5E3: Terrible Sounding Distortion

[JohnTheConqueror]

So before I start, I'd like to mention that I am a complete n00b when it comes to building tube amps (please don't make fun of me...I had to start somewhere). I've lurked around here for a while trying to learn as much as possible from people a lot more knowledgeable than myself. I just completed my first build (a 5E3 deluxe kit) and I need some help.

Initially I thought the amp sounded great. My living situation doesn't permit me to play loudly very often, so the volume knob was restricted to about 3 or so. Several days ago I took it over to a buddy's house and was finally able to crank it, and immediately something didn't sound quite right. With the volume past about 4 or 5, it starts to break up in a really strange way. I'm not exactly sure how to describe it, but the distortion is extremely fuzzy, and the tone has a strange inflection (for lack of a better term, I suppose) sort of like an "eeeeeooow" sound, if that makes any sense. This effect is more pronounced with more forceful picking, and the tone returns to a more sonically pleasant region as the note decays.

I had hoped that maybe this was just an effect of extreme clipping somewhere in the preamp, but I checked the bias of all the tubes, all of which seem to be operating at reasonable quiescent points (I say reasonable because the schematic I had didn't list any voltages, so I'm not entirely sure what B+, Vk, etc. is supposed to be for any of the tubes). I checked the output of each preamp stage with an oscilloscope, and everything looked relatively clean up until the phase inverter. The waveform at the phase inverter looked extremely similar to that described on this page of Merlin B's site... The Valve Wizard
(look at the section titled "avoiding unpleasant overdrive tones"). He suggested adding a large grid stopper resistor at the grid of the cathodyne, which I did. Unfortunately, while this cleaned up the output of the cathodyne significantly, the strange sounding overdrive persisted.

The only other thing I could really think of to do was move around some heater wires, in the hope that somehow AC ripple was getting into the signal path or something. This didn't work (I don't know...like I said I'm a n00b). Hopefully someone out there has heard an amp with similar problems and can suggest a possible solution, or at least a clue as to where to look.

Thanks in advance for your help. And I apologize for the huge block of text...I realize I'm long-winded.

[Chuck H]

It sounds like you have a parasitic oscillation. Lead dress and grounding are critical in amps that get their distortion by overdriving. One problem with hunting down parasitics is that the stage affected may not be where the problem is. It could, for example, be that the grid lead for that stage is fine but some other lead is in close proximity and "talking" to it. Same with grounds. If two things are sharing a ground that really shouldn't there could be an AC decoupling issue causing oscillation on one stage when it's another stage that should be grounded elsewhere. Read up here and abroad about layout and grounding and consider how it related to what you have built. I'll assume you followed some kind of diagram but that doesn't guarantee success. Best to understand the principals at work so you can avoid or correct problems.

Another thing it may be is crossover distortion. If the PI is driving the power tubes too hard the power tube grids will try to draw current (grid conduction). The time constant of the PI circuit is typically too high to allow the grids to draw this current or discharge the voltage so it appears to the tube as a bias shift causing the tube to go into cutoff. Smaller PI coupling caps can help. So can larger power tube grid resistors.

That or if the power tubes are actually biased too cold you can get cutoff distortion. A smaller value cathode resistor will bias the tubes hotter. A 5E3 is often biased real close to class A (100% dissapation) but pay attention to the tubes dissapation limits.

Another thing that comes to mind is that all stock 5e3's have the potential to allow too much bass through the circuit for good overdrive. Too much bass when overdriving can cause the above mentioned grid conduction as well as just sound really fizzy. Many, many 5e3 owners do circuit mods to increase upper mids and/or reduce bass through the circuit to achieve better overdrive. This can be done with smaller coupling caps between stages and smaller preamp cathode bypass caps.

My first attack would be to look for parasitic oscillations. Even though it's easier to just start reducing capacitor values. You could end up reducing bass response when it wasn't needed unless you know you don't have an oscillation or crossover distortion.

[JohnTheConqueror]

I had read a few posts on this and other forums regarding oscillation but as I have a limited amount of experience building amplifiers, I wasn't entirely sure that's what I was hearing. Since it appears that that is probably the issue, I'll go ahead and re-do all the wiring. It's a bit of a drastic step, but in hindsight I was slightly more relaxed with the layout of the wiring than maybe I should have been (the inside of the chassis looks more chaotic than I'd like).

I feel fairly confident that I know where/how to ground the PT center taps, safety ground, filter caps, etc., to avoid oscillation, but are there any rules of thumb you can think of for the preamp as far as what can/cannot be grounded together? Also, considering that the sleeves of the input jacks are physically/electrically connected to the chassis, is it necessary to actually ground these with a wire? Many layouts I've seen have the input jacks grounded to a solder terminal somewhere in close proximity, or connected to the pots.

I checked the output at the speaker jack and didn't see anything that would indicate the problem is crossover distortion (that being said, I'm still not completely comfortable with the theoretical operation of push/pull amplifier circuits, so I will have to do some research and make sure the output tubes are biased correctly). However, it does appear that the phase inverter is being driven extremely hard; the output is starting to look like a square-wave at high volume. Right now the plate resistors for the first two gain stages are 100k. Correct me if I'm wrong, but wouldn't reducing these to 68k or so reduce the voltage swing on the plates and give me a little more headroom for the phase inverter? I think I calculated a gain of maybe 25V/V or so for the 12AY7 at it's current operating point. I realize some of this is attenuated by the insertion loss of the tone stack (although there's not much to it in this amp) but considering a guitar is capable of producing a signal of a few volts, it's certainly enough to deliver a signal that would cause the next stage, and subsequently the PI, to clip pretty heavily. Not that this is of too much concern right now, but I'd like to avoid overdriving the preamp as much as possible.

