advertised amp power output ratings vs actual power tube output (when calculating plate current etc)

Discussion in 'Amp Tech Center' started by owlexifry, May 12, 2021.

  1. owlexifry

    owlexifry TDPRI Member

    Age:
    33
    Posts:
    19
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia
    last night i played through a mates ENGL savage 120 and noticed it only has x2 output power tubes (2 x KT88)
    this popped up a lot of questions in my mind.

    KT88 is typically rated for 40-45w max plate dissipation.
    if biased at 66% max plate dissipation, that’s (2 x 29.7) = 60W

    isn’t the ENGL savage 120 supposed to be a 120W amp?

    am i missing something?
    what is going on here?

    the same could be discussed regarding marshall amps with EL34 tubes.
    EL34 max plate dissipation = 25w.
    if biased at 66%, thats 16.5w per tube.
    so really, a 50w el34 marshall is only capable of up to ~33w output.
    or a 100w el34 —> ~66w

    on my 2204/jcm800 50w clone, i took out the el34 and replaced them with kt77 tubes.
    these are rated for 35w max plate dissipation.
    with these i biased them up to 60-65% and which calculates to around 45w. still not even the full ‘50w ouput’.

    my 40w SLO-inspired build has 2 x 6L6 power tubes.
    6L6 max plate dissipation = 30W
    can easily bias these up to 20W each (66%)
    and this actually achieves the 40W output (40w output transformer)

    i’m starting to believe that advertised output ratings are only a reflection of the maximum power output that the amp’s output transformer is capable of... even though the power tubes will never achieve this power output..

    is this right?
    what else is going on here?

    i just can’t get my head around the ENGL savage being advertised as a 120W amp when it has just x2 power tubes with a 40-45w max plate dissipation rating...

    there’s gotta be something i’m missing..
     
  2. Lowerleftcoast

    Lowerleftcoast Friend of Leo's

    Posts:
    3,675
    Joined:
    Dec 30, 2019
    Location:
    california
    Well... amp companies take liberties. There is no standard when listing the power of an amp.

    66% max plate dissipation is idle current. When the amp is driven with signal it can go past 100%. It can spike well over 100% Max PD but it can't do that very long.

    Then there is wall voltage consumption. A 40w Pro Reverb lists Power Consumption as 175w.

    What wattage would you like the amp to achieve? Let's design a test that can measure that number.;)
     
    owlexifry likes this.
  3. Dacious

    Dacious Poster Extraordinaire

    Posts:
    9,755
    Joined:
    Mar 16, 2003
    Location:
    Godzone
    English amp manufacturers rated amps at pre-clipping. A vintage 100 watt Plexi can crank 160watts. A Vox AC30 can hit 36.

    Marshall rated the JTM45 at 33 watts RMS.

    The Engl may hit a peak 120 watts with everything maxed in one frequency range. How much voltage swing the amp can apply to a clean signal is a predicator of power. How much can it boost it up?

    If you measure a constant varied signal through the amp at maximum gain then measure volts and current at the speaker terminals it's likely you'd find at some point power would flash up to the rated peak.

    Most Deluxe Reverbs probably put out 15-20 real watts but if pushed with the right guitar it might just touch 22. That's way past a pair of 6V6 spec.

    In reality speaker efficiency might have a lot to do with how loud you perceive an amp to be. Older speakers from the 70s might have ratings in the 95db 1 watt, 1 metre SPL range.

    Just putting 98 db speakers in would double the perceived volume. All of a sudden a 20 watt amp could be as 'loud: as a 40 watt amp.eveb if it's character is different.
     
    owlexifry likes this.
  4. radiocaster

    radiocaster Poster Extraordinaire

    Posts:
    8,586
    Joined:
    Aug 18, 2015
    Location:
    europe
    A 60W RMS output is about 120W peak-to peak.

    But it's more complicated than that. Power output is not the same as idle plate dissipation ratings. To measure the output, you have to run a sine wave to the point of distortion. The output transformer also plays a role.
     
    owlexifry likes this.
  5. Jon Snell

    Jon Snell Tele-Holic

    Posts:
    839
    Joined:
    Aug 31, 2015
    Location:
    Jurassic Coast. Great Britian.
    60W RMS = 60 X 1.414 = 84.84W Music Power as the HiFi boffins say.
    120W X 0.707 = 84.84W so I would conlude that 120W is in fact the peak rating.

