Leo and his 420V 6V6

[Helmholtz]

That cathode drive circuit is pretty slick .

Yes, that's a cascode circuit, where the transistor controls the tube current.
I guess that avoids the risk of runaway with that high plate voltage (being B+ minus transistor voltage).

[g1]

I guess that avoids the risk of runaway with that high plate voltage (being B+ minus transistor voltage).

The MM cathode voltage runs around 80V. So that still puts amps like the 2475-150 in 'violation' of the datasheets by 120V.

[Helmholtz]

The MM cathode voltage runs around 80V.

Yes, but that's not my main point.

What I meant is that with this hybrid cascode circuit the transistor has full control of the tube current and DC bias.
- The collector output is high impedance and the cathode input is low impedance, so a perfect match.
In other words, the transistor acts as a current source to the tube.
- The high impedance cathode load also means strong cathode NFB.
- Even a small unwanted increase in cathode current will be countered by an increase of the transistors emitter voltage, thereby lowering transistor bias and current.
- The grid is connected to a fixed low voltage source having very low impedance.

I think all these effects help to stabilize the tube against runaway.

[g1]

I think all these effects help to stabilize the tube against runaway.

Agree. I also believe low screen voltage is a factor. So these are all good points as far as 'how do they get away with it'.
But as far as I know the tube is still subject to the same datasheet, and that was a point I was trying to make.

[Helmholtz]

Agree. I also believe low screen voltage is a factor. So these are all good points as far as 'how do they get away with it'.
But as far as I know the tube is still subject to the same datasheet, and that was a point I was trying to make.

Yes, low screen voltage and not exceeding other tube limits as well seems essential.
Same with operating EL34s at a B+ of 800V (which is actually in spec).
Still I'm hesitant regarding a general principle.

[loudthud]

Note that the numbers you see on a Tube data sheet are usually grouped under a broad type of service such as Class A1 Power Amplifier, Push Pull Class AB1 Amplifier etc. Some tubes also have a type of service named Vertical Deflection Amplifier which has different ratings. You would expect different ratings if the category was Push Pull class B Cathode Drive.

[K Teacher]

No. Leo didn't get lucky. The use of higher than recommended voltages, was a calculated risk, to get more power out of the 6V6's (treated more like an earlier metal envelop 6L6 = 19W tube).

But the downside of this practice, is that it will shorten the 6V6's useful life. Now, if the average home user plays his amp for about 8 hrs./week, it will take a couple of years for it to need power tubes replacement. By this time, the amp is out of the warranty period.
Most of the other amp manufacturer's just followed Fender's practices. That strategy could have been acceptable back in the 60s and 70s, when you easily could get good USA 6V6s for not too much money.

If you ever had the opportunity to work on a Valco built amp, you can verify that they didn't abuse the 6V6's voltages and many of them still have the original tubes, in good operating condition.

Now, consider this – a pair of 6V6’s will put out about 14W, usually with an 8K OT. The Deluxe squeezes approx. 22W out of the 6V6's running at supply voltages of 420VDC – 6K6 OT (more like 6L6 territory). However, double of 14W, that is, 28W is +3dB SPL. It's documented by psychoacoustics studies, that a +3dB SPL increase, is the maximum increase that cannot be perceived by the average human ear. In other worlds, most people can't tell the difference.

So, why sacrifice the 6V6's? I guess is bragging rights. You can measure it, but you can't HEAR IT !!! (I mean the loudness… distortion-wise is another story).
That explains why there are a lot of 5W amps but not many 10W ones! Same goes for 15W and 30W or 50W and 100W ones… Doubling the power doesn't cut it.

I know… every time that this comes up, there's always someone that replies saying that they can hear +1dB difference !!!

My suggestion is -- tested this yourself with two amps side-by-side and get to your own conclusion…
ALERT – it doesn't count, if you have double of the power PLUS a speaker that has a +3dB or higher sensitivity than the lower power amp !!!

