1965 Super Reverb AB763 with too early distortion

[Apexelectric]

Been wrestling with a real 65 Blackface Super that just doesn’t want to behave no matter what I do. It isn’t as clean and as loud as it should be and I’ve been through the whole amp and can’t find the culprit.

The amp sounds good on both channels at lower volumes but when you drive it past 4 on the volume level, it starts clipping and doesn’t get any louder. If anything it looses a little volume. The amp came to me with this issue and a variety of others. Someone had messed with it before I got it. I replaced all the filters and most of the tone stack caps and coupling caps as a lot were non stock values. The tubes have all been swapped for known good ones, resistors have been tested and some swapped that were in question, a new OT has been tried out as well. All voltages seem stable with only about a 20V difference from idle to under load in the B+. More distortion is present on the vibrato channel than the normal channel but what you would expect from the usual difference in the two. They both exhibit the volume loss. I can turn this up to 10 while being next to it without it being uncomfortable so it’s not what it should be.

The clipping is pleasing and the amp just seems to sound like it’s going through an attenuator. There is a quite a bit more distortion that you would get from a Super turned up though.

Looking for a little guidance on how I might be able to troubleshoot this issue further as I’m running out of ideas at this point and a little stumped.

[sluckey]

Speakers?

[Apexelectric]

Bench speaker, dummy load or the 4 x 10 cab have the same result

[dude]

Properly biased?

[pdf64]

Bench speaker, dummy load or the 4 x 10 cab have the same result

What is that result, eg power output at the onset of clipping?

[Apexelectric]

The output is properly biased and the clipping occurs at only 11-12W output to the speaker.

[Apexelectric]

Bench speaker, dummy load or the 4 x 10 cab have the same result

What is that result, eg power output at the onset of clipping?

Correction, Puts out about 10V RMS into a 4 ohm dummy load before it starts clipping. This is about 4-1/2 on the volume pot. Volume level doesn’t increase after that, just the output distortion. With the signal generator at 1K the plate dissipation hits about 50ma at the onset of clipping and increases from there. The plate dissipation at idle is about 35ma

[sluckey]

Correction, Puts out about 10V RMS into a 4 ohm dummy load before it starts clipping. This is about 4-1/2 on the volume pot.

That's 25W. Should be plenty loud.

[Apexelectric]

Correction, Puts out about 10V RMS into a 4 ohm dummy load before it starts clipping. This is about 4-1/2 on the volume pot.

That's 25W. Should be plenty loud.

Loud enough for normal inside playing but not loud enough for this amp. I shouldn’t be able to stand right next to the speaker with ths amp dimed. There is most certainly an issue.

Just wondering what not so obvious of an issue I could be dealing with? Power supply seems stable, OT was subbed out for a new one with no change, all the components test fine as do the tubes, so what’s left? Grounding? I played with that as well with no changes. Oscillations? What would be a good way to test that theory if it’s viable?

[sluckey]

If you can't hear any oscillations then an oscilloscope is the easiest way to tell if there is oscillation.

Are the speakers connected to the correct jack? Surely they are. Just grasping straws since you seem to have covered everything.

You never said anything about biasing. What are the voltages on the 6L6s pins 3, 4, and 5?

[Apexelectric]

If you can't hear any oscillations then an oscilloscope is the easiest way to tell if there is oscillation.

Are the speakers connected to the correct jack? Surely they are. Just grasping straws since you seem to have covered everything.

You never said anything about biasing. What are the voltages on the 6L6s pins 3, 4, and 5?

It’s biased at 35ma at idle and plate and screen voltage hover around 450V at idle and drop a bit , as you would expect to 420ish when the output is pushed hard and drawing around 50ma plus. Away from the amp at the moment and can check specific voltages under specific conditions when I get back. I don’t recall anything odd about the voltages at the grids but can double check that. Other than the early onset distortion, it doesn’t sound bad at all, just uncharacteristic of this amp.

[punkykatt]

What are the voltages of the other tubes?

[pdf64]

Bench speaker, dummy load or the 4 x 10 cab have the same result

What is that result, eg power output at the onset of clipping?

Correction, Puts out about 10V RMS into a 4 ohm dummy load before it starts clipping…

Super Reverb is intended for a 2 ohm load.
Just to note that 10V across 2 ohms is 50W.
I suggest you recheck with a 2 ohm load.
It probably won’t manage 10V, but I suspect the power output will look much more reasonable.