    Many years ago, B Babani published a valve data book and stated that two EL34s in class Ab2 with a 750v HT could realize 120Watts. Maybe they were working on that and using old time rugged Mullard valves, not new production with smaller anodes.

    Push pull amplifiers do not normally run with 66% plate current, there is no need in a good design as it just wastes power, it's not Class A.
     
    owlexifry likes this.
  6. 2L man

    2L man Tele-Meister

    Age:
    61
    Posts:
    455
    Joined:
    Nov 23, 2020
    Location:
    Finland
    Simply said in a push pull circuit other tube makes first wave half and other tube makes second half so voltage doubles and when power is produced the voltage is multiplied with current so power quadruples.

    However audio tube PP amps operate in A/B to reduce distortion and it reduce efficiency so two PP tube power comes less than four times what one tube can produce.
     
    owlexifry likes this.
  7. VintageSG

    VintageSG Friend of Leo's Ad Free Member

    Posts:
    4,579
    Joined:
    Mar 31, 2016
    Location:
    Yorkshire
    Lies, damned lies, statistics and power ratings.
    Vox rate a few of their solid state amps at greater than their wall consumption, which is a neat trick. A 60W amplifier with 36W wall consumption is over unity, and to be applauded. Vox aren't alone. It may be the case that it can hit the stated power on a transient peak with the smoothing caps dumping their guts.
    It's an interesting exercise, for limited values of interesting, to work through an amp, tracing the signal, to max it out at various frequencies. Start with a sine wave of 1KHz. Bring the gain up until the gain stage displays clipping, then back it off until the clipping just stops, then do the same with the output stage. measure the power at that point.
    Laney, Marshall, Orange and Vox valve amps are generally rated at ~1% clip/distortion[1], at least in Laneys case where they're not giving the model name as a number based on driver size. The Cub 10 is a solid 12W clean, 14W with a little clipping and somewhere north of 17W with an unstable 'scope trace.
    Then try at 100Hz, 500Hz, 2.5KHz and 5KHz
    For squits and giggles, try a square wave with a 50% duty cycle, then a 10% duty cycle. Don't run a square wave too long though as things can get a bit hot.
    Now crank everything and single shot a square, 1.25V pulse. I bet a penny to a dried turd that that is the measure quoted for some amps.

    RMS at the onset of, or a few % distortion is the true figure ( to me )
    Music Power, if represented honestly, is fine, as long as it's stated.
    Peak Music Power Output can be the transient, one shot measurement, which is optimistic at best, but possible, or it can be a number made up that appears to be possible, or it can be pure numerical willy-waving waffle.

    I've found the spec sheets for the valves to be a good starting point.




    [1] Admittedly, my sample size is limited to amps I've had my head inside. I'm quite happy to be corrected.
     
    hepular, 2L man and owlexifry like this.
  8. owlexifry

    owlexifry TDPRI Member

    Age:
    33
    Posts:
    19
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia
    crikey! some excellent info here! thanks very much for all your responses. you’ve saved me a tonne of searching
     
  9. danlad

    danlad Tele-Meister

    Posts:
    345
    Joined:
    Mar 24, 2015
    Location:
    Here and there
    Isn't there a recipe for 100w from a pair of EL34 in the datasheet?

    Just looked - 800v anode, 400v screen, 11k anode to anode transformer.

    Wonder how many guitar amp manufacturers actually did that? Can't be many!

    Amps never come with the info on quite how they are achieving the numbers they claim, or the distortion figures even if they really are achieving them. I'm doubtful if that would make much sense to me anyway. What does 1% distortion vs 3% sound like? Would I even notice?

    Generally it seems to me the advertised power ratings are just convention to not confuse guitar players. Just enough to let them know which classic amp market segment it's aiming for.

    Single 6V6 or EL84? Tell everyone it's a 5 watter.

    Cathode bias push pull 'big bottle'? Call it 30 watts.

    Fixed bias push pull 'big bottle'? 50 watts.

    You can easily complete the list, dropping the classics into each category as you go!
     
    Last edited: May 13, 2021
    owlexifry likes this.
  10. Paul G.

    Paul G. Friend of Leo's

    Posts:
    3,678
    Joined:
    Mar 17, 2003
    Location:
    Rhode Island
    All amps are loud. Treat the quoted power rating as a rough guide to suggest the amount of clean headroom. It is nothing more. There is no standard for measuring output power, and guitar amp manufacturers never specify how they're measuring, so...
     