[Leo_Gnardo]

We had a go-round about a decade back, either here or The Amp Garage, discussing the fate of 6V6's in high voltage amps. Most specifically the Jim Kelly models that put around 500V on a quad of 6V6's. Who showed up to the party but Jim Kelly himself, recommending NOS Sylvania or RCA for minimum trouble. Here it is a decade later, still no winner for currently made 6V6's that will survive 420V much less 500. JJ's seem to minimize outright failures in challenging conditions but ones I encounter recently all jingle and rattle. Guitarists don't like that. Tried a few from PSVane and expensive TAD. Too many spectacular failures with either, sparks, buzzes and baleful purple glows. Not a fan of Russian tubes. What's left?

[Helmholtz]

[FONT=Arial]
. It's documented by psychoacoustics studies, that a +3dB SPL increase, is the maximum increase that cannot be perceived by the average human ear. In other worlds, most people can't tell the difference.

What is your source?

To my knowledge it is generally accepted in psychoacoustics that the minimum noticeable difference in SPL is around 1dB, depending on frequency and absolute level.
IIRC, a 3dB difference in SPL means a difference of 23% in loudness, so should definitely be audible.

[g1]

Of all the amps that are hard on tubes, Fender's that run 6V6's are not high on my list. When those black glassed Sovtek's were around in the 90's, we saw a fair bit of failures of those particular tubes. Other than that, replacing blown 6V6's in Deluxe Reverb amps is not especially commonplace for me.

[K Teacher]

What is your source?

To my knowledge it is generally accepted in psychoacoustics that the minimum noticeable difference in SPL is around 1dB, depending on frequency and absolute level.
IIRC, a 3dB difference in SPL means a difference of 23% in loudness, so should definitely be audible.

I don’t remember from the top of my head, but this information is available on the net; just look around...
What I remember from school, is that the max. and min. amp frequency response points are defined where the output drops 3bB in relation to a reference point (400Hz or 1KHz).
Why do you think this is defined this way?

[K Teacher]

To Leo_Gnardo's point, if you look at the Deluxe RI schematic, it has all kinds of circuit protections to guard the amp from power tubes catastrophic failures, that were not in the original design…

Regarding the JJ's 6V6's, there are several treads that claim this is not a real 6V6, but some other Russian tube similar to the 6V6, but more robust, so it can take the abuse… now tone is another story…

[Chuck H]

We had a thread here not long ago regarding high-ish voltages and actual output from the bigger bottle 6L6 tubes in Fender amps. I believe the subject amp was a home brew that was eating 6L6 screens. It was speculated that Leo and Fenders designers may have limited available current from the PT in their 4x6L6 amps to make the circuits work as intended. It seems likely that a similar ideal could have been used for the high voltage 6V6 amps. The high voltage makes for a great dynamic response when run clean and the current limiting protects the tubes under higher gain conditions. Best of both worlds and something I can only qualify as well considered and good design. In other words, I don't think it was a calculated risk. I think it was a calculated design.

Like g1 I don't hear much about the high voltage Fenders eating 6V6's. I've had a lot more trouble with big bottles in home brews that don't even get close to their rated specs before giving you a light show.

[Helmholtz]

I don’t remember from the top of my head, but this information is available on the net; just look around...

I couldn't find anything confirming your statements.
Please show credible sources.

My sources are:
1) German electro-acoustics standard literature: Zollner/Zwicker "Elektroakustik", confirming the 1dB audibility threshold on page 289.
2) http://sengpielaudio.com/calculator-levelchange.htm , confirming that + 3dB SPL means +23% in loudness.

What I remember from school, is that the max. and min. amp frequency response points are defined where the output drops 3bB in relation to a reference point (400Hz or 1KHz).
Why do you think this is defined this way?

Certainly not because a difference of 3dB is "inaudible" .

[g1]

3db is commonly referred to as the 'half power point'.