Sizing the output transformer for a single ended tube power amp design

[ricothetroll]

Hi,

I'd like to build a tube guitar amp with a single-ended output stage. As I love the EL84 sound, I decided to put two of those in parallel, for a ~8-10W power rating. In Richard Kuehnel's book "Guitar Amplifier Power Amps", I've read that when you set up N tubes in parallel in such a configuration, you get N times the plate current flowing into the o/p trafo and thus N times the power.

Taking a look at Hammond's catalog, I saw that a 5W transformer can handle an idle current of 45mA, about what I need for one EL84, but for a current of 90mA, the trafo is then rated 20W ! Its weight also increases by the same amount : 1,1kg > 4,5kg. That a HUGE trafo ! Here's the link to the 125xSE series :
http://www.hammondmfg.com/125SE.htm

Is there an error in Hammond's catalog ? Or is there a error in my reasoning ? Will I really have to put a trafo that big for two tiny EL84s ?

Best regards.

Eric

[Seljer]

Yep, theres a reason why most the singled ended tube amps are little 5 to 10 watt things. If you start scaling the power up components just start getting too big.

[PRR]

> 5W transformer ...45mA, ..., but for a current of 90mA, the trafo is then rated 20W

Yes. These *SE irons are really 5K impedance. Same impedance, double the current, you imply double the voltage, thus _four_ times the power.

In fact you won't change the supply voltage, won't get double voltage. You will cut the impedance in half, double current, same voltage, double power.

The *SE irons are targeted at 5K but are also workable at 2.5K or 10K. So you have your half-impedance.

What I would expect to work: 125DSE. Yes, it is rated 70mA. If you flow 90mA, the inductance will drop. But you do not need full inductance, since you are working at lower impedance. Also you are not expecting full 20Hz Hi-Fi response; guitar is usually better with some iron-strain in the lowest notes.

No, the 125DSE will not "burn up" at 90mA. The DC resistance is small, the DC heat is well inside what a 2x3" core will shed.

Alternatively, look for "super Champ" OTs rated for either 6V6 or _6L6_ and typically 8K/7K and 4K/3.5K nominal impedance. Your two-EL84 is pretty-near one 6L6 in terms of horsepower and loading.

JFYI: two tubes parallel is "dumb" because with two tubes you "could" run push-pull, use half the iron and half the power filtering and lower distortion. Push-Pull is such an "obvious" win-win-win that very few commercial amps were ever built paralle-SE. (Yes, guitarists are different.)

[R.G.]

As an aside that's somewhat related.

The best distortion amplifier I ever built for myself was built from the insides of a Magnavox hifi (not stereo  ) console. It used a couple of 12A?7s and one 6AQ5. The 6AQ5 is the little brother to the 6BQ5 and its EL84 ilk. I got this by poking through the goods at a used-furniture shop, buying the console for $25, removing the chassis with a borrowed screwdriver, then selling the now-lighter console back to the furniture shop for $10.

I'm sure it never got over 3W at any time. But the quality of the watts far surpassed lots of louder brothers. And today, it's simple to hook up a kilowatt or so of "booster" to the thing if you like that.

IMHO, there's not a lot of need to worry with paralleling SE tubes. A Celestion (for example) will likely do between 96 and 100db with one watt of drive.

[ricothetroll]

Hi,
Thanx a lot for your great answers !
I"m giving a though to the single-ended solution because I love the sound of Vox AC4, and decided to increase the power because I'm afraid those 4W would not give me enough clean headroom at band levels...
I think I'll give the 125DSE solution a try ! At worst I'll get an amp for LOUD bedroom playing
Best regards.
Eric

[PRR]

> 6AQ5 is the little brother to the 6BQ5 and its EL84 ilk.

The numbering suggests that, but no.

6V6 was a great US tube. When minis came in there was call for a 6V6-like device in the mini bottle. 6AQ5 is a slightly downrated 6V6 on 7 pins.

EL84 was a great Euro tube. Lots more gain (and a bit more output) than 6V6/6AQ5. For US marketing it got the 6BQ5 number.

Significant difference around the knee, the near-overload behavior.

Yes, good old console amps make fine guitar amps with minor mods. I cringe every time someone finds one and immediately guts it. Those consoles were sold 90% on cabinets, true, but 10% on sound and the old guys knew what sounded good to the ear.

> afraid those 4W would not give me enough clean headroom at band levels...

If you play with a drummer, it's tough; and if 4 Watts is not enough then 8 Watts probably won't bring total happiness. It takes more than double the Power to make a real difference.

I've seen 13 Watts (nearly clean) from one 6550 and a 125ESE. I did it with 600 Volts; I do not advise this. Power supply cost and complexity became a real problem. And while double Watts is not a big difference to the ear, double Voltage is a BIG difference to the finger.