  11. LightningPhil

    LightningPhil Tele-Meister

    Posts:
    394
    Joined:
    Dec 1, 2018
    Location:
    Nottingham, UK
    At Culham, UK, there’s a 270MW amp that does deliver full power. It is for a physics experiment. But who cares about the experiment. If your in the building with it, the sound from the conductors alone is awesome! Occasionally explodes and dumps several percent of UK’s grid into the air around in it.
     
    tubedude and owlexifry like this.
  12. printer2

    printer2 Poster Extraordinaire

    Posts:
    7,381
    Joined:
    May 24, 2010
    Location:
    Canada
    Actually I am reading a lot of information that does not give you the answer you need to answer your question. Let's back way up.

    Let's pull up some diagrams.

    [​IMG]
    I am not going to bog down with math, I just don't feel like it. Let us say we have a tube biased at its maximum plate dissipation, a 6L6 at 20W. The dc bias is at the Q point, in the middle of the load line so that you can have as much clean signal up as down (or as speakers go, in or out). A lot of power is just going up as heat and we have our maximum output amplitude as the sine wave on the left side.

    But what if we cool down the bias as in the next diagram? We are dissipating less power but one portion of the waveform is chopped limiting how much clean signal we get out.


    [​IMG]

    But if we throw a bigger signal at it the portion above the Q point can get pretty large. Larger than the above diagram. Now because we have two tubes the one side of the signal gets clipped by the one tube, the other side gets clipped by the other tube. But when we combine the power out we get a lot more power than if we had the two tubes operating at 20W each.

    [​IMG]

    This is for a transistor amp, they bias at a much lower power percentage than a tube amp. Ignoring the crossover section (you take what you get when you look for diagrams) you can see that the power dissipated with no signal is basically nothing. This is called Class B. We do not bias our amps in class B as the crossover distortion is annoying (in radio transmitters class B is used). So we bias both devices a little on, even with no signal, as in the second diagram.

    Getting back to your example with the output being biased at say 60%, this leaves more room to send power through the tubes rather than being dissipated in them. One thing that is not said often if at all, the colder you bias the tubes the more power you can get out of the amp. Rather than bias it at 60% how about we bias it at 30%? The plate remains cooler so we can push more power through the tube. So rather than an amp making 100W when biased at 60% it might be able to put out 140W if biased at 30%.

    But you run into crossover distortion. And unless you are blasting power out of the amp continuously (say elevator music going to a big complex with a lot of speakers, always distorted due to clipping, would never hear the crossover distortion) the crossover distortion bothers you because it sounds gritty. Mind you sometimes we like a little bit of grit in the sound, part of the Marshall 18 Watt amp design.

    So how can an amp make more power than being dissipated at idle? We are operating in Class AB, where you get more power "through" the amp than what you dissipate in the amp. You can increase the bias up to 80% but you will be making less power than you could at 60%. A lot more is going on as far as load lines and the like but this is the stripped down explanation. Hope it makes sense.
     
  13. printer2

    printer2 Poster Extraordinaire

    Posts:
    7,381
    Joined:
    May 24, 2010
    Location:
    Canada
    Tube amps used to be quoted at 5% or there abouts. It is where you can actually see some clipping of a sine wave on an oscilloscope. Datasheets can have between 2% and 10% listed for power output. A table radio with a 25L6 putting out 2W may be at 10% distortion but a hifi designed tube like the 6973, used in Wurlitzer organs, is rated at 24W at 2% distortion.
     
    Abu Twangy and owlexifry like this.
  14. tubedude

    tubedude Tele-Meister

    Posts:
    102
    Joined:
    Jul 26, 2013
    Location:
    east georgia
    I wish all amps quoted power at RMS values with the SPL recorded at 1 Meter with a sweep generator.

    That gives the user expected clean headroom, freq response, and actual volume.
     
    Paul G. likes this.
  15. Jowes_84

    Jowes_84 Tele-Meister

    Posts:
    261
    Joined:
    Mar 18, 2018
    Location:
    Kraut-Island, Chiemsee
    Check John Geloso Amps - Italian from the 20-60s
    He had some cool, powerful two tube designs. I have a thing for these, and own a lot… only 3 working at this point ;)
    They always stated a nominal rating and a max rating. Initially Mic/Phono amps… they can sound like Marshalls turned up.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.