Output tube matching pot on ultralinear fender super reverb (and fried screen grid resistors)

[judge734]

I just picked up a 1979 Super Reverb off Facebook for cheap. I used it for a few weeks never turning the volume above 3 and it sounded great except there was some background hiss and hum that I couldn't get to go away with swapping tubes and using the hum balance and the output tube matching pots. It also seemed to distort some bass notes even with the volume turned down. The output tube matching pot did help but didn't get rid of everything. Also the tremolo did not work.

On opening it up I saw it was basically all original with no mods inside. I decided to make the following changes

I went ahead and swapped the master volume push pull for plain master volume pot, and got rid of all the excess wire for the push pull circuit (the push pull pot was very scratchy and noisy and I already had a plain 1M pot handy).

I put everything back to together went online and ordered some parts. Everything worked fine.
I did the following after my antique electronics order came in

added cathode bypass e-cap and 2.2k resistor on reverb driver in place of silverface 680 ohm reverb resistor.
replaced all cathode bypass e-caps
replaced trem roach.
replaced all e-caps under doghouse. (reservoir caps, filter caps, bias caps) I used all the same values for all the caps except I upped the bias caps from 70uf/100v to 100uf/100v because they were cheap and available.
I took off one component at a time and replaced it so I couldn't get confused about what goes where.

I powered the amp up after making changes (I should have used a current limiter but I didn't) with all controls on 0 it made a loud pop and hum and blew the fuse and I smelled smoke.

On inspection the two 470 ohm screen grid resistors were burnt. I decided to trace all the wires and verify all the changes I made. Everything I hooked up looks to be done correctly. I have quadruple checked the orientation of all the e-caps and they are right. I also checked values of all existing components in the power and bias circuits.

I did find a few bad parts. The 30k 20w resistor from B+ to ground is open. Also coming from the legs of the output tube matching pot there is a 47k and a 68k resistor in line to pin 5 of the power tubes. Those have both drifted the 47k is 52k and the 68k is 95k.

My theory is that with the 30k 20w resistor open (I assume it was already open when I bought the amp) and all new filter caps the B+ was running higher than it had in years and possibly higher than designed because there was no shunt to ground and no leaky old caps. Then with value of the resistors in the tube balance circuit so far drifted the bias on the output tubes was way off. I assume the the screen grid resistors may have already been failing because of the distorted bass notes. I assume because of the blown screen grid resistors and the fact it made a loud pop and hum with the master on 0 that the issue is on the power side, not the preamp. Perhaps because I raised the bias caps to 100uf instead of 70 they were charging slowly so when I powered up the bias was not fully charged?

So my plan is to replace the screen grids, the 20K 30w and the 47k and a 68k resistors from the balance pot, swap the old 70uf caps back in place (they were working when I bought the amp) Then power up with a current limiter and no tubes and test voltages.

I got to thinking about the balance pot though and how it works. It does not seem like you could balance the bias all the way with it? because it is a 10 k pot and the center is hooked to the bias voltage and the legs are hooked to 47k and 68k. It seems to me that if you had the pot swept all the way to one side then you would have 47K + 10K < 68K and that would be the closest to balanced that it could be. However, it is designed that way and works fine in most amps so what am I not understanding about that circuit? Does anyone have anything else I should check? Are there any other theories as to what is wrong?

Here is a link to the schematic
https://www.thetubestore.com/lib/the...-Schematic.pdf

Thanks!

[Helmholtz]

Bias circuit:

- difference between 100µ and 70µ is insignificant, modern equivalent to 70µ is 68µ. But correct polarity is essential: positive end goes to ground. If ever operated with wrong polarity chances are to cap is bad (leaky).

- different bias feed resistors (47k/68k). These resistors don't carry a DC current, so they don't drop voltage and have no influence on grid bias.
They do load the PI outputs and the different values are likely chosen to improve PI balance.

- pull power tubes and measure grid (pin 5) voltages at the sockets.

You might need new power tubes.

[judge734]

Thank you Helmholtz. Yes I am very curious what the voltage at pin 5 is. It may be about a week before I can measure it because I want to replace the 30K 20W wire wound resistor before I turn it back on. None of the places I normally order from seem to have it...

[judge734]

FYI here is a view under the doghouse and a reading from the bias caps to ground. Had to double check to verify the bias caps.

[g1]

27K or 33K would be the common values for the resistor. Either should be close enough. Or you could go two 15K 10W in series.
Agree with Helmholtz the power tubes are suspect and probably most likely cause of fuse blowing. Burning smell was probably screen resistor(s).
Recommend keeping new bias filter caps.

If still hum issues, make sure hum balance pot and associated resistors are good, fairly common to find the hum pot burnt/open.

[judge734]

Ok reporting back in. First off the amp is “working” again. But it still has some ghosting and distorts bass notes even at very low volume. It was doing that before I started any repairs as well so I don’t think it is anything I did.

here is what I have done so far.

replaced All electrolytic caps. (Filter caps, bias caps, cathode bypass caps etc.)
replaced the 30k 20w resistor.
replaced the 470 screen grid resistors.
added 22k resistor + 20k trim pot to back of tube balance pot.
replaced 47k and 68k resistors after tube balance pot.
resoldered ground connections by power transformer bolts.
added 1 ohm resistors for checking bias.
removed push pull boost on master volume


I am still plagued by a nasty ghosting or inner modulation distortion on mostly low notes but a few other spots on the neck too. Even with the volume most of the way down. The reverb is also got a decent amount of hum to it actually the reverb is humming worse now than when is started fixing things. I pulled and swapped all the preamp tubes and eliminated them as a culprit. I have not tried any other power tubes because I don’t have another set of 6l6 tubes.The issue is in both channels so I assume it is in the phase inverter or later.