[Apexelectric]

Bench speaker, dummy load or the 4 x 10 cab have the same result

What is that result, eg power output at the onset of clipping?

Correction, Puts out about 10V RMS into a 4 ohm dummy load before it starts clipping…

Super Reverb is intended for a 2 ohm load.
Just to note that 10V across 2 ohms is 50W.
I suggest you recheck with a 2 ohm load.
It probably won’t manage 10V, but I suspect the power output will look much more reasonable.

My bench top dummy load is 4-8-16 ohm so I saw no harm in using the 4 ohm setting. I realize it wants to optimally see a 2 ohm load but there’s no denying the issue is present, regardless of the load it sees. The 4x10 speakers in the combo cabinet and my 8 ohm bench speaker yield the same audible result as the scope shows with the dummy load.

[Apexelectric]

Bench speaker, dummy load or the 4 x 10 cab have the same result

What is that result, eg power output at the onset of clipping?

Correction, Puts out about 10V RMS into a 4 ohm dummy load before it starts clipping…

Super Reverb is intended for a 2 ohm load.
Just to note that 10V across 2 ohms is 50W.
I suggest you recheck with a 2 ohm load.
It probably won’t manage 10V, but I suspect the power output will look much more reasonable.

I rigged up a 2 ohm load and it managed 8V so about 32W before clipping so it seems to behave fairly normally it the undistorted range but when it starts clipping, it ceases to increase in volume past about 4-1/2 on the dial and just gets more and more distorted and between 8-10 on the volume range, the volume actually drops.

[Apexelectric]

What are the voltages of the other tubes?

I went through the whole preamp and checked voltages, couldn’t find anything out of the ordinary. Both channels have the same issue so it appears to be a phase inverter/output issue. Focused on that half and couldn’t find any leakage or voltage abnormalities

[Apexelectric]

Picked up my near mint 66 Super from the shop today. Will do a side by side comparison and see if I can find any differences in voltages and see what the scope says too regarding the output and when it starts clipping and what the undistorted output maximum is. Will report back with my findings.

[pdf64]

If there is an issue, my guess is that it may be in regard of the speakers.
eg low sensitivity due to small alnico magnets, bad corroded / oxidised connections.

[Apexelectric]

They both seem to behave similarly. Don’t think I’ve ever tried to turn my other one up past 5 since it was plenty loud and sounded great. But the OT on these amps must saturate pretty easily once they get up around 5 on the dial. I wouldn’t have believed it but it’s not as loud of an amp as I believed it was.

[tubeswell]

You say it’s been tinkered with extensively. Are the coupling caps stock values? What about bypass caps?


(I had a 66 Super Reverb in a couple of years back where someone had subbed a 2nF coupling cap in the preamp, killed all the bass, put the stock cap back in, issue was fixed. Not saying that’s your problem- just a stab)

[Apexelectric]

You say it’s been tinkered with extensively. Are the coupling caps stock values? What about bypass caps?

Someone else tinkered with it, I put it back to stock values. Much better tone now. My customer said he was surprised that it wasn’t louder than it was compared to his Blackface Bassman. I told him that it probably had to do with the beefier output transformer but I was questioning it too. After now comparing it to a known stock, unmolested, one, I realized that it’s actually normal. Still seems odd to me but I’ve always played them clean and never really had a need to turn it up loud enough to get it to compress as much as it does past 5 on the dial.

(I had a 66 Super Reverb in a couple of years back where someone had subbed a 2nF coupling cap in the preamp, killed all the bass, put the stock cap back in, issue was fixed. Not saying that’s your problem- just a stab)

[pdf64]

They both seem to behave similarly. Don’t think I’ve ever tried to turn my other one up past 5 since it was plenty loud and sounded great. But the OT on these amps must saturate pretty easily once they get up around 5 on the dial. I wouldn’t have believed it but it’s not as loud of an amp as I believed it was.

Why ascribe the clipping to OT saturation?
There’s a lot of nonsense on the internet about OT saturation. The output pentodes we use are incapable of drawing enough HT current for it to happen, anode resistance is too high.
The clipping will either be grid or anode current limiting, most likely the former. Most likely the former, but it can be verified using a scope.

Edit- just to note that the 022855 Super Reverb OT is about twice the mass of the 022848 used in Bandmaster, Vibrolux etc models.

