Insulation stress in tube output transformers

That sounds a bit much for Class-AB/B push-pull configurations.

Even with square wave drive, one tube is fully conducted and since the center tapped primary acts as an auto transformer, the opposite tube will be at twice the battery voltage Vb. Thus, a bass amplifier with two EL34 at Vb=800 V delivering 100 W would have at most +1600 V at the non-conductive anode and never dropping below +0 V.

However, in Class-C, one tube is fully conducting, then the current stops, before the other tube starts to conduct, causing back-EMF problems before the other tube starts to conduct.

Traditional AM transmitters are essentially Class-C stages, in which the power supply voltage Vb varies according to the audio waveform through a big modulation transformer.

The problem is in inductive kick-back of the transformer if the load is too light or lost. It is far more than twice of the DC feed.

Class C is suited for single-frequency tuned amplifier use. The topic was an audio amplifier.

Right - and it was the modulation transformer which needed the protection. The RF part was in Class C and its plate (anode) supply was modulated with the audio. Tetrodes and pentodes needed some modulation in the screen grid supply also. The needed audio power is half of the RF amplifier input power.

In my old AM transmitter, a Champion spark plug did a nice job in protecting the modulator pair of 807's feeding another pair of 807's in the RF final.

--

-TV

On a sunny day (Sun, 22 Feb 2015 14:44:33 +0200) it happened Tauno Voipio wrote in :

Were you driving the 2 807 with a transformer in zero bias mode?

Zero bias mode has the screen grid and normal grid connected, no negative bias.

I remember somebody doing an audio amp with 2 x 807 in balance that way, sounded very good.

The 807 was quite similar to a 6L6GC, wasn't it? I never got to play much with the big tubes. This was due as much to availability in my part of the world as to financial constraints. IIRC the PA amp in the hospital where I worked in another part of India used multiple pairs of 807s to pump out something like

200W.

On a sunny day (Sun, 22 Feb 2015 20:52:31 +0530) it happened "Pimpom" wrote in :

I don't know those 'merrican numbers for tubes very well. I was in Europe at that time, and the 807 was very much avaiable, and still is:

I did not use that tube myself, but the PE1/100 penthode:

the socket was almost as expensive as the tube.

EL34 was used for audio a lot IIRC in Europe.

I build a nice linear (for 7 MHz SSB) with the PE1/100.

No - I did not dare to do it. Later, with two pairs of EL500's made a bass guitar amplifier in screen-driven configuration. It needed some series resistors in the control grid lines, to limit control grid dissipation.

Actually, the amplifier was not too linear, but I think that the iron-wire folks like it, and that's what they are seeking with the current tube amplifier boom.

--

-TV

IIRC, 807 is the transmitting version of 6L6, which was originally steel-cased. 6L6GC is the same in a glass bulb.

The sockets are different: 6L6's use octal, and 807 uses a 5 pin socket.

Jan: The Philips marking for 807 was QE06/40.

--

-TV

On a sunny day (Sun, 22 Feb 2015 20:29:54 +0200) it happened Tauno Voipio wrote in :

I remember the QQE06/40 used for 2 meters with a Lecher system. There was some army dump set that used that IIRC.

This was typical for those days:

What is the difference beteen a QE06/40 and a QE06/50? I see QE06/50 sold here as 807?

Or did you mean QE06/50?

Not quite: the 6L6, -G (glass, ST "coke bottle" bulb), -GA (improved) and -GB (straight bulb) are all very similar. I forget what the difference is between G and GA, but between these four versions, the specs (in particular, plate dissipation max) are all very similar.

-GC is the improved version (straight "GB" bulb) with "full" plate dissipation ratings.

Then there's the versions that are incrementally fancier still, like 5881, KT66, and so on. But those are generally not called 6L6s because the characteristics do differ.

In addition, the 807 has much more internal shielding, resulting in lower Cg1-p.

The 6BG6 is essentially a 6L6 with plate cap connection, for voltage standoff purposes (the 807 could be used for horizontal deflection, I suppose -- and probably was in the early sets, but I suppose it's always been more expensive?).

Not sure offhand what the total genealogy of the 6L6 family is, but I believe it's at least a few more than just these.

Someone should write a side-bar about it. Like the "History of the 741" sidebar in AoE2. :-)

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

You're right. It seems that there were both markings, but the more powerful QE06/50 seems to be the correct one. My 807's were from USA, and I did not have the 1972 Philips pocketbook at hand.

QQE06/40 is a different beast, for VHF push-pull use. I have seen last one at work in an old VOR beacon in 1992.

--

-TV

** But I and many others have MEASURED it!!

It does not occur with a matched, resistive load - but DOES with a speaker load, simulated speaker load or even open circuit.

The phenomenon is called "spiking" and is the main cause of socket arc over and transformer burnout. EL34 output stages more prone than 6L6.

As I posted earlier, many guitar amp makers fit high voltage diodes to stop this. Music Man, Mesa Boogie, Peavey and Ampeg, to name a few.

