I need advice on how to troubleshoot really loud squeal

[coreysan]

HI everyone! I'm a new amp builder, and I just finished building a Super Reverb kit. I'm using the 8 ohm lead for output,
and connected to a 8 ohm speaker rated at 130 watts (brand new speaker), and all new and matched tubes from
JJ Electronics.

I powered up to standby, and it was great. Heater voltage is at 6.2 volts, and the bias voltage is 4.8 volts. So far
so good. I rolled off all the pots.

The split second I flipped to "on" I got the *loudest* squeal I have every heard!!! Wow it was loud. What can I
check? Do I pull each tube one at a time?

[eleventeen]

You have to be sure the output tranny is properly phased. Correct phasing is generally unpredictable, the idea is to leave the leads (usually brown & blue) long and sloppy, and temp-solder them to the plates of the 6L6s. Pin 3-pin 3. One way will be grossly obviously wrong (causing raucous squeal as you describe; or, sometimes low thumping throbbing aka "motorboating") and the other way will be right. The reason is that the negative feedback used to control the gain structure of the output stage, if out of phase, is POSITIVE feedback aka an oscillator. Just flip the output tranny plate wires, and in fact, do NOTHING ELSE if you have not resolved this point.


On another point, your "4.8 volts bias" is nowhere near correct. There are any number of places to measure this, but pin 5 of either output tube is fine. This voltage wants to be NEGATIVE -45 or -50 volts, absolutely no way a positive voltage. That voltage should show up in standby. Do NOT operate the amp with +5 volts on the output tube grids! You will burn the 6L6s up in VERY short order.

Let me be clear: These are two completely separate faults. They BOTH need to be addressed, and shouldn't be too much of a big deal. Do not turn the amp on with the bias at +5!!

This voltage "turns off" the tube, but it is a necessary control voltage to prevent the output tubes from overheating, overcurrenting, running too hot, however you wish to term it. The LOWER, more negative it is, the more the 6L6 tubes are shut off. Myself, on a brand new build, I like to have bias set to a too-cold level, eg; more like -60 or even -65 volts if it will adjust there. Keeps things cool while I fiddle. 

[coreysan]

Thanks so VERY much for your response!!! I'll definitely switch the OPT wires. Being new to this, I never would have thought
of this on my own, so  I appreciate your insights more than you know!

On the 5v comment, that's totally my fault. I stated it all wrong. My brain was thinking about the 5VAC filament voltage for the
rectifier tube, but I said "bias," and right now I have no idea why I said that!

I'll be more careful how I communicate. In fact, your explanation about bias is equally helpful, so I'm printing that out.

I'll address the OPT wiring first, then triple check the bias voltage.

Just out of curiosity (and you don't need to respond to this if you don't want to) the brown and blue OPT wires for pin 3 match the schematic and layout I have. Is it just the nature of transformers that it can be this way, or did I wire something else in reverse to make this symptom show up?

[shooter]

I believe the answer is, transformers have ""matching"" or "phasing" from input to output.  Sometimes you can see them on the mechanical drawings, usually a dot on primary and secondary.  But then again, I could be all wet with your question

[eleventeen]

"Brown and blue" are the standard colors for the plate leads on an OPT. Just like Green is the standard color for the 6.3 volt winding & yellow is the standard for the 5v and black is the std for the 120 V winding.


Since these are AC, for the most part we simply do not care which way the primary or secondary wires are connected on a POWER tranny.


If I have an old Fender Super and fry the OT and replace it with a real Fender part it's overwhelmingly likely that if I connect the new blue to where the old blue went, I will be fine.


If I buy a kit, it's reasonably possible that the kit vendor one month got a good deal on brand "X" output trannies, sold them all, and 6 months later got a good deal on brand "Y" OT's. Their basing diagram (mounting holes) are the same, the impedances are the same, so they are interchangeable. But he still has the same kit wiring diagram & instructions. Similarly, if you find an old amp and decide to convert it to a guitar amp or salvage the output tranny, you will once again probably have a blue and brown wire which you now know for sure are the plate leads. One of them is called (and you don't need to know this) "start" and one is called "finish". The kit seller, maybe for brand X the blue is the start and the brown is the finish. It could be different on brand "Y". The salvage job, same story. You cannot know and if you DID know and it was wrong you would STILL have to flip them.


Point being, unless you are replacing a specific part with the same specific part, you really cannot predict which is the start and which is the finish. And/or, you may decide to hook up the negative feedback lead to the opposite phase of the phase inverter or phase splitter (assuming push-pull dual output tube config) And so, it (flipping the plate wires) is just one of those things you have to be prepared to do on a fresh build. You could search specs for hours, you could argue with color codes all you want, but you have to make it right and clumsy as it may be, you just have to try it one way and then the other. One way will be grossly obviously wrong. So we want to leave the leads long, not run them all neat and tidy, and fire up the amp to determine which way is right.


If I was Fender and getting my trannies from a known source and I could specify  that "blue" was always the "start", then I could wire up 1000 of them and they would all work fine. If not, it's just a 50/50 chance and there's no point getting exercised about it. If it's wrong, flip 'em. End of story.

It's probably the case that there is an official EIA color code designating "blue" as the "start" and "brown" as the finish (or vice versa) but few seem to pay attention. And again, it it turned out wrong, you'd STILL have to flip them.

[coreysan]

Eleventeen - wow, you're a wealth of knowledge! Thank you very much! It all made sense. I flipped the OPT leads and the amp
works really well now, so thank you a million times.