[R O Tiree]

> EL84 was a great Euro tube...

Was? It still is

[brett]

Hi
because the 50Hz or 60Hz of line power is about the lowest frequency that a regular guitar puts out, and because power transformers are efficient at transferring that frequency, and because it is the low frequencies that determine the size of the output transformer - a power transformer of the same power rating tells you how big and heavy you need an output transformer to be.
That's why PTs and OTs are a similar size in larger amps. (in small amps the PT is larger because a few W of power is needed for heaters etc as well as signal)
Note that if you don't mind some bass roll-off - say at 100Hz instead of 50Hz - the size of the transformer can be greatly reduced (weight reduced by half in this example)
cheers

[R.G.]

because the 50Hz or 60Hz of line power is about the lowest frequency that a regular guitar puts out, and because power transformers are efficient at transferring that frequency, and because it is the low frequencies that determine the size of the output transformer - a power transformer of the same power rating tells you how big and heavy you need an output transformer to be.

Close, and a good way to think of it. This applies only if there's no significant DC through the windings, or if the windings are arranged to "cancel" the resulting magnetic fields. That means, in short, that a class B push-pull is remarkably close. One other caveat. Lowest note on a guitar in standard E tuning is 82.407Hz, and in dropped-D, 73.41. So you do pick up a little more power handling because a guitar's lowest frequency is a bit higher than power line. On bass, it's worse, because E-tuned four-string bass is an octave lower, so you get lower power handling.

This is only the power handling on the lowest notes; power handling goes up directly with frequency (if you can ignore certain secondary effects), so if your preamp cuts some bass or boosts treble, you get away with a smaller OT for the same power, or more power with the same OT.

As you note:

Note that if you don't mind some bass roll-off - say at 100Hz instead of 50Hz - the size of the transformer can be greatly reduced (weight reduced by half in this example)

And this is one good illustration of why SE OTs are so big. For the same power, they only use 1/4 of the possible magnetic working potentials inside the iron, so they wind up being about 4X the mass for the same power. Fully half the iron's capability is used up in "resisting" the DC that SE has to have. The other half is lost because you can't swing both directions with the DC offset. Class (a)b pushpull makes much better use of the iron.

@Paul: Dang! That's what I get for remembering what my ham friends said instead of looking it up!

[PRR]

> a power transformer of the same power rating tells you how big and heavy you need an output transformer to be.

As a guide, yes. But (as R.G. says).....

Audio transformers should have low distortion on their lowest note. In small power transformers (under 1000W), large distortion is normal and acceptable.

Good hi-fi iron should be low distortion somewhat below power frequency. In "good" hi-fi you expect the output iron to be *bigger* than the power iron (even though the PT carries more total power than the OT). That's partly snootiness: I've had quite fine hi-fi with OT smaller than PT. OTOH bench-test or disco-test with solid woofer will show the small iron corrupting the deep bass.

Guitar is 82Hz, well above 50/60Hz power tone, and distortion is not bad, may be good ("false bass" overtones phatten-up the sound with less weight). Guitar OTs are almost always smaller than PTs.

Changing the frequency makes a real difference in the iron required. Aircraft ran 400Hz power because it made all iron (alternators and transformers and motors) far smaller.

As R.G. says, SE requires the OT to back-fight the one tube to get a both-ways output; energy is *stored* in the iron, not just transferred. He says 4X the iron per Watt and observation suggests 2X; I suspect very few SE OT designs can afford to go 4X and instead give a bit less bass/distortion performance than the usual P-P OT.

-------
> what my ham friends said

Hams generally do not dig the fine details of audio tubes.

Historically: there was a patent on the Power Pentode. This led to a series of tubes which were much more efficient than power triodes, but still had soft knees and large screen power waste. RCA wished to avoid paying patent fees, threw several other tricks together and called it "beam tetrode", got sharp knee and low-low screen power. Eventually the patents expired or were traded, the difference between "pentode" and "beam tetrode" seemed to lessen.

Pretty much, if you stay below 80% power (most hi-fi) or beat it to death (highly clipped ham speech), you don't care which tube. But the classic low-cost Power Pentodes like EL34 and EL84 have a zone around max clean power where they do sound "different" from the beam tetrodes. And that's exactly where small-medium guitar amps are mostly used.

FWIW: lots of players "love" EL84 both P-P and SE; SE 6V6 Champ is a Rock Classic; I hear very little love for 6AQ5 amps. While 6AQ5 is clearly based on 6V6, classic 6V6 guts are as big as a 6AQ5 bottle, so they must have re-designed the guts to get it in. Maybe they squeezed the mojo out.