.... Phil

On a sunny day (Sun, 22 Feb 2015 23:27:19 +0200) it happened Tauno Voipio wrote in :

Yes, those tubes make comebacks, nothing wrong with those.

But, now I have this power amp for 2.4 GHz:

about 60 to 70 W linear.

Still have to build it into a box with cooling. No more heaters :-)

nd so >prevents large plate to plate AC voltages in operation. "

I think that ability to go negative and allow for a hair more cone exceursi on is one of the reasons for that sweet tube overdrive sound, when that's t he way it is used. Now they usually get the overdrive at the input. However some actually mike the amp. Doesn't B B King do that with Lucille ? Or use d to ?

Someone told me I was wring on that but the laws of physics are on my side. The cone is in motion and that is causing back EMF. Transistors short it out but tubes let it fly. This means it is not square waves like a solid st ate amp, not exactly anyway.

And that's why there are guitar amps with tube outputs, and digital effects . Worked on one last week. Thought to myself "now there's something ytou do n't see in the TV biz".

Did those amplifiers contain some kind of "autobias" circuits, i.e running in class AB at low signal levels, going to class B at high power ? How did that "autobias" circuit work when _severely_ overdirven ? Going to Class-C ?

I have replaced several 100 W 4xEL34 or 6xEL34 sockets but these stages operated only at Vb = +400 V.

I have never seen these diodes in Geloso or Marshall amplifiers.

Phil Allison

** My tests were with 100w Marshalls using EL34s - loaded with a 4x12 cab or a choke in series with resistive dummy load. Then turn all knobs to 10. ** Or Fender, which Marshall simply copied and substituted EL34s.

Part of the problem is coupling caps to the grids of the output tubes charging under heavy overdrive - due to grid current.

You cannot just add such diodes to a Marshall or most Fenders either, cos they will fail short if the HT fuse opens or the standby switch on a Fender is operated while being played loud.

.... Phil

** ROTFL ! Plate voltages only go negative when there is "spiking" - the voltage spike last about 50uS, at most. ** No it isn' t- spiking sounds horrible. ** Only at low frequencies - below about 120 Hz. ** Utter drivel - just another stupid, muso's myth.

Most high powered bass amps are SS and have low output impedance - and players love them.

A great many SS guitar amps have high output impedance - higher even than tube amps and players are unaware as it is never speced.

... Phil

Yes. It was not intentional and in fact a Zener diode would have been great . but going through a resistor as it was, in class AB part of the time one tube is cut off. The other tube is going into more current usually when you get near max power output and that will cause a higher total current throo ugh the output tubes when the opposite tube cannot conduct less than zero. This will raise the cathode voltage and send the thing into class B when yo u load it enough. Overdrive it and you will get to class C but probably bur n up the cathode resistor.

However, when you understand the nature of crossover distortion you realize this is not all that bad. When you cross over that center voltage you are moving so fast that nobody can hear it. Well actually some can. And others can hear the more subtle differences it makes in the tonal quality because of the negative feedback and the damping effect. But the fact is it is most noticable at low levels.

What's more, most speakers have obscene levels of distortion, mostly odd or der. Has to do with the linearity of the strength of a magnetic field as th e actign object passes throught the range. Actually it is a fascinating sub ject studied in depth at certain speaker makers' research departments. Bott om line, once you gt to those higher levels that speaker is gong to creatre a hell of alot of harmonic distortion. Fact of life.

Don't want the distortion ? May I suggest a pair of Quad ESL-63s. thiose ha ve super low distortion for speakers. If you can hear the difference on any thing, that's them. Of course if you can pry someone's cold dead fingers of f a pair you will have to pay their widow more than the cost of a modest bu ngalow.

(is "widow" supposed to be capitalised ?)

ke >last about 50uS, at most. "

You can hear 50uS. At least "they" can.

The "Horrible" is due to come out on effct boxes and maybe even some pedals . you will see it right there between COMP2 and LARGE HALL1.

Ummm, those laws of physics. Speaking of which, how come you can't retrofit those amps with high volatge diodes ?

Yaeh, another "muso yth" like Bob Carver ? He proved his point about amps a nd he did it in the hifi paradigm where they do not want distortion. These musicians seem to like distortion. And that distortion must be just right.

I will have to agree that most DC coupled amps all sound pretty much the sm ae when it comes to output. When it comes to guitar amps and overdrive, ver y few people are able to use an old Fender Champ with a single 6V6 miked in to a venue's system. Takes a little doing. And even mixing boards by themse lves sound good sometimes. Got the gain, you got it.

Bottom line, they pay. Not that well but they pay.

layers love them. "

What they love is irrelevant. you know that if you took one output transist or (or a bank) out of their amp they might like it better.

I got a half a mind to do just that. All I need is the genny, a resistor an d a scope.

have super low distortion for speakers. If you can hear the difference on a nything, that's them. Of course if you can pry someone's cold dead fingers off a pair you will have to pay their widow more than the cost of a modest bungalow.

electrostatics make sweet speakers, but no balls, you still need subwoofers with them.

NT