[MWJB]

Search the thread "5E3 inadequate voltages", OP is is Strat 57. It has voltages plus tips on layout.

[jrfrond]

Sounds like good ol' blocking distortion to me, and happens when the grids of the power tubes are driven positive.

What is happening is that the time constant of the high-pass filter formed by the 220K grid leak resistors and .1uF coupling caps before the power tubes is too high. When presented with a constant signal with a significant duty cycle (i.e. clipping), the power tube grids have no time to recover and begin to charge positive, the result being some really nasty, alien-like distortion. To complicate matters, this is something you won't notice until you drive the amp hard.

You need to lower the RC time constant of the high-pass filter, either by lowering the cap value, resistor value or both. I've seen this before with .1uF coupling caps. I will drop the value to .056 or .068, possibly even .047 or .02, depending on the player's requirements. Or, you can drop the value of the grid leak resistors. Or you can reduce BOTH slightly. Either way works. The trick is to reduce the values only enough to reduce blocking distortion, while still maintaining good bass reproduction.

[JohnTheConqueror]

I will first try to lower the PI filter caps as jrfrond suggested, simply because it's an easy fix. If that doesn't work then it looks like I'll be spending the better part of the day re-thinking my grounding scheme. Wish me luck. Thank you everyone for your input.

[Chuck H]

I will first try to lower the PI filter caps as jrfrond suggested

After Chuck H suggested it As noted, smaller grid bias resistors and larger grid stop resistors also help. As does reducing overal bass through the rest of the amp.

[jrfrond]

After Chuck H suggested it As noted, smaller grid bias resistors and larger grid stop resistors also help. As does reducing overal bass through the rest of the amp.

Oops! Sorry Chuck. Didn't mean to hijack your suggestions. Mea Culpa. I shoulda read the whole thread.

[Chuck H]

Nothing on you John. Sometimes more than one person has to say something before it seems valid. I was just noting that the OP disregarded me.

[JohnTheConqueror]

My appologies, Chuck. I didn't mean to devalue your input. I had no idea how to fix this issue before posting, so all advice is greatly appreciated. I actually ended up desoldering everything and starting from scratch. Perhaps PI filter caps/power tube grid stoppers are the cause of the problem, but lead dress and a vastly inadequate grounding scheme certainly aren't helping anything. If the problem persists after these issues are resolved then I'll worry about PI caps. At least I will have eliminated one possibility and improved my grounding scheme in the process.

[jrfrond]

My appologies, Chuck. I didn't mean to devalue your input. I had no idea how to fix this issue before posting, so all advice is greatly appreciated. I actually ended up desoldering everything and starting from scratch. Perhaps PI filter caps/power tube grid stoppers are the cause of the problem, but lead dress and a vastly inadequate grounding scheme certainly aren't helping anything. If the problem persists after these issues are resolved then I'll worry about PI caps. At least I will have eliminated one possibility and improved my grounding scheme in the process.

Lead dress might contribute to oscillation, and improper grounding will often manifest itself as hum or buzz, but improving either or both of those issues won't help blocking distortion one bit. Also, it is the PI COUPLING caps you need to address, not filter caps.

You can try this very easily BEFORE you do any cleanup work. I'm all for neat work and proper grounding, but always try the simplest solution first.

[JohnTheConqueror]

You can try this very easily BEFORE you do any cleanup work. I'm all for neat work and proper grounding, but always try the simplest solution first.

Point duly noted. Unfortunately, I'm a bit obsessive about neatess and the lead dress/grounding was bothering me so I just decided to deal with it first.

I replaced all the .1uF coupling caps with .022uF, reduced the value of the grid leak resistors slightly, and put in larger grid stoppers on the power tubes. This helped a little, but the strange distortion still occurs with the volume more than about halfway up. This is what the signal looks like at the output jack at full volume. The input signal is a 200Hz, 1v p-p sine wave.



I'll be the first to admit, I'm not entirely sure what I'm looking for on the output, but this doesn't look quite right to me. Any further thoughts or suggestions would be greatly appreciated.

[MWJB]

"This helped a little, but the strange distortion still occurs with the volume more than about halfway up." Sorry, are you saying that you expect the amp to stay clean up to full sweep on the volume pot, or thereabouts?

At what output Wattage can you keep a good sine wave with 100mVAC on the input, vol & tone controls set 12 o'clock?

I think we need to see some pictures of the build.

[GregS]

A 1V p-p input signal is extremely large for a guitar. A typical humbucker guitar has a signal around 150 mVAC when played hard. I expect your pre-amp is being overdriven during your oscilloscope test. That waveshape is probably originating there and accumulating throughout the rest of the amp. I wouldn't assume that all of that is occurring in the power amp section.

[JohnTheConqueror]

Sorry, are you saying that you expect the amp to stay clean up to full sweep on the volume pot, or thereabouts?

No, not at all...but the type of distortion I'm getting sounds terrible. I was just wondering if it was normal for the output to look like that, or if that shape was indicative of crossover and/or blocking distortion. I can't say I know exactly what blocking distortion sounds like or looks like on an o-scope, and I was hoping that someone whos has more experience troubleshooting tube amps than me could definitively say whether or not that is the problem.

I'll be able to take some measurements this weekend and I'll get back to you with pictures and regarding your question about output wattage. Thank you for your input.