got extremely hot... I noticed the solder on pin 3 melting while I was poking it with my multimeter probe... the joint started moving and I immediately shut the amp off. It started to smell like burnt plastic and smoke was coming from pin 3.

After the first round of measurements I decided to tighten up the sockets as the tubes were very loose in their sockets. I don't know if this was a bad idea because after this tube #1 made pin 3 melt the solder of its joint and the plate dissipation has increased to max for the rest of the tubes. That is if I have made my calculations correctly... if you guys have the time to look through the numbers it would be so kind of you!


First round was before tightening the octal socket pin receivers:

6L6 Old Tubes, # ---->	1	2	3	4
Plate to ground	485.3	477.7	477.7	478.7
Rectifier to ground	484.5	479.9	480.9	480.8
Voltage drop measured plate to rectifier	0.000	1.49	1.38	1.37
Voltage drop calculated	0.8	2.2	3.2	2.1

Pin 3/Blue transformer lead to OT- primary center tap: 37.5 Ohms (V7 and V8 side)
Pin 3/Brown transformer lead to OT primary centre tap: 35.1 Ohms (V9 and V10 side)

Plate current=voltage drop / resistance

0.000/37.6 = 0 mA
1.49/37.5 = 39.7 mA
1.38/37.5 = 36.8 mA
1.37/37.5 = 36.5 mA

Plate dissipation= plate voltage*plate current

485.3*0 = 0 W
477.7*0.0397=18.96 W
477.7*0.0368=17.58 W
478.7*0.0365=17.47 W

18/30 = 0.6 = 60 % of max PD

Tube #	1	2	3	4
Plate current (mA)	0	39.7	36.8	36.5
Plate dissipation (W)	0	18.96	17.58	17.47

Second round, after tightening the octal socket pin receivers:

6L6 Old tubes, # ----->	1	2	3	4
Plate to ground	---	466	467.6	466.4
Rectifier to ground	---	467.5	469.7	468.5
Voltage drop measured plate to rectifier	---	2.60	2.56	2.55
Voltage drop calculated	---	1.5	2.1	2.1

Pin 3/Blue transformer lead to OT- primary center tap: 38.4 Ohms (V7 and V8 side)
Pin 3/Brown transformer lead to OT primary centre tap: 35.9 Ohms (V9 and V10 side)

Plate current=voltage drop / resistance

2.6/38.4 = 67.7 mA
2.56/38.4 = 66.6 mA
2.55/38.4 = 66.4 mA

Plate dissipation= plate voltage*plate current

466*0,0677 = 31.5 W
467.6*0,0667 = 31.2 W
466.4*0,0664 = 31.0 W

31/30 = 1.03 = 103 % of max PD

Old Tube # ---->	1	2	3	4
Plate current (mA)	----	67.7	66.6	66.4
Plate dissipation (W)		31.5	31.2	31.0

New tube # --->	1
Plate to ground	464.0
Rectifier to ground	465.6
Voltage drop measured plate to rectifier	2.8
Voltage drop calculated	1.6

Plate current new tube: 2.8/38.4 = 72.9 mA
Plate dissipation new tube: 464*0,0729 = 33,83 W

34/30 = 1.13 = 113 % of max PD


I'm quite confused as to what happened...perhaps I tightened the sockets in a bad way? Or something else in the amp is wrong

[g1]

I'm quite confused as to what happened...perhaps I tightened the sockets in a bad way? Or something else in the amp is wrong

I think the solder joint at pin3 is bad and causing some arcing. Remove the old solder and re-solder. Do the same for pin3 of all the power tube sockets.
Examine both sides of V7 socket for any signs of burning. If there are any, the socket should be replaced.

Sounds like one of the original tubes was pretty much dead, which is why the matching pot needed to be so far from center.
When you do the measurement, you need to do it direct, not by calculation (subtraction). The voltages will be slightly changing constantly.
Put one meter probe at the rectifier connection and the other probe at pin3. Then you will have direct measurement of the voltage drop across the OT.

[JAelec]

The values that I am using in calculations above are from direct measurement pin 3 to rectifier, but thanks for clarifying I was a little unshure of what was the best approach.

I’ll try your suggestions.

What about the plate dissipation and how it suddenly increased with 40-50%

Is that because of the solder joints arcing on pin 3?

[JAelec]

I noticed some resistors in under the sockets have look a bit strange… would this be some indication of something?

[JAelec]

Here you can see pin three of V7 and the black spot on pin three left after the pin started to smoke

[Tom Phillips]

I noticed some resistors in under the sockets have look a bit strange… would this be some indication of something?

Your photos show residual old flux on the end of the screen resistors. This is not unusual. It is not something bubbling out of the resistor and does not indicate a resistor failure.

[Helmholtz]

I don't like the solder joints of the screen resistors.
The wire insulation should never get into the solder.
Some insulation types like PVC release acid when heated, giving rise to corrosion.
Also the big solder blobs extend to the body of the resistor.
Means that the resistor was exposed to ecessive heat during soldering.
This in turn might have increased resistor value and/or might have caused internal cracks resulting in unstable resistance.

[JAelec]

Your photos show residual old flux on the end of the screen resistors. This is not unusual. It is not something bubbling out of the resistor and does not indicate a resistor failure.

Thanks for this info... so it's just flux then