[tubeswell]

What B+ voltage have you got?

[shooter]

It’s biased at 35ma at idle and plate and screen voltage hover around 450V at idle and drop a bit , as you would expect to 420ish when the output is pushed hard and drawing around 50ma plus.

[Apexelectric]

They both seem to behave similarly. Don’t think I’ve ever tried to turn my other one up past 5 since it was plenty loud and sounded great. But the OT on these amps must saturate pretty easily once they get up around 5 on the dial. I wouldn’t have believed it but it’s not as loud of an amp as I believed it was.

Why ascribe the clipping to OT saturation?
There’s a lot of nonsense on the internet about OT saturation. The output pentodes we use are incapable of drawing enough HT current for it to happen, anode resistance is too high.
The clipping will either be grid or anode current limiting, most likely the former. Most likely the former, but it can be verified using a scope.

Edit- just to note that the 022855 Super Reverb OT is about twice the mass of the 022848 used in Bandmaster, Vibrolux etc models.

Any reason for this behavior in a Super more so than another Blackface fender, like a Twin?

[Williamblake]

So maybe the output stage is doing fine after all? What is going on? I do very much like to have a spreadsheet where i can enter all the relevant voltages of all tubes of an amp. Like every tube, every pin. Every power node. Also enter the plate and cathode and dropping resistors. Have your spreadsheet calculate the currents. Sounds like a lot of work? Well no it is not that much work and it is a great help to get an overview. Unrelated example attached. Sounds like a lot of footwork but the benefit being you only have to do the footwork once. To whom it may concern. Just noticing you cannot post .ods on this forum, the .pdf is an export from that spreadsheet.

[pdf64]

…
Any reason for this behavior in a Super more so than another Blackface fender, like a Twin?

Maybe the HT is a bit saggier, due to the valve rectifier?
You mentioned that when measured, it didn’t sag much though 

FYI the usual cause of volume dropping when heavily overdriven is bias shift, leading to the early stage of blocking distortion.

The effect can be reduced by reducing the value of the coupling caps between V6 anodes and the output valve control grids, eg try 47nF.
Reducing the C HT node voltage a bit will help too, eg try increasing the 1k dropper to 2k2.

[PRR]

....Just noticing you cannot post .ods on this forum...

.ZIP it.

FWIW, I do not know what a .ods is. OK, now I know, but a single amp tube-list does not need all that frillery.

Most spreadsheets can export as .XLS{x}, the (ug!) Microsoft number mangler & spreader. Going way back: Lotus 1-2-3 used to rule the world (after it slew VisiCalc .VC), that would be .123 or WKS. There is no math in there so there is always dBase II. These all open in my eXcel 2003; I would bet Mac and Open/Free/Office opens them. Any of these probably still needs to be ZIPed to get past the forum upload filter; if you do it a LOT then Doug can add it to the list.

[shooter]

or just screenshot and post a .jpg

it the undistorted range but when it starts clipping,

Is that from a scope or ears?
IF scope;
what does the signal look like feeding the pi? (TS @ 5), adjust gain/volume til  the signal distorts, what numbers did they land on?

[Apexelectric]

or just screenshot and post a .jpg

it the undistorted range but when it starts clipping,

Is that from a scope or ears?
IF scope;
what does the signal look like feeding the pi? (TS @ 5), adjust gain/volume til  the signal distorts, what numbers did they land on?

Actually both, signal looks pretty normal coming out of the phase inverter so almost all of the distortion is coming from the output tubes. I don’t have the amp on the bench any more but there was some distortion in the signal from the PI when you got past 6 or 7 but by then the output was in fairly heavy clipping.

[tubeswell]

Is it a completely stock circuit?

[Apexelectric]

Is it a completely stock circuit?

All stock values, yes.

[Apexelectric]

…
Any reason for this behavior in a Super more so than another Blackface fender, like a Twin?

Maybe the HT is a bit saggier, due to the valve rectifier?
You mentioned that when measured, it didn’t sag much though 

FYI the usual cause of volume dropping when heavily overdriven is bias shift, leading to the early stage of blocking distortion.

The effect can be reduced by reducing the value of the coupling caps between V6 anodes and the output valve control grids, eg try 47nF.
Reducing the C HT node voltage a bit will help too, eg try increasing the 1k dropper to 2k2.

Thanks for that, helpful info!