I measured voltages, and I wanted to show you my measurements just to see if you have any comments or reaction. If not,
that's fine too - I don't want to overstay myself!

Connection                                   My Measurement           Schematic or MGF Stated Measurement
==================            ===============    ============================
 Preamp Heaters                                    6.1 vac                              6.0 vac
 Rectifier Tube Heater                            5.0 vac                               5.0 vac
6L6gc plate voltage                               433 vdc                             465 vdc
ch. 1 preamp plate                                  245 vdc                            270 vdc    (pin 1 only)
ch. 2 preamp plate                                  237 vdc                            270 vdc       ""
tremolo tube plate                                 356 vdc                             390 vdc       ""
PI plate                                                  240 vdc                             230 vdc       ""
Bias voltage                                             -36 vdc

[eleventeen]

Don't worry about asking questions---that's what the forum is for!


Congrats on the OT flip!


Looking at an AB763 Super Reverb, you are looking pretty good, except your bias is a little "hot" @ -36 volts (remember that the more negative, the more the tube is shut off compared to the schematic ( that wants -52 volts per the Super Reverb AB763 schemo) +/-10 or 15 volts on the bias *is* a fairly big deal, even though 50 volts +/- on a plate generally isn't. I play jazz, quieter, cleaner, so myself, I tend to bias cold with the idea that my 6L6 tubes will last longer, run cooler.


You can't really perform a topnotch bias adjustment without measuring *current* through the output tubes, and for that (or for *any* current measurement) you have to interrupt the circuit. The most widely used method is to insert a precision 1 ohm resistor between each 6L6 cathode and ground. Then you measure voltages across that resistor. Using ohm's law, every 1/1000th of a volt (millivolt) dropped across that resistor is 1 milliamp of tube current. Most people around here would reco you add such a resistor under your 6L6 tubes so that you make a better bias measurement and make sure it's within range. To get fancier, most people around here bring that junction of the cathode and the top of the 1 ohm out to tip jacks so they can measure bias easily. You have probably seen those "bias adjust" kits for sale...those do nothing but insert that 1 ohm R under the 6L6. The tube does not know anything about a 1 ohm resistor, it does not affect the tube's operation, thus it can be left in.


That said, most Super Reverbs I have encountered have their bias at around -45 volts. It is somewhat ddependent upon the brand of 6L6 used. The 6L6 tubes should not get red plates at -36 volts, but in my experience they will be running hottish. If it were me and I didn't have those resistors I would dial that bias voltage more towards -45. It's far from the best way to measure bias but IMO you want to shade towards the cold side (lower, more negative volts) for the sake of your 6L6 life, especially if you are flying blind (which you are).

Congrats again! Nice to have sound come out!




---Oh and by the way....if as I am implying your bias is a bit too hot and you running a tad bit too much current through your 6L6 tubes, as you crank that bias colder, your output tube plate voltage will RISE a little bit. Not much and nothing to be concerned about, but could be 15-20 volts.

[sluckey]

Where did you get this kit? Are the instructions available on the net?

[coreysan]

I got the entire kit from MojoTone. It's a nice kit, and I've enjoyed every second of it.
The tubes are from JJ Electronics, caps are Orangedrops.

[PRR]

> If it's wrong, flip 'em. End of story.

Aside from the OT primary, you can flip phase at

- Power tube grids
- (sometimes) driver tube grids
- OT secondary (tho this may have other consequences)

When kit documentation is derived from a prototype, it is very easy for the draftsman to get something mixed-up. Once upon a time, HeathKit could afford to hire ordinary people (probably their workers, families and friends) to build from a mimeograph and prove the documentation right (or wrong). Sales today are smaller (the cheapest kits can't compete with Asian pre-made gear) so the documentation may be flawed.

And the OT could be wound either way, but MojoTone probably has "captive" suppliers who know to wind the production iron just-like the prototype. (But there could be a design-change before production which did not get told to the prototype's documenter.)

There are also an infinite number of OTHER ways an amp can go into a full-power squeal due to wiring errors. But power section (OT and tubes) phase wrongness is not uncommon.

[eleventeen]

"Aside from the OT primary, you can flip phase at- Power tube grids- (sometimes) driver tube grids- OT secondary (tho this may have other consequences)"


All true, but the best (IMO) place, assuming you have built the whole amp up, are the plate leads:


Using a real Fender OT or a salvaged OT with only a single, untapped secondary, although nobody really cares, we would *prefer* to have the green output wire go to ground and the black to the "hot" of the speaker jack. Presumably we would wire it up this way before applying any power necessary to perform our motorboat test.

Using a tapped tranny, we really do not want to undo our nice switch or barrier strip wiring to properly phase the OT

Changing the PI wiring is probably equivocal, unless, again, you have done it neatly.

Power tube grids: That method would be fine, usually those wires are single leads off a tag board and are easily moved.

I remain a voter for the method of NOT cutting the OT plate leads to length, leave them long & sloppy, tack them to the plates. Least confusing way. Complete wiring, prove the phasing. Flip if necessary. Dress the wires nicely with proven phase. Done, and all the nice wiring/dressing you have done everywhere else remains as is.

[PRR]

> the best (IMO) place

Yes. Most convenient. (Which is why, as you say, you don't cut OT primary wires short until you test.) Least side-effects.

I was pointing out that there are other places where "correct phase" can go astray, even when following a picture. The artist may have gotten "neat". Or just gone cross-eye staring at the prototype too long. Of course the builder can do the same mistakes with the best intentions.