I am wondering if maybe it is a coupling cap leaking dc? Or perhaps it is a conductive board? I did measure voltage on the board and found a few spots with 250 to 350 millivolts is that enough to cause a problem? Also I measured coupling caps to see if they were blocking dc to the next stage. I found a few of them where one side of the cap had around 250 volts and the other side had 150 or 200 millivolts. Is there another way to test caps that is better or to diagnose leaking coupling caps.

[Helmholtz]

Did you verify schematic voltages (plates, cathodes, power tube screens and grids)?
Did you measure 6L6 idle currents?

Hum/ghosting is not likely caused by leaky coupling caps.

Do you have a scope?

[judge734]

Unfortunately I don’t have a scope. I have used one in the past so I have a pretty good idea how to trace the signal with one. But I haven’t owned one in years.

I did measure all the plate voltages and they are super close to the schematic.

I did not measure the mA on the tubes. I had issues with the 1 ohm resistors. They are supposed to be 5% tolerance but one of them is reading 1.3 ohm and the other is 1.2 ohm after installing them. My meter only reads one decimal place on resistance measurements so with the resistor value being sort of unknown past the first decimal a voltage reading there would not give me enough information to accurately determine the mA.

I am going to look at it again tonight and measure it with the voltage drop on the transformer and transformer resistance.

In the meantime I just adjusted the bias trimmer setup to 33k so it would match the schematic. That gives me about -56 v at the pins. (I forgot to write it down but it was about that)

maybe I am describing it wrong it sounds almost like you split the signal and put a trashy fuzz pedal on the bass notes. Then mixed that back with the clean signal. I keep listening to YouTube audio clips about blocking distortion/ inner modulation distortion/ crossover distortion etc. to see if I can identify it by sound.

There is a hum but only when I turn up the reverb. It was not there before. I am wondering if that could have been caused by the changes I made to the reverb driver cathode.

[Helmholtz]

Well two 5% resistors can be 10% apart, so it's best to use 1% types.

But you can still use your cathode voltages.
Idle current is given by cathode voltage divided by actual resistance.

As said above, also measure screen voltages and verify PI schematic voltages.

[judge734]

The problem with that is that my meter only reads the resistance to one decimal place. So even though I read 1.2 ohms that could mean anything between 1.15 and 1.24 ohms.

if I measured 40 mV at the cathode and 490 v at the plate the value could be anywhere from

40/1.15=34.78 mA
490 * .03478 = 17.04
17.04/30 = 56.8%

To

40/1.24=32.25 mA
490 * .03225 =15.80
15.80/30 = 52.6%

that is a pretty big spread… but the last decimal place won’t make as much difference measuring the transformer windings.

I don’t know maybe the range on the cathode measurement is close enough?

[Helmholtz]

I think the inaccuracy is acceptable for our troubleshooting.
We're looking for a gross imbalance or very low idle current.
You should be able to balance actual idle currents using the bias pot by adjusting for least hum.

[Tom Phillips]

...that is a pretty big spread; but the last decimal place won't make as much difference measuring the transformer windings .I don't know maybe the range on the cathode measurement is close enough?

That's not really a big spread for a guitar amp and certainly not enough to indicate a significant problem. Note that the Fender schematics usually state that all the specified voltage readings are ±20%.

It is an interesting exercise to measure the voltages and component values of an old amp that is reported to sound spectacular. You will often find anomalies that seem to need correcting but they may actually be the source of the mojo or at least inconsequential. This fact doesn't point to strict rule but supports the requirement to know how to proceed when servicing a guitar amp.

Regarding your meter that only reads one decimal place for resistance measurements. It's also a good idea to check the "zero reading" by shorting the test leads together and noting the reading you get. You will likely find that the reading will not be zero Ohms. What ever reading you get, you can subtract that value from the reading you obtain when you measure the resistor. The is especially important when measuring low value resistors such as your 1 Ohm units.

[judge734]

Thanks for the advice about the multimeter it does read .3 just touching the leads. I actually came up with another way to much more accurately measure the 1 ohm resistors and they are 1.02 and 1.03 ohms. So that’s good enough for me.

I had another chance to work on the amp so I took some of the requested voltages

B+ measured at standby switch 514
at top of 30k resistor 428
at tube balance wiper -59.6
at PI plate 1 272.7
at PI plate 2 275

inner power tube
plate 513
grid 512
screen -59.0
cathode with 1 ohm resistor 19 mV
by my math that’s 9.75 watts or 32%

outter power tube
plate 509
grid 509
forgot to write the screen down
cathode 18.3 mV
9.31 watts 31%

so I changed the bias trimmer because bias in the 30s is too low

it is now inner tube 505 v 32.8 mA
outer tube 505 v 31.4 mA
both of those should be about 55% which is what I think most old fenders were close to when new.

anyway after all that I am still getting the same weird distortion and hum. This time I recorded a short 1 minute YouTube video so you could hear it. It is just with my phone so the audio is not great. But if you listen on headphones you can hear the hum before I start playing, and you can hear the odd distortion in the background when I am holding out notes.

https://youtu.be/XegSmANJD0A

[Helmholtz]

inner power tube
plate 513
grid 512
screen -59.0
cathode with 1 ohm resistor 19 mV
by my math that’s 9.75 watts or 32%

outter power tube
plate 509
grid 509
forgot to write the screen down
cathode 18.3 mV
9.31 watts 31%

Seems you're mixing up grid (G1) and screen (G2).
Grid voltage should be